Patient Safety Culture in Primary Care: Natasha Verbakel

Patient safety culture in
primary care
Natasha Verbakel
Patient safety culture in primary care
Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht.
PhD Thesis - with a summary in Dutch.
University of Utrecht.
ISBN: 978 - 94 - 6203 - 800 - 4

Author: N.J. Verbakel
Coverdesign: M. Fraters
Printed by: CPI Koninklijke Wöhrmann B.V., Zutphen, the Netherlands
The research in this thesis was financially supported by the Dutch Ministry of Welfare and
Sport (Object number: 1154503).
All rights reserved. No part of this thesis may be reproduced without prior permission of the author.
Patiëntveiligheidscultuur in de eerste lijn

(met een samenvatting in het Nederlands)
Proefschrift
ter verkrijging van de graad van doctor aan de Universiteit Utrecht

op gezag van de rector magnificus, prof. dr. G.J. van der Zwaan,
ingevolge het besluit van het college voor promoties
in het openbaar te verdedigen op donderdag 19 maart 2015
des ochtends 10.30 uur
door
Natasha Johanna Verbakel
geboren op 18 april 1984 te Exeter, Verenigd Koninkrijk

Promotoren: Prof.dr. Th.J.M. Verheij
Prof.dr. C. Wagner
Copromotoren: Dr. D.L.M. Zwart

Dr. M. Langelaan
TABLE OF CONTENTS
Chapter 1 General introduction 7
PART I Patient safety culture in primary care 21
Chapter 2 Measuring safety culture in Dutch primary care: psychometric 23

characteristics of the SCOPE-PC questionnaire
Chapter 3 Exploring patient safety culture in primary care 43
PART II Patient safety culture interventions in general practice 59
Chapter 4 Improving patient safety culture in primary care: a systematic 61

review
Chapter 5 Cluster randomized, controlled trial on patient safety improvement in 77

general practice: a study protocol
Chapter 6 A cluster randomized trial on the effects of patient safety culture 93

interventions in general practice
Chapter 7 How does it work? An interview study on improving patient safety 117
culture in general practice
Chapter 8 General discussion 135
Summary 153
Samenvatting 161
Acknowledgements 169
Curriculum Vitae 173

CHAPTER 1
General introduction
“I am a solo practitioner, without assistant or other staff. Therefore, naturally there are
hardly any adverse events as everything goes through me, delightfully.”
CHAPTER 1
(GP; argumentation not to participate in our study)
“As a GP, you should always ask yourself: is it safe enough for my staff to say what they
want to say? You will not always be certain about that. Although we do try to. You see that
people say things and dare to say stuff. Even things about which you think, well, amazing
that you say that. I’m not sure I would dare to say that to my boss.
So, considering, it seems safe.”
(GP, during interview one year after our intervention)
“I think that, since we had the workshop, we all improved, or at least things are set up. We
already had a sort of reporting procedure. But due to the workshop, there came a sort of
awareness in the whole team.”
(Assistant, during interview one year after our intervention)
These citations, - one from the invitation reply form and two from post intervention
interviews - give a wonderful glimpse of the versatile considerations regarding patient safety
and culture in general practice. The importance of patient safety in the care of patients
is naturally indisputable. Though, as a subject for policy and research, patient safety only
gained momentum since the last two decades and, in primary care it has just gotten in the
picture. Since 2008, patient safety in primary care has become a priority in Dutch health care
policy. In order to start up and engage all different primary care professions in patient safety
activities, the ministry of Health, Welfare and Sport launched a platform (Zorg voor Veilig;
Care for Safety).1 This platform facilitated regular meetings with professional associations
in primary care to discuss the development of patient safety management and necessary
tools. Patient safety culture improvement was recognized as an essential topic leading to
the need for tools for improving culture in primary care practices. At that time, a Dutch
questionnaire to assess patient safety culture had been validated only in general practice.2
This thesis focuses on patient safety culture in primary care. In short, culture is described
as ‘the way we do things around here’.3 Below, the concepts of patient safety and patient
safety culture will be introduced. Also, the primary care setting will be described. Finally,
the objectives and outline of this thesis are delineated.
8 CHAPTER 1
PATIENT SAFETY
Several definitions of patient safety have been described. The Institute of Medicine (IOM)
CHAPTER 1
defines patient safety as the ‘freedom from accidental injury’. The World Health Organisation
(WHO) speaks of ‘prevention of errors and adverse effects to patients associated with health
care’.4 A more elaborate definition was described by Wagner & van der Wal: ‘The (almost)
absence of (the risk of) patient harm (physical/mental) that is caused by not acting according
to the professional standard of care providers and/or failure of the health system’.5 Safety
science in itself is not a new concept, it originates from high-hazard industries such as the
petrochemical industries. After the establishment of healthcare being a high-hazard industry
as well, safety issues got more into focus. Particularly the IOM report ‘To Err is Human’
spurred to action and accelerated patient safety as a policy priority and research subject.6
This report stated that approximately 44.000 people died in the United States as a result of
medical errors. Subsequently, other countries conducted similar studies. In the Netherlands
medical record studies were carried out in 2004, 2008 and 2011/12, the latter showing that
7.1% of patients encountered care related harm and 2.6% of those (968 patients) may have
died of potential preventable harm.7-9 Errors are suggested to be the result of both active
and latent failures, meaning that an error is not only the result of a fallible person (mostly
at the end of the chain), but also from errors occurring earlier in the process.10 Woolf et al.
called this ‘a string of mistakes’ also stating that most of the time, events are not isolated
but the result of a sequence of mistakes.11 Incident analysis, therefore, was even more
important. Medical errors can be an important source of experiences and knowledge which
can be used to improve patient safety. And although the shift from the personal blame
approach to the system approach has been embraced and encouraged, still, there is a need
for a non-punitive safety culture to get healthcare workers to report.
PRIMARY CARE AND PATIENT SAFETY
At first most of the patient safety research and policy recommendations were focused on
hospitalised care. The last decade, this focus was broadened to include primary care as well.
Primary care is a key part of healthcare and a strong primary care is associated with better
healthcare outcomes.12, 13 In the Netherlands the gatekeeper systems supports a strong
primary care. More than 90 percent of healthcare is performed in primary care, emphasizing
its importance and relevance.14 In 2013, the number of GPs in the Netherlands was estimated
at 11075.15 Primary care is easily accessible and GPs serve as a gatekeeper for specialised
care. Besides general practice, primary care consists of a broad array of different disciplines
such as speech therapy, dental care, physiotherapy and midwifery. Unlike secondary care,
General introduction 9
practices in primary care are relatively small. Practices can be single-handed or group
practices, either mono- or multidisciplinary. Managerial and organisational tasks are mostly
carried out by professionals themselves, sometimes supported by nurses or administrative
CHAPTER 1
assistants. An important difference between primary and secondary care is the type of
care provided which might influence the kind of errors that could occur. Though, the risk of
serious harm might seem smaller in primary care compared to incidents in hospitals, it is of
significance to give attention to patient safety in primary care as well because of the large
amount of patients contacts.16 Several studies investigated the occurrence of adverse events
in primary care settings. In Spain, 773 adverse events were identified in 48 practices.17 Gaal
et al. examined 1000 medical records in the Netherlands and found that in 2.5% of patient
contacts an incident had occurred.18 An equivalent record study in other Dutch primary
care professions showed low percentages of harm: 0.8% in dental care practices, 2.5% in
midwifery practices and 1.0% in paramedical practices.19 Communication was stated to be
one of the prominent causes for incidents.11, 17, 18, 20-22
PATIENT SAFETY CULTURE
The safety culture of an organization is described as: ‘‘the product of individual and
group values, attitudes, perceptions, competencies, and patterns of behaviour that
determine the commitment to, and the style and proficiency of, an organization’s health
and safety management. Organizations with a positive safety culture are characterized by
communications founded on mutual trust, by shared perceptions of the importance of safety
and by confidence in the efficacy of preventive measures.’’ (Advisory Committee on the
Safety of Nuclear Installations).23 In essence, culture is “the way we do things around here”.3
Sammer et al. performed a review on safety culture and classified culture properties into
seven subcultures: leadership, teamwork, evidence-based, communication, learning, just,
and patient-centered.24 A constructive, open culture is seen as a facilitator in the success of
implementations of safety interventions.25, 26 Besides reporting the rates of adverse events
and preventable deaths, the IOM report gave recommendations for improvement, including
to develop a culture of safety. Likewise, the National Patient Safety Agency stated the
creating of a positive safety culture as the first step in their ‘Seven steps to patient safety’.27
In addition, a Dutch report by the former director of Shell in The Netherlands stated that
it poses a safety risk when acknowledging mistakes is considered taboo.28 An open safety
culture is also suggested to be positive for healthcare outcomes.29, 30
10 CHAPTER 1
CULTURE OR CLIMATE?
Measuring the prevailing culture is often one of the first steps undertaken when intending
CHAPTER 1
to improve patient safety and culture. The attempts to assess culture fuelled the debate
whether one should speak of ‘culture’ or ‘climate’ when conducting a survey. To describe
the distinction Guldenmund uses the framework by Schein31 on organizational culture.32 It
distinguishes three levels: basic assumptions, espoused values and artefacts. The first level
concerns the ‘core’ of the culture, these are the underlying convictions, and is equated
with culture. The second level, the espoused values, are the manifestations of culture;
the attitudes, which are compared with climate. Artefacts are any other manifestation of
culture, such as clothes and symbols.
Guldenmund states that culture is explanatory to climate, why things are done in a certain
way. From a methodological perspective, it is said that with a quantitative approach,
i.e. questionnaires, only the superficial climate can be captured and, that longitudinal,
qualitative methods, are necessary to examine culture.33 The usability and relatively low
cost, however, makes a survey an attractive instrument to assess patient safety culture. In
the literature, both terms - climate and culture - are used interchangeably. In this thesis we
will use the term culture because at the start of our patient safety research in primary care
the term ‘culture’ was more commonly used. Though, we acknowledge that a survey will
only tap into the espoused values, as called above ‘climate’.
ASSESSING AND IMPROVING PATIENT SAFETY CULTURE
Data on the prevailing culture are not only useful as a starting point but also provide
outcome measurements to be able to evaluate interventions.34 Nieva and Sorra 23 state that
the assessment of culture can be used to:
1. diagnose safety culture to identify areas for improvement and raise awareness about
patient safety;
2. evaluate patient safety interventions or programs and track change over time;
3. conduct internal and external benchmarking;
4. fulfil directives or regulatory requirements.
Today, several instruments have been developed and validated for various healthcare
settings.3, 23, 35, 36 After a thorough literature review the Linneaus Euro-PC group identified two
tools to be most useful for assessing safety culture in primary care settings37: the primary
care version of the Manchester Patient Safety Framework (MaPSaF)38, 39 and the Medical
Office short version40 or the Nursing Home version of the Agency for Healthcare Research
and Quality (AHRQ) survey.41
The MaPSaF is both an assessment and a discussion tool based on typologies of
organizational communication described by Westrum38 and later expanded by Kirk et al.42-
44
It describes nine patient safety culture dimensions according to five increasing maturity
CHAPTER 1
stages: pathological, reactive, bureaucratic, proactive and generative, see Figure 1. Nine
dimensions are described for primary care:
1. Overall commitment to quality;
2. Priority given to patient safety;
3. Perceptions of the causes of PSIs and their identification;
4. Investigating patient safety incidents;
5. Organisational learning following a patient safety incident;
6. Communication about safety issues;
7. Personnel management and safety issues;
8. Staff education and training about safety issues;
9. Team working around safety issues.
A practice may not only use it to identify the current state of affairs and to discuss strengths
and weaknesses, but it is also a useful tool to learn about differences in perspectives
between staff.44 The latter is very important because during the subsequent discussions staff
can elaborate on their viewpoints and get closer to each other.
Generative
The nirvana of all
Proactive safety
organisations in
Organisations
Bureaucratic which safety is an
that place a high
integral part of
value on
Organisations everything they
improving safety,
Reactive that are very do. In a
actively invest in
paper-based and generative
continuous safety
Organisations safety involves organisation,
improvements
Pathological that only think ticking boxes to safety is truly in
and reward staff
about safety after prove to auditors the hearts and
who raise safety
Organisations an incident has and assessors minds of
related issues
with a prevailing occurred that they are everyone, from
attitude of ‘why focused on safety senior managers
waste our time on to frontline staff
safety’ and, as
such, there is
little or no
investment in
improving safety
Figure 1. Culture ladder42
12 CHAPTER 1
The series of AHRQ patient safety culture surveys started with the Hospital Survey on
Patient Safety Culture (HSOPS) consisting of 12 dimensions: Teamwork across hospital units;
teamwork within units; hospital handoffs and transitions; frequency of event reporting;
CHAPTER 1
non-punitive response to error; communication openness; feedback and communication
about error; organisational learning – continuous improvement and supervisor/manager
expectations and actions promoting patient safety; hospital management support for
patient safety; staffing; overall perceptions of safety.45 Smits et al. translated and validated
the HSOPS into a version for Dutch hospitals.46 This version was followed by the adaption
of this questionnaire for general practice, the SCOPE (Systematisch Cultuur Onderzoek
Patientveiligheid Eerstelijn).2 SCOPE is an acronym for systematic culture inquiry on patient
safety.
The SCOPE for general practice consists of 43 items divided over eight dimensions:
1. Handover and teamwork;
2. Support and fellowship;
3. Communication openness;
4. Feedback about and learning from error;
5. Intention to report events;
6. Adequate procedures and adequate staffing;
7. Overall perceptions of patient safety management;
8. Expectations and actions of managers.
In addition, two outcome questions are included on the assessment of a patient safety grade
and frequency of error reporting.
In search for effective culture interventions two reviews found leadership walk rounds
and broad programmes to have a positive effect on safety culture.47, 48 Team training and
communication were stated to be key in improvement efforts.48 Improving patient safety
culture is a complex intervention. Healthcare in itself is a complex organization as both
practices and interventions cannot be fully standardized and succes, among others, depends
on the context in which it is deployed. Singer and Vogus aptly described it as ‘it is not a
matter of taking a pill or flipping a switch’.49 The success of the intervention and direction
of the results are highly dependent on the context.50,51 Moreover, the failure to tailor
complex interventions to the practice where it will be implemented limits its effectiveness.52
Designing an intervention study using mixed-methods will provide insight in the success of
the intervention and sheds light on how the intervention worked.
OBJECTIVE AND OUTLINE OF THE THESIS
The general objective of this thesis was to measure patient safety culture and to assess the
CHAPTER 1
effect of patient safety culture interventions in primary care. Figure 2 depicts the outline of
the thesis and it shows an overview of the chapters and associated objectives.
Two studies were conducted. Part I of this thesis concerns patient safety culture assessment
in Dutch primary care. The following research questions were addressed:
1. What is the validity of the patient safety culture questionnaire SCOPE-PC?

2. What is the prevailing patient safety culture in Dutch primary care professions and are
there differences between professional groups?
Research question 1 concerned the validity and reliability of the patient safety culture
questionnaire in all professions of primary care. The process of adapting the questionnaire
to a generic multidisciplinary questionnaire is described in chapter 2. This questionnaire,
SCOPE-PC, will add to available tools in primary care for patient safety activities. It will
allow for the exploration and comparison of the prevailing culture in Dutch primary care
professions. Research question 2, subsequently, regarded the prevailing culture in this setting
and possible differences between professions. In chapter 3 we present the perceptions of
nine primary care profession groups on patient safety culture. Also, the differences between
the professions are examined.
Patient safety culture
Assessing Improving Experimenting
Chapter 2 Validating a Chapter 4 Searching Chapter 6 Studying the

generic patient safety and examining existing effect of two culture
culture questionnaire tools that affect safety interventions in general
for primary care culture practice
Chapter 3 Exploring the Chapter 7 Examining

prevailing patient safety the factors that explain
culture in primary care the effect of two
culture interventions in
general practice
Figure 2. Outline of the thesis: chapters and objectives
14 CHAPTER 1
Part II of this thesis concerns the SCOPE Intervention Study and focusses on improving
patient safety culture in general practice. The majority of intervention tools and research
are focused on secondary care. Consequently, little is known about how to create an open
CHAPTER 1
and positive culture in primary care practices. Therefore, the following research questions
were addressed:
1. Which tools are available for patient safety culture improvement in the primary care
setting and what is their effectiveness?
2. What is the effect of two culture interventions in general practice?
3. How can the effect of the two culture interventions be explained?
To learn about existing patient safety interventions in primary care (research question 3) and
to choose our improvement strategy a systematic review was conducted which is presented
in chapter 4. Together with the knowledge on assessment this was used to design a study
to test two culture interventions in general practice. In chapter 5 the study protocol of our
trial on the effect of two patient safety culture interventions in general practice is reported.
Here, both interventions are described in detail, i.e. a patient safety culture questionnaire
(SCOPE) and a practice based workshop. Also, the methodology of the trial is elaborated on.
Research question 4 regarded the effect of the interventions in daily practice. In chapter
6, the quantitative results of the trial, performed in thirty general practices are presented.
The last research question concerned our understanding of the culture interventions. The
qualitative results of the trial are described in chapter 7. The trial was conducted as a mixed-
methods study using interviews to gain more insight in how the intervention worked in the
different practices. In conclusion, we discuss our research and implications in the general
discussion in chapter 8. Chapter 9 summarises the thesis in English and Dutch.
REFERENCES
1. ZorgvoorVeilig. Available at: www.zorgvoorveilig.nl. Accessed 5,1, 2014.
2. Zwart DLM, Langelaan M, van de Vooren RC, Kuyvenhoven MM, Kalkman CJ, Verheij TJ, Wagner
C. Patient safety measurements in general practice. Clinimetric properties of ‘SCOPE’. BMC family
CHAPTER 1
practice 2011;12:117.
3. Scott T, Mannion R, Davies H, Marshall M. The Quantitative Measurement of Organizational
Culture in Health Care: A Review of the Available Instruments. Health services research
2003;38(3):923-45.
4. WHO. Patient safety. Available at: http://www.euro.who.int/en/health-topics/Health-systems/
patient-safety. Accessed 7/21, 2014.
5. Wagner C, van der Wal G. Voor een goed begrip. [For a good understanding]. Medisch Contact
2005;47:1888-91.
6. Kohn L, Corrigan J, Donaldson M. To err is human: building a safer health system. Washington
D.C.: National Academic Press; 2000.
7. de Bruijne MC, Zegers M, Hoonhout LHF, Wagner C. Onbedoelde schade in Nederlandse
ziekenhuizen: dossieronderzoek van ziekenhuisopnames in 2004. Amsterdam: EMGO Instituut
en NIVEL; 2007.
8. Langelaan M, Baines RJ, Broekens MA, Siemerink KM, van de Steeg L, Asscheman H, de Bruijne
MC, Wagner C. Monitor zorggerelateerde schade 2008: dossieronderzoek in Nederlandse
ziekenhuizen. Utrecht: NIVEL; 2010.
9. Langelaan M, de Bruijne MC, Baines RJ, Broekens MA, Hammink K, Schilp J, Verweij L, Asscheman
H, Wagner C. Monitor zorggerelateerde schade 2011/2012 [Monitor care related harm]. Utrecht:
NIVEL en EMGO+ Institute/VUmc; 2013.
10. Reason J. Human error: Models and management. Brit Med J 2000;320(7237):768-70.
11. Woolf S, Kuzel A, Dovey S, Phillips R. A string of mistakes: the importance of cascade analysis in
describing, counting, and preventing medical errors. Annals of family medicine 2004;2(4):317-26.
12. Kringos DS, Boerma W, van der Zee J, Groenewegen P. Europe’s Strong Primary Care Systems
Are Linked To Better Population Health But Also To Higher Health Spending. Health affairs
2013;32(4):686-94.
13. Starfield B, Shi L, Macinko J. Contribution of Primary Care to Health Systems and Health. The
Milbank quarterly 2005;83(3):457-502.
14. Wiegers T, Hopman P, Kringos D, de Bakker D. Overzichtstudies. De eerste lijn. Utrecht: NIVEL;
2011.
15. van Hassel DTP, Kasteleijn A, Kenens RJ. Cijfers uit de registratie van huisartsen. Peiling 2013
[Figures from the registration of general practitioners. Poll 2013]. Utrecht: NIVEL; 2014.
16. Gandhi TK, Lee HL. Patient Safety beyond the Hospital. N Eng J Med 2010;363(11):1001-3.
17. Aranaz-Andres JM, Aibar C, Limon R, Mira JJ, Vitaller J, Agra Y, Terol E. A study of the prevalence of
adverse events in primary healthcare in Spain. European journal of public health 2012;22(6):921-
5.
18. Gaal S, Verstappen W, Wolters R, Lankveld H, van Weel C, Wensing M. Prevalence and
consequences of patient safety incidents in general practice in the Netherlands: a retrospective
16 CHAPTER 1
medical record review study. Implementation science 2011;6:37.
19. Harmsen M, Giesen P, Martijn L, Mettes T, Verstappen W, Nijhuis-van der Sanden R, Wensing M.
Patiëntveiligheid in de Nederlandse eerstelijnszorg anno 2009. [Patient safety in Dutch primary
care anno 2009]. Nijmegen: Scientific Institute for Quality of Healthcare, UMC St Radboud; 2009.
CHAPTER 1
20. Makeham M, Dovey S, Runiciman W, Larizgoitia I. Methods and measures used in primary care
patient safety research. Results of a literature review. World Health Organization; 2008.
21. Zwart DLM, Steerneman AHM, van Rensen ELJ, Kalkman CJ, Verheij TJM. Feasibility of centre-
based incident reporting in primary healthcare: the SPIEGEL study. BMJ Quality and Safety
2011;20(2):121-7.
22. Dovey SM, Meyers DS, Phillips RL, Green LA, Fryer GE, Galliher JM, Kappus J, Grob P. A preliminary
taxonomy of medical errors in family practice. Quality and safety in health care 2002;11(3):233-8.
23. Nieva VF, Sorra J. Safety culture assessment: a tool for improving patient safety in healthcare
organizations. Quality and safety in health care 2003;12(Suppl II):ii17-23.
24. Sammer CE, Lykens K, Singh KP, Mains DA, Lackan NA. What is Patient Safety Culture? A Review of
the Literature. Journal of nursing scholarship 2010;422:156-65.
25. McCarthy D, Blumenthal D. Stories from the sharp end: case studies in safety improvement. The
Milbank quarterly 2006;84(1):165-200.
26. Zohar D, Livne Y, Tenne-Gazit O, Admi H, Donchin Y. Healthcare climate: a framework for measuring
and improving patient safety. Critical care medicine 2007;35(5):1312-7.
27. National Patient Safety Agency. Seven steps to patient safety in primary care. London: NPSA-NHS;
2006.
28. Willems R. Hier werk je veilig, of je werkt hier niet. Sneller beter - De veiligheid in de zorg. [Here
you work safe, or you do not work here. Faster better _ Safety in healthcare]. Den Haag: Shell
Nederland; 2007.
29. Haynes A, Weiser T, Berry W, Lipsitz S, Breizat AH, Dellinger,,E,Patchen, Dziekan G, Herbosa T,
Kibatala P, M, Merry A, Reznick R, Taylor B, Vats A, Gawande A. Changes in safety attitude and
relationship to decreased postoperative morbidity and mortality following implementation of a
checklist-based surgical safety intervention. BMJ Quality and Safety 2011;20(1):102-7.
30. Mardon RE, Khanna K, Sorra J, Dyer N, Famolaro T. Exploring relationships between hospital
patient safety culture and adverse events. Journal of patient safety 2010;6(4):226-32.
31. Schein E. Organizational Culture and Leadersip. San Francisco: Jossey-Bass; 1985.
32. Guldenmund FW. The nature of safety culture: a review of theory and research. Safety science
2000;34(1-3):215-57.
33. Flin R, Mearns K, O’Connor P, Bryden R. Measuring safety climate: identifying the common
features. Safety science 2000;34:177-92.
34. Sexton J, Helmreich R, Neilands T, Rowan K, Vella K, Boyden J, Roberts P, Thomas E. The Safety
Attitudes Questionnaire: psychometric properties, benchmarking data, and emerging research.
BMC health services research 2006;6(44).
35. Colla JB, Bracken AC, Kinney LM, Weeks WB. Measuring patient safety climate: a review of
surveys. Quality and safety in health care 2005;14:364-6.
36. Singla AK, Kitch BT, Weissman JS, Campbell EG. Assessing Patient Safety Culture: A Review and
Synthesis of the Measurement Tools. Journal of patient safety 2006;2:105-15.
37. Linneaus Euro-PC. Available at: www.linneaus-pc.eu. Accessed April 12th, 2013.
38. Westrum R. Cultures with requisite imagination. In: Wise JA, Hopkin VD, Stager P, editors.
Verification and validation of complex systems: human factor issues New-York: Springer-Verlag;
CHAPTER 1
1993. p. 401-16.
39. Parker D. Managing risk in healthcare: understanding your safety culture using the Manchester
Patient Safety Framework (MaPSaF). Journal of nursing management 2009;17:218-22.
40. Medical Office Survey on Patient Safety Culture. Agency for Healthcare Research and Quality,
Rockville, MD. Available at: http://www.ahrq.gov/professionals/quality-patient-safety/
patientsafetyculture/medical-office/index.html. , July 2014.
41. Nursing Home Survey on Patient Safety Culture. Agency for Healthcare Research and Quality,
Rockville, MD. Available at: http://www.ahrq.gov/professionals/quality-patient-safety/
patientsafetyculture/nursing-home/index.html. , July 2014.
42. Kirk S, Parker D, Claridge T, Esmail A, Marshall M. Patient safety culture in primary care: developing
a theoretical framework for practical use. Quality and safety in health care 2007;16(4):313-20.
43. Parker D. The Manchester Patient Safety Framework. Primary care. Manchester: NPSA - NHS.
44. Parker D. Manchester Patient Safety Framework (MaPSaF). Facilitator guidance. Manchester:
NPSA - NHS.
45. Sorra J, Nieva V. Hospital Survey on Patient Safety Culture. Westat, Rockville: Agency for Healthcare
Research and Quality; 2004.
46. Smits M, Christiaans-Dingelhoff I, Wagner C, van der Wal G, Groenewegen P. The psychometric
properties of the ‘Hospital Survey on Patient Safety Culture’ in Dutch hospitals. BMC health
services research 2008;8:230.
47. Morello RT, Lowthian JA, Barker AL, McGinnes R, Dunt D, Brand C. Strategies for improving patient
safety culture in hospitals: a systematic review. BMJ Quality and Safety 2013;22(1):11-8.
48. Weaver SJ, Lubomksi LH, Wilson RF, Pfoh ER, Martinez KA, Dy SM. Promoting a Culture of Safety as
a Patient Safety Strategy. A Systematic Review. Annals of internal medicine 2013;158(5):369-74.
49. Singer SJ, Vogus TJ. Safety climate research: Taking stock and looking forward. BMJ Quality and
Safety 2013;22:1-4.
50. Stevens DP, Shojania KG. Tell me about the context, and more. BMJ Quality and Safety
2011;20(7):557-9.
51. Pawson R, Tilley N. Realistic evaluation. London: SAGE Publications Ltd; 1997.
52. Corrigan M, Cupples M, Smith SM, Byrne M, Leathem CS, Clerkin P, Murphy AW. The contribution
of qualitative research in designing a complex intervention for secondary prevention of coronary
heart disease in two different healthcare systems. BMC health services research 2006;6(90).
18 CHAPTER 1
CHAPTER 1
PART I

CHAPTER 2
Measuring safety culture in Dutch primary care:

psychometric characteristics of the SCOPE-PC
questionnaire
Published as:
Verbakel NJ, Zwart DLM, Langelaan M, Verheij TJM, Wagner C.

Measuring safety culture in Dutch primary care: psychometric characteristics of the
SCOPE-PC questionnaire.
BMC Health Services Research 2013;13:354
ABSTRACT
Background
Patient safety has been a priority in primary healthcare in the last years. The prevailing
culture is seen as an important condition for patient safety in practice and several
tools to measure patient safety culture have therefore been developed. Although
Dutch primary care consists of different professions, such as general practice, dental
care, dietetics, physiotherapy and midwifery, a safety culture questionnaire was only
CHAPTER 2
available for general practices. The purpose of this study was to modify and validate
this existing questionnaire to a generic questionnaire for all professions in Dutch
primary care.
Methods
A validated Dutch questionnaire for general practices was modified to make it
usable for all Dutch primary care professions. Subsequently, this questionnaire
was administered to a random sample of 2400 practices from eleven primary care
professions. The instrument’s factor structure, reliability and validity were examined
using confirmatory and explorative factor analyses.
Results
921 questionnaires were returned. Of these, 615 were eligible for factor analysis.
The resulting SCOPE-PC questionnaire consisted of seven dimensions: ‘open
communication and learning from errors’, ‘handover and teamwork’, ‘adequate
procedures and working conditions’, ‘patient safety management’, ‘support and
fellowship’, ‘intention to report events’ and ‘organisational learning’ with a total of
41 items. All dimensions had good reliability with Cronbach’s alphas ranging from
0.70 – 0.90, and the questionnaire had a good construct validity.
Conclusions
The SCOPE-PC questionnaire has sound psychometric characteristics for use by the
different professions in Dutch primary care to gain insight in their safety culture.
24 CHAPTER 2
BACKGROUND
One of the main focuses in patient safety research is patient safety culture. A supportive
patient safety culture is seen as an important condition for patient safety.1 Patient safety
culture refers to values, attitudes, norms, beliefs, practices, policies, and behaviours about
safety issues in daily practice. In essence, culture is ‘‘the way we do things around here’’.2
In a review, Sammer et al. identified seven subcultures of patient safety culture: leadership,
teamwork, evidence based, communication, learning, just, and patient-centred.3 Gaining
CHAPTER 2
insight in the prevailing safety culture is therefore seen as a first pivotal step towards an
adequate patient safety system.4 Various instruments have been developed to measure
patient safety culture.2,5-8 They help to identify weak areas in the perceived safety culture
and thus enable designing tailored improvement strategies.
In recent years, increasing attention has been given to patient safety in primary care.9-14
Primary care is directly accessible and consists of a broad array of professions, e.g. dental
care, general practice, physiotherapy, midwifery, speech therapy. Despite this wide range
of care, practices have many similarities in organisational structure. As most practices are
small, managerial and organisational tasks -including safety improvement- are mostly done
by the professionals themselves. Moreover, primary care professionals increasingly work
together in broad healthcare centres, collaborating in disease management programmes
and consulting one another in managing the care of individual patients.
Because of the increase in collaboration within primary care, developing a generic patient
safety culture instrument was desirable. It will enable comparison between different
primary care providers and in a later stage of safety management, may generate exchange
of learning and improvement strategies. As a tool for patient safety culture already exists in
the Netherlands: the SCOPE, it has been developed and validated for general practice only.15
SCOPE is a Dutch acronym for systematic culture inquiry on patient safety. Other primary
care professionals were already familiar with it, therefore, we choose to modify this tool
into a generic questionnaire for all professions in primary care: the SCOPE-Primary Care
(SCOPE-PC).
METHODS
First adjustment to the questionnaire

The Dutch SCOPE questionnaire for general practices, is a modification of the Dutch Hospital
Survey on Patient Safety (HSOPS).15,16
Psychometric characteristics of the SCOPE-PC questionnaire 25

This original SCOPE for general practice consists of eight dimensions:
1. Handover and teamwork (8 items);
2. Support and fellowship (5 items);
3. Communication openness (6 items);
4. Feedback about and learning from error (6 items);
5. Intention to report events (3 items);
6. Adequate procedures and adequate staffing (7 items);
7. Overall perceptions of patient safety management (4 items);
CHAPTER 2
8. Expectations and actions of managers (4 items).
We made adjustments to this SCOPE through an iterative process. First, the research team
revised the terminology of the questionnaire. Secondly, professionals from all primary care
professions assessed the questionnaire individually clarity and applicability to their own
setting. In total 27 professionals (1 midwife, 4 pharmacists, 1 physician, 2 dieticians, 2
physician assistants, 2 physiotherapists, 2 skin therapists, 2 general practitioners, 1 speech
therapist, 2 dental hygienists, 2 exercise therapists, 2, dentists, 2 dentist assistants, 1
occupational therapist, 1 general practice nurse and 1 nurse working in an anticoagulation
clinic) gave feedback by e-mail. Lastly, the research team reached consensus on the version
to be used for the further validation process.
Adjustments were limited to a few changes of terminology, for example ‘general practitioner’
was changed to ‘professional’ and ‘physician assistant’ was changed to ‘support staff’. None
of the original patient safety culture items were deleted. Three questions were added for
routing purposes, where if this question prompted a negative answer the respondent was
not shown the other questions regarding this topic. One question was added: ‘Are incident
reports discussed in meetings on a structural basis? Structural means that it is a permanent
feature on the agenda’.
Besides patient safety culture questions the existing questions about patient safety
characteristics of the practice were included: whether or not events were discussed in an
informal way, the frequency of event reports filled in in the last 12 months and a patient
safety grade for the total practice (answer categories: failing, poor, acceptable, good,
excellent).
The final questionnaire consisted of 43 patient safety culture items (see Table 2). Items had
to be answered using a five-point Likert scale ranging from strongly disagree (1) to strongly
agree (5) or never (1) to always (5). On request of the individual professionals we added the
option ‘not applicable’ to the questions about the practice organisation and collaboration.
Background questions addressed demographics and work-related information, such as how
long and in which profession the respondent had been working in this practice.
26 CHAPTER 2
Data collection and respondents
Data collection for validation of the questionnaire took place from March until May 2011.
An online system managed by the Dutch Practices Accreditation Organisation was used for
collection and storage of the data.17
Eleven primary care professions participated: dental care, dental hygienist care, dietetics,
exercise therapy, physiotherapy, occupational therapy, midwifery, anticoagulation clinics,
general practice, skin therapy and speech therapy. A random sample of 200 members
was drawn from the national databases of each professional association. These members
CHAPTER 2
were asked to participate and to invite colleagues from their own practice too. The key to
sign in to the digital questionnaire was included in the invitation. It was emphasized that
the questionnaire was to be filled out individually. In addition, practices were promised a
feedback report regarding the patient safety culture of their practice.
The selection process differed for one of the professions: the physiotherapists were invited
directly by their professional association. Because of this extra step, a lower response
rate was expected. To anticipate on this, the sample for physiotherapists was doubled to
400. Once enrolled, the inclusion and the following steps were the same as for the other
professions. All practices received a first invitation followed by two reminders with an
interval of three weeks to all the contact persons. Invitations and reminders were preferably
sent by e-mail but if not available by post.
Analyses
Preliminary analyses
As culture is a feature of a group, single-handed practices without employees were excluded
from analyses. In addition, as it takes time to absorb the culture of an organisation, we
excluded respondents with less than half a year experience in their current practice. Further,
respondents with more than five missing values on the patient safety culture items were
excluded. The answer category ‘not applicable’ was not counted as missing. Items that were
negatively worded were recoded so that high scores always reflect a positive response.
Subsequently, distributions of variables were examined to assess response variability and
missing data. Inter-item correlations were studied, as well as Bartlett’s test of sphericity
and the Kaiser-Meyer-Olkin Measure of Sampling Adequacy (KMO) were performed to see
whether a factor analysis could be performed. When Bartlett’s test is significant (p<0.001) it
indicates that the data are appropriate for factor analysis. For KMO a value near 1 indicates
that patterns of correlations are relatively compact and factor analysis should yield distinct
and reliable factors.18 Regarding the rule of thumb of 10 respondents per patient safety
culture item, at least 430 completed questionnaires were needed.19

Factor analyses
As we built on an existing questionnaire, a confirmatory factor analysis (CFA) was performed
to investigate whether the structure of the original SCOPE for general practices could be
confirmed for these data. The chi2 and RMSEA were used as parameters for goodness of fit. A
non-significant chi2 means that the discrepancies between the hypothesized model and the
empirical data are negligibly small and thus indicate a good fit. The RMSEA measures how
well the empirical model approaches the theoretical model. A value of <0.05 is considered
a close fit of the model, a value of <0.08 fair or a reasonable error of approximation, and
CHAPTER 2
values >0.1 are regarded as not acceptable.20,21

To examine whether a different structure would give a better fit to the data, an explorative
factor analysis (principal component analysis, promax rotation) was performed. To determine
how many factors should be retained the eigenvalues and the scree plot were examined.
Also, the total explained variance was taken into account.
Reliability
Internal consistency of the factors was measured using Cronbach’s alpha. A Cronbach’s
alpha of >0.60 indicates that different items measure the same concept.18 A positive rating
for internal consistency is met when Cronbach’s alphas range between 0.70 and 0.95.19 We
also examined the deleted-item reliability coefficients.
Construct validity
For all respondents, sum scores were calculated by obtaining the mean score of all items
within one dimension. One missing value per dimension was allowed. Subsequently,
intercorrelations between dimensions were calculated with Pearson correlation coefficients.
We expected that the various dimensions would correlate moderately as they cover an
aspect of the same construct: patient safety culture. However, the correlations should not
exceed 0.70 because this would mean that the dimensions are too similar and measure the
same concept. Furthermore, correlations of the dimensions with the patient safety grade
were computed. It was expected that all dimensions would have a positive correlation with
the grade.
All Statistical analyses were conducted using SPSS 17.0 and Lisrel 8.8 for the CFA.20
Ethics statement
The Medical Research Ethics Committee of the University Medical Center Utrecht concluded
that no WMO approval for this study was needed.
28 CHAPTER 2
RESULTS
In total, 921 individual questionnaires were returned from 519 practices. 306 questionnaires
were excluded for further analysis: 200 from single-handed practices, 11 from respondents
with less than half a year experience at the particular practice and 94 with more than 5
missing values, resulting in 615 questionnaires eligible for the study. Bartlett’s test was
significant (p<0.001) and the KMO was 0.91 indicating that the data were appropriate for a
factor analysis.
CHAPTER 2
Table 1 gives a description of the study population by gender and age and response rate per
profession. Overall, the age and gender distribution of the sample was representative for
the Dutch professions population, where reference data were available (data not shown).
However, in both dental care and general practice females were overrepresented.
Table 1: Study population by age and gender and response rate per profession
Professions n Response Age % women
rate Mean (SD)

Exercise therapy practices 35 17.5% 42 (9.7) 97.1
Dental care practices 42 21% 42 (11.8) 67.5
Dental hygiene practices 14 7% 40 (10.1) 100
Dietetic practices 18 9% 41 (10.3) 100
Midwifery practices 123 61.5% 37 (10.3) 96.7
Occupational therapy practices 34 17% 41 (8.5) 93.5

Physiotherapy practices 147 36.8% 40 (12.2) 50.4
Anticoagulation clinics 95 47.5% 49 (9.1) 87.8
General practices 16 8% 48 (10.7) 50
Skin therapy practices 24 12% 35 (11.7) 100
Speech therapy practices 67 33.5% 37 (10) 100

Total 615 25.6% 41 (11.4) 81.7
Confirmatory factor analysis

A confirmatory factor analysis showed that the original factor structure did not fit well with
the data (X2=2855.56 df=832, p<0.001). The RMSEA was .06 (90% confidence interval: 0.06-
0.07). Although some authors consider a value of 0.06 a fair fit, the cut-off point of <0.05 is
usually used. Therefore, also an exploratory factor analysis was conducted to examine if a
different structure would fit the data better.

Exploratory factor analysis
Exploratory factor analyses showed the best fit with seven dimensions: (1) open
communication and learning from errors, (2) handover and teamwork, (3) adequate
procedures and working conditions, (4) patient safety management, (5) support and
fellowship, (6) intention to report events and (7) organisational learning. The original
dimension ‘communication openness’ was divided in two dimensions: ‘open communication
and learning from error’ and ‘adequate procedures and working conditions’. Table 2 provides
an overview of the seven dimensions, the number of items with their factor loading, mean
CHAPTER 2
score and standard deviation (SD) per dimension.

Two items did not have a satisfactory loading above 0.40 on any of the factors and one item
loaded on two factors. Two of these items: ‘Staff are afraid to raise questions if something
does not seem right’ and ‘Disciplines work together well to provide the best care for patients’
were deleted from the questionnaire, since the content of these items were covered to a
great extent by other items. As the content of the item ‘Professionals discuss errors that
occurred with other disciplines’ was not covered by other items, and in view of the expected
increase of interdisciplinary collaboration in the near future, it was retained as a separate
item. A few items fell under a different dimensions than in the original SCOPE. The answer
category ‘not applicable’ was retained for the question ‘When one area in this practice gets
really busy, others help out’ and questions regarding collaboration. The seven dimensions
jointly explained 58.9% of variance.
Reliability and construct validity

The internal consistency was excellent with Cronbach’s alphas ranging from 0.70 to 0.90
(see Table 2). Table 3 shows the mean dimension scores with the SD and correlations
between the seven dimensions and the patient safety grade. Overall, the correlations
between the dimensions were moderate to good. The highest correlations were between
‘open communication and learning from error’ and ‘patient safety management’ (r=0.61)
and between ‘adequate procedures and working conditions’ and ‘handover and teamwork’
(r=0.57). ‘intention to report events’ did not correlate with other dimensions (r= 0.11 - 0.18)
except for ‘open communication and learning from error’ (r=0.38). The correlation with the
patient safety grade showed a similar pattern: for all dimensions there was a moderate to
good positive correlation ranging from 0.34 to 0.55 except for ‘Intention to report events’
(r=0.21).
30 CHAPTER 2
Table 2. Mean scores and factor loadings of the items of the SCOPE-PC questionnaire
Item Description Mean SD F1 F2 F3 F4 F5 F6 F7 α
1. Open communication and learning from error

C1 We are given feedback about changes put into place based on 3.95 1.27 0.84
event reports
C2 Staff will freely speak up if they see something that may negati- 4.53 0.65 0.59
vely affect patient care
C3 We are informed about errors that happen in this practice 4.22 0.88 0.86
C4 Staff feel free to question the decisions or actions of those with 4.08 0.89 0.72
more authority
C5 In this practice, we discuss ways to prevent errors from hap- 4.42 0.76 0.69
pening again
C7 Professionals discuss errors that occurred with each other 4.30 0.78 0.73
C9 We are given personal feedback about our own event reports 4.09 0.99 0.66
B4n My supervisor/manager overlooks patient safety problems that 3.96 0.81 0.40
happen over and over
2. Handover and teamwork
F1n Problems often occur in the exchange of information across 3.50 1.01 0.67
disciplines in our practice
F2n The fact that patients are treated by different professionals in 4.12 0.71 0.77
Psychometric characteristics of the SCOPE-PC questionnaire

our practice is causing problems
F3n Disciplines in the practice that we co work with do not coordi- 3.88 0.90 0.85
nate well with each other
31
CHAPTER 2
CHAPTER 2
32
Table 2 continued
F4 There is a good exchange of information between professionals 4.30 0.76 0.52
in this practice
CHAPTER 2
F5 There is a good exchange of information between supporting 4.21 0.72 0.45
staff in this practice
F7n Things “fall between the cracks” when transferring patients 3.89 0.88 0.83
between different disciplines in this practice
F8n Important patient care information is often lost because pa- 4.01 0.85 0.81
tients see different professionals
3. Adequate procedures and working conditions
A5n It is just by chance that more serious mistakes don’t happen 4.34 0.78 0.77
around here
A7n We use more agency/temporary staff than is best for patient 4.40 0.78 0.80
care
A8n Staff feel like their mistakes are held against them 4.23 0.80 0.54
A10n In this practice we work longer hours than is best for patient 3.89 0.92 0.76
care
A12n When an event is reported, it feels like the person is being writ- 4.06 0.80 0.65
ten up, not the problem
A13n We work in “crisis mode” trying to do too much, too quickly 3.80 0.95 0.59
A14n Staff worry that mistakes they make are kept in their personnel 4.17 0.77 0.58
file
A15n We have patient safety problems in this practice 4.39 0.70 0.59
Table 2 continued
B3n Whenever pressure builds up, my supervisor/manager wants us 4.02 0.84 0.43
to work faster, even if it means taking shortcuts
4. Patient safety management
B1 My supervisor/manager says a good word when he/she sees a 3.32 0.96 0.71
job done according to established patient safety procedures
B2 My supervisor/manager seriously considers staff suggestions for 3.96 0.73 0.86
improving patient safety
B6 The actions of my supervisor/manager show that patient safety 3.76 0.88 0.90
is top priority
B7n My supervisor/manager seems interested in patient safety only 4.09 0.74 0.43
after an adverse event happens
5. Support and followship
A1 People support one another in this practice 4.56 0.62 0.90
A2 We have enough staff to handle the workload 3.93 0.94 0.60
A3 When a lot of work needs to be done quickly, we work together 4.18 0.75 0.85
as a team to get the work done
A4 In this practice, people treat each other with respect 4.51 0.63 0.92
A11 When someone in this practice gets really busy, others help out 4.12 0.74 0.79
Psychometric characteristics of the SCOPE-PC questionnaire

33
CHAPTER 2
CHAPTER 2
34
Table 2 continued
Intention to report events
CHAPTER 2
D2 When a mistake is made, but is caught and corrected before 3.56 1.19 0.91
affecting the patient, how often is this reported?
D3 When a mistake is made, but has no potential to harm the 3.59 1.14 0.93
patient, how often is this reported?
D4 When a mistake is made that could harm the patient, but does 4.01 1.04 0.90
not, how often is this reported?
Organisational learning
A6 We are actively doing things to improve patient safety 3.95 0.82 0.62
A9 Mistakes have led to positive changes here 3.97 0.68 0.57
A16 Our procedures and systems are good at preventing errors from 4.00 0.66 0.53
happening
Deleted items
C6n Staff are afraid to ask questions when something does not seem
right
F6 Disciplines work together well to provide the best care for
patients
Seperate item
C8 Professionals discuss errors that occurred with other disciplines 3.55 1.08
The letter ‘n’ in an item-code means that it concers an item in negative wording.
Table 3: Mean dimension scores, correlation with patient safety grade and intercorrela-
tions of the seven dimensions
Mean patient
Dimensions n (SD) safety 1 2 3 4 5 6
grade
1 Open communication 588 4.22 0.44**
and learning from error (0.64)
CHAPTER 2
2 Handover and teamwork 456 3.99 0.43** 0.50**
(0.62)
3 Adequate procedures 457 4.12 0.47** 0.53** 0.57**
and working conditions (0.54)

4 Patient safety 294 3.81 0.55** 0.61** 0.53** 0.52**
management (0.65)
5 Support and fellowship 606 4.26 0.34** 0.40** 0.42** 0.46** 0.39**
(0.60)
6 Intention to report 590 3.72 0.21** 0.38** 0.11* 0.15** 0.17** 0.18**
events (1.03)
7 Organisational learning 609 3.97 0.42** 0.41** 0.38** 0.33** 0.49** 0.54** 0.20**
(0.59)
** Correlation is significant at the 0.01 level (2-tailed)
* Correlation is significant at the 0.05 level (2 tailed)
DISCUSSION
Main findings
Validation of the SCOPE-PC showed that the scale consisted of seven dimensions, slightly
differing from the original SCOPE questionnaire with eight dimensions. The main difference
was that the original dimension ‘open communication’ in the current study was divided in
two dimensions: ‘open communication and learning from error’, and ‘adequate procedures
and working conditions’. Internal consistency and construct validity were good.
Interpretation of findings
It is interesting to note the absence of a correlation between ‘intention to report events’ and
all other dimensions but one: ‘open communication and learning from error’. The absence
of correlation between ‘intention to report’ and most other dimensions may be explained
by a difference in perspective. The questions about reporting relate to actual steps to be
undertaken when an error occurs, they ask about one’s personal intentions: What would
you do if? In contrast, questions regarding collaboration, support, the notion of abiding and
employing the procedures about patient safety relate to how everybody feels or thinks of
the atmosphere in their practice, and is concerned with how this is at the moment.
Another explanation for the absence of correlation could be the fact that reporting is still very
uncommon in primary care. The dimension ‘Intention to report events’ does therefore not
‘behave’ the way the other dimensions do. Additionally, the fact that ‘open communication
and learning from error’ does correlate may indicate that this is an important precondition
for reporting. Subsequently, one would expect that the coherence of all dimensions will
CHAPTER 2
become stronger as safety management activities become common practice.

The uneven distribution of the response rate in the eleven professions was striking. A partial
explanation is that some professionals, like general practitioners, often receive requests to
participate in a study and are therefore less likely to respond. Also, some general practitioners
had already completed the original questionnaire during their accreditation process. The
original questionnaire was already available for them. Furthermore, it is possible that
the interest for patient safety differs between primary care professions. Despite these
differences we considered all professions as one group; similarities in the targeted patient
group, organisational structure and educational level of employees justify this.
Strengths and limitations

The strength of this questionnaire is that it serves all primary care professions with one
generic questionnaire. This facilitates comparison in further research.
This study has some limitations. First, selection bias due to the fact that more innovative
practices and practices that are more enthusiastic about the topic were probably more
willing to participate, could not be excluded. Because data collection was done through a
contact person, we may assume that personal drive and maybe even authority played a role
in participation of individuals. However, for a psychometric study this is less important since
the focus is on clustering of items.
Second, the response rate of 38.4% for individual questionnaires is not very high, yet it is
not unusual for an open population study. The low response could be due to the fact that
because of the organisational structure, primary care professionals have no overhead time
for such activities. On top of this, data collection was hindered because the membership
records of some groups were not up to date. In addition, we were not able to distinguish
the single-handed practices beforehand. Still, with a total of 615 questionnaires the rule of
thumb of a ratio of ten respondents per item is amply met with a ratio of 14:1.
Third, a general drawback of measuring culture with a survey is that it will not capture the
heart of the current culture, for which a more sophisticated method is necessary.23 Options
given are participant observation, interviews and focus groups combined with attitudinal
36 CHAPTER 2
surveys and established cultural assessment tools. Indeed, it would be interesting to
combine the SCOPE-PC questionnaire with qualitative methods in a future study aiming at
describing patient safety culture. However, for professionals themselves, as final users of
the product, a survey has the advantage that it is feasible and easy to use.
Conclusions and implications

In conclusion, this study showed that the SCOPE-PC questionnaire for primary care has sound
psychometric properties. The questionnaire is slightly different from the original SCOPE, but
CHAPTER 2
overall the main part of the factor structure is the same and only two items were deleted.
In future, when sufficient data will be available, it would be interesting to perform cross-
validation of the questionnaire. The use of the questionnaire will enable all professions
in primary care to gain insight in their safety culture status and to take steps from there
to improve patient safety in their practices. In our opinion, the next step in research is to
explore the status and possible differences between professions in the Dutch primary care
regarding patient safety culture.

REFERENCES
2003;38(3):923-45.
CHAPTER 2
2006.
Synthesis of the Measurement Tools. Journal of patient safety 2006;2(3):105-15.
8. Pronovost P, Sexton B. Assessing safety culture: guidelines and recommendations. Quality and
safety in health care 2005;14(4):231-3.
10. Wallis K, Dovey S. Assessing patient safety culture in New Zealand primary care: a pilot study
using a modified Manchester Patient Safety Framework in Dunedin general practices. Journal of
primary health care 2011;3(1):35-40.
11. Hoffmann B, Domanska OM, Albay Z, Mueller V, Guethlin C, Thomas EJ, Gerlach FM. The Frankfurt
Patient Safety Climate Questionnaire for General Practices (FraSiK): analysis of psychometric
properties. BMJ Quality and Safety 2011;20:797-805.
12. de W, Spence,W, Mash,R, Johnson,P, Bowie,P. The development and psychometric evaluation of a
safety climate measure for primary care. Quality and safety in health care 2010; 12;19(6):578-84.
medical record review study. Implementation science 2011;6:37.
14. Gaal S, Verstappen W, Wensing M. Patients safety in primary care: a survey of general practitioners
in the Netherlands. BMC health services research 2010;10:21.
practices 2011;12:117.
17. NHG-praktijkaccreditering®. Dutch College of General Practice-Practice Quality Accreditation.
2011.
38 CHAPTER 2
18. Field A. Discovering Statistics using SPSS for Windows. London: Sage Publications; 2000.
19. Terwee CB, Bot SDM, de Boer MR, van der Windt DAWM, Knol DL, Dekker J, Bouter LM, de Vet
HCW. Quality criteria were proposed for measurement properties of health status questionnaires.
Journal of clinical epidemiology 2007;60:34-42.
20. Hox JJ. Principes en toepassing van structurele modellen [Principles and application of structural
models]. Kind en Adolescent 1999;20(3):200-17.
21. Arbuckle J. AmosTM 18 User’s Guide. Crawfordville: Amos development corporation; 2007.
22. Jöreskog KG, Sörbom D. Lisrel for Windows [Computer Software]. Lincolnwood, IL: Scientific
CHAPTER 2
Software International Inc 2006;8.80.
23. Marshall M, Parker D, Esmail A, Kirk S, Claridge T. Culture of safety. Quality and safety in health
care 2003;12:318.

APPENDIX I Description confirmatory factor analysis (CFA)
A CFA provides a formal statistical test of how well the data fits the predisposed structure of
factors. This fit is indicated by a X2 score and goodness-of-fit indices. The CFA was performed
in Lisrel version 8.8. Because CFA was performed on a data file with missing data, Lisrel
automatically uses Full Information Maximum Likelihood estimation. This method makes
maximal use of all data available from every respondent in the sample. When performing
a confirmatory factor analysis with missing values, Lisrel only gives X2 and the Root Mean
CHAPTER 2
Square Error of Approximation (RMSEA).
Interpretation of results
A non-significant X2 means that the discrepancies between the hypothesized model and the
empirical data are negligible small and thus indicate a good fit. However, the X2 is sensitive
to sample size and therefore little discrepancies can be statistical significant. The RMSEA
measures how well the empirical model approaches the theoretical model. The assumption
is that all models can only be an approximation and therefore a perfect fit cannot be
obtained. A value of <0.05 is considered a close fit of the model, a value of <0.08 fair or a
reasonable error of approximation, and values >0.1 are not acceptable.
40 CHAPTER 2
APPENDIX II Scree plot
CHAPTER 2

CHAPTER 3
Exploring patient safety culture in primary care
Published as:
Verbakel NJ, van Melle M, Langelaan M, Verheij TJM, Wagner C, Zwart DLM.
Exploring patient safety culture in primary care.
International Journal of Quality in Health Care, 2014.
ABSTRACT
Objective
To explore perceptions of safety culture in nine different types of primary care professions
and to study possible differences.
Design
Cross-sectional survey.
Setting
Three hundred thirteen practices from nine types of primary care profession groups in the
Netherlands.
Participants
Professional staff from primary care practices. Nine professions participated: dental care,
CHAPTER 3
dietetics, exercise therapy, physiotherapy, occupational therapy, midwifery, anticoagulation

clinics, skin therapy and speech therapy.
Main Outcome Measure(s)
Perceptions of seven patient safety culture dimensions were measured: ‘open communication
and learning from error’, ‘handover and teamwork’, ‘adequate procedures and working
conditions’, ‘patient safety management’, ‘support and fellowship’, ‘intention to report
events’ and ‘organisational learning’. Dimension means per profession were presented and
multilevel analyses was used to assess differences between professions. Also a so-called
patient safety grade was self-reported.
Results
Five hundred and nineteen practices responded (response rate: 24%) of which
313 (625 individual questionnaires) were included for analysis. Overall, patient
safety culture was perceived as being positive. Occupational therapy and
anticoagulation therapy deviated most from other professions in a negative way,
whereas physiotherapy deviated the most in a positive way. In addition, most
professions graded their patient safety as positive (mean =4.03 on a 5-points scale).
Conclusions
This study showed that patient safety culture in Dutch primary care professions on average
is perceived positively. Also, it revealed variety between professions, indicating that a
customised approach per profession group might contribute to successful implementation
of safety strategies.
44 CHAPTER 3
INTRODUCTION
The establishing of an open, constructive patient safety culture is believed to be important

for improving patient safety. Culture refers to the shared values, attitudes, norms, beliefs,
practices, policies, and behaviours about safety issues in daily practice.1 Several safety
culture surveys were developed to assess safety culture in healthcare.1-4
Until recently, patient safety research mostly addressed hospitalized care whilst a major
part of health care is delivered in primary care settings. Although the risk for patient harm is
lower in primary care, due to the high numbers of patient contacts absolute numbers seem
significant.5 Few studies have assessed patient safety culture in a primary care setting.6-11
Some studies adapted and validated existing questionnaires developed for hospitals, others
CHAPTER 3
developed own questionnaires. Also, the Manchester Patient Safety Framework (MaPSaF),
a discussion tool, was customized.12,13
In the Netherlands, primary care is easily accessible and for medical care it serves as a
gatekeeper to hospital care. Most practices consist of staff members stemming from several
different disciplines but practice sizes are relatively small. When primary care gained more
attention in patient safety research, the Dutch hospital version of the HSOPS14, 15 firstly was
adapted for general practice.16 Subsequently other primary care professional associations
expressed their interest in this assessment tool and patient safety culture as a topic. The
ambition of one questionnaire for all professions in primary care was voiced because
primary care professions increasingly collaborate and work together in one health centre. A
generic questionnaire would be in line with these developments and enhance exchange of
lessons learned.
Following this need, the questionnaire for general practice was modified to a generic primary
care version, the SCOPE-PC.17 SCOPE is a Dutch acronym for systematic culture inquiry on
patient safety in primary care. The aim of the current study was twofold: firstly, to explore
primary care professionals perceptions of patient safety culture and, secondly, to examine
whether there are differences between the primary care professions and in which area.
METHODS
Setting, participants and data collection

The SCOPE-PC was administered to practices in nine primary care profession groups from March
until May 2011. Professions that participated were: dental care (which consists of both dentists
and dental hygienists), dietetics, exercise therapy, physiotherapy, occupational therapy,
midwifery, anticoagulation clinics, skin therapy and speech therapy. Anticoagulation clinics
Exploring patient safety culture in primary care 45

monitor coagulation markers in patients’ blood and ensure correct anticoagulant medication
dosage. Skin therapy is professional care for the treatment of diseased or damaged skin.
In primary care, teams are generally multidisciplinary. In this study we aimed to compare
professions as whole teams. Therefore, within each profession we also included supporting staff,
for instance healthcare assistants and nurses. We did not study individual professional disciplines.
Per profession a random sample of 200 professionals was drawn from their national
professional association member database. The research team invited professionals to
participate and also to invite their colleagues at their practice location to complete an
individual questionnaire. Only for physiotherapy, professionals were approached by the
national physiotherapy association itself, in return asking them to contact the research
team for an invitation and login keys. Anticipating on a lower response by this invitation
procedure, a sample of 400 physiotherapists was drawn. Practices could participate if they
CHAPTER 3
consisted of at least two staff members and the questionnaire was to be completed by staff
that worked at the practice location at least half a year. We chose to keep the minimum
number of employees low so that also small practices could take part in our study. Primary
care consists of many small practices and the sample would not be representative if these
were excluded. The reasoning to set the limit at two was because also two people in one
practice create a way of working and collaborating, in essence, a culture will therefore be
present even in such a small quantity. For collection and storage of data an online system
was used.18
Measurements
The SCOPE-PC questionnaire has been validated and showed sound properties with Cronbach
alpha’s ranging from 0.70-0.90.17 It consists of 41 items divided over seven dimensions:
‘open communication and learning from error’(1), ‘handover and teamwork’(2), ‘adequate
procedures and working conditions’(3), ‘patient safety management’(4), ‘support and
fellowship’(5), ‘intention to report events’(6) and ‘organisational learning’(7). Items were
rated on a five-point Likert scale, ranging from ‘strongly disagree’ to ’strongly agree’ or from
‘never’ to ‘always’. In addition, in dimensions two, three and five some questions had the
answer option ‘not applicable’. Respondents were also asked to rate the level of patient
safety in their own practice between ‘poor’ and ‘excellent’ (Patient Safety Grade, PSG).
Data analysis
Questionnaires from single-handed practices, from respondents working less than half
a year at the practice or responses with more than 50% missing values in patient safety
items were excluded from further analyses. Also, per dimension, respondents scoring “not
applicable” on >50% of the items were excluded.
46 CHAPTER 3
First, the average of the scaled items was computed per profession. Second, for each
dimension a grand mean was calculated over all professions. While calculating the mean
score, one missing item per dimension was allowed. When respondents indicated that
there was no formal management layer in their practice, items concerning patient safety
management were disregarded in the missing count (concerning items in dimension one,
three and four).
To assess perceptions of patient safety culture we examined the mean scale scores of the
seven dimensions per profession group and the PSG. A score of four or higher represents
a positive attitude. Next, to examine whether professions differed from each other we
compared the mean of each profession to the grand mean of the dimension using multilevel
analyses in order to adjust for clustering of respondents in practices. A linear mixed model
CHAPTER 3
with a random intercept was used for the analyses. To interpret differences and their
relevance we adhered to the size of a difference of a half standard deviation (SD).19 All
statistical analyses were conducted using SPSS 20.0.
RESULTS
In total, 906 individual questionnaires were returned from 519 practices, the response
rate was 23.6%. From these, 281 questionnaires were excluded from analysis: 200 from
single-handed practices (mainly exercise therapy, speech therapy and dietetics), 11 from
respondents with less than half a year experience and 70 respondents with >50% missing
values in the patient safety culture items. This resulted in a total of 625 questionnaires
(313 practices) eligible for analysis (see Figure 1). The distribution varied over the seven
dimensions due to the fact that some respondents had >50% of the dimension items
answered with not applicable. The low number of subjects in dimension four resulted from
the majority of respondents not having formal management and therefore not able to
answer the items in this dimension.

CHAPTER 3
48
2200 practices were invited
(200 practices per profession,
400 practices for physiotherapy)
CHAPTER 3
n=906 single respondents
(519 practices)
Exclusion of single handed practices (n=200)
n=706
Exclusion of respondents <0.5 yr working experience (n=11)
n=695
Exclusion of respondents >50% missing values (n=70)
n=625 single respondents

(313 practices)
1. Open 2. Handover 3. Adequate 4. Patient 5. Support and 6. Intention 7. Organisational

communication and teamwork procedures safety fellowship to report learning
and learning and working management events
from error conditions
Exclusion of
>50% not
applicable
n = 611 n = 446 n = 523 n = 302 n = 615 n = 580 n = 617
Figure 1. Flowchart of numbers and exclusions of study population and distribution over the seven dimensions
Respondents characteristics
Table 1 shows characteristics of the participating practices sorted by profession. The largest
groups to respond were physiotherapists (n=150), midwives(n=125) and anticoagulation
clinics (n=99). The smallest numbers of respondents were for dietetics (n=19) and skin
therapy (n=26). The high percentage of female respondents stands out (82,6%), only in
physiotherapy practices the percentage of male and female employees were equal. With
regard to practice size, skin therapy, exercise therapy and speech therapy practices were
small, whilst anticoagulation clinics were large. Working experience was shortest in skin
therapy, anticoagulation clinics and midwifery practices and longest in exercise therapy.

The mean dimension scores, SD and PSG are presented in Table 2. At the bottom of Table
CHAPTER 3
2 the grand mean of each dimension is presented. In general, primary care professions
perceived dimensions positively. There were two dimensions that scored below four: (6)
‘intention to report events’ scored the lowest (3.73) and (4) ‘patient safety management’
(3.79). The highest dimension scores were for (1)‘open communication and learning from
error’ (4.25) and (5)‘support and fellowship’ (4.26). Dimension (6) ‘Intention to report
events’ showed the largest variation within the profession groups itself. In addition, the
PSG was rated positively (four or higher) with a mean of 4.03 ( range 3.62 – 4.16). Two
professions, occupational therapy (3.62) and anticoagulation therapy (3.83), scored below
four on the PSG.
Differences between professions

When comparing each profession to the grand mean, in general, deviations were small.
Differences larger than half a standard deviation are underlined in Table 3. Two professions
showed only negative deviations from the overall mean: occupational therapy and dietetics.
These professions also had the largest deviations overall. In addition, anticoagulation
therapy also perceived safety culture more negatively on most dimensions when compared
to the other professions. In contrast, physiotherapy was the only profession that showed
solely positive deviations on all dimensions. However, none of these differences were
larger than half a SD. Dental care deviated slightly positive on all dimensions but (1) ‘open
communication’, where they deviated negatively. With regard to the dimensions, largest
deviations were found for (6) ‘intention to report events’ which showed two large negative
deviations, 0.84 (occupational therapy) and 0.63 (dietetics) respectively.

CHAPTER 3
50
Table 1 Characteristics of respondents and practices

CHAPTER 3
Dietetics
Physiotherapy
Dental care
Skin therapy
Exercise
therapy
Occupational
therapy
Anticoagula-
tion therapy
Midwifery
Speech
therapy
Respondents (n) 19 150 61 26 36 39 99 125 70
Practices (n) 13 52 46 22 27 28 14 70 41
Professionals per practice
2-4 5 13 15 18 21 14 0 34 35
5-9 2 18 14 3 4 10 0 32 6
10-14 1 13 5 1 1 0 4 4 0
≥15 5 5 12 0 1 4 10 0 0
Age of respondents* 43 37.5 42 29 44 43 50 35 37
median (range) (24-56) (22-64) (24-63) (25-63) (25-58) (27-56) (24-63) (22-61) (22-61)
Gender (% women)* 100 50.3 75.9 100 97.1 93.8 87.8 96.8 100
Working experience in years 17 11.5 13 5 20 16 7 8 13
median (range)
(1.5-30) (0-40) (0.5-40) (2.5-25) (3-33) (1-35) (0.5-36) (0.5-40) (1-38)
* The distribution of age and gender is representative for Dutch primary care professionals.
Table 2 Mean scores per dimension and PSG, presented by profession
learning from error (mean, SD)
working conditions (mean, SD)
4 Patient safety management

3 Adequate procedures and
1 Open communication and
6 Intention to report events

2 Handover and teamwork
7 Organisational learning
5 Support and fellowship
Patient safety grade

(mean, SD)
(mean, SD)
(mean, SD)
(mean, SD)
(mean, SD)
(mean, SD)
Number of re- 611 446 523 302 615 580 617 605
CHAPTER 3
spondents
Dietetics 4.00 3.78 4.06 3.86 4.23 3.24 3.98 4.00
(0.67) (0.28) (0.39) (0.28) (0.37) (1.22) (0.45) (0.77)

Physiotherapy 4.20 4.11 4.22 3.93 4.29 3.72 4.07 4.12
(0.56) (0.49) (0.49) (0.59) (0.49) (0.81) (0.52) (0.64)

Dental care 4.13 4.03 4.20 4.05 4.34 3.92 4.03 4.16
(0.66) (0.56) (0.42) (0.61) (0.42) (1.03) (0.50) (0.73)

Skin therapy 4.28 3.86 4.20 3.95 4.50 3.67 3.92 4.12
(0.58) (0.23) (0.26) (0.62) (0.35) (0.86) (0.57) (0.53)

Exercise therapy 4.38 3.98 4.40 4.27 4.13 3.55 3.67 4.14
(0.33) (0.31) (0.57) (0.31) (0.23) (0.51) (0.33) (0.49)

Occupational 3.57 3.74 3.90 3.55 3.96 2.84 3.55 3.62
therapy (0.51) (0.33) (0.37) (0.71) (0.40) (1.02) (0.35) (0.74)

Anticoagulation 3.90 3.38 3.68 3.69 3.86 3.76 3.99 3.83
therapy (0.54) (0.72) (0.63) (0.54) (0.56) (0.84) (0.43) (0.56)

Midwifery 4.44 4.09 4.01 4.02 4.07 4.02 3.82 4.08
(0.43) (0.55) (0.65) (0.77) (0.93) (0.93) (0.97) (0.52)

Speech therapy 4.35 4.08 4.23 4.00 4.39 3.79 3.88 4.05
(0.64) (0.23) (0.24) (0.23) (0.27) (0.75) (0.38) (0.59)

Grand mean (SD) 4.25 3.99 4.14 3.79 4.26 3.73 3.98 4.03
(0.59) (0.62) (0.54) (0.65) (0.56) (1.01) (0.58) (0.62)

Table 3: Deviations from the grand mean presented by profession
7 Organizational learning
5 Support and fellowship
4 Patient safety manage-
and learning from error
3 Adequate procedures
1 Open communication
and working conditions

2 Handover and team-
6 Intention to report
events
ment
work
Dietetics (n=19) -0.35* -0.21 -0.05 -0.19 -0.03 -0.63* -0.34*
Physiotherapy (n=150) 0.08 0.22** 0.16* 0.13 0.12* 0.08 0.11
Dental care (n=61) -0.26** 0.04 0.08 0.05 -0.03 0.19 0.11
CHAPTER 3
Skin therapy (n=26) 0.17 -0.05 0.25* 0.17 0.31** -0.02 0.10
Exercise therapy (n=36) 0.17 0.10 0.28* 0.52 0.30** 0.13 -0.01
Occupational therapy (n=39) -0.43** -0.18 -0.16 -0.39* -0.13 -0.84** -0.37**
Anticoagulation therapy -0.25* -0.57** -0.35** -0.23 -0.27** 0.10 0.08
(n=99)
Midwifery (n=125) 0.22** 0.13 -0.03 0.14 0.03 0.06 0.00
Speech therapy (n=70) -0.05 0.18 0.17* 0.07 0.04 -0.22 -0.20*
Multilevel analyses of professions in relation to the grand mean of the SCOPE-PC dimensions, adjusted for
clustering in practices. Differences larger than half a standard deviation are underlined.
* p<0.05
** p<0.01
DISCUSSION
In exploring perceptions of patient safety culture in nine Dutch primary care professions,
we found that all professions perceived safety culture fairly positive and graded patient
safety in their practice as very well. Differences in perception of patient safety between the
professional groups were small.
Comparison with literature

Real comparison with other studies on perceived safety culture in primary care was difficult
because of heterogeneity both in the applied questionnaires and the reporting of outcomes.
In addition, almost all were focused on family practice, only one study reported separately
52 CHAPTER 3
on the results of midwifery.9 Compared to these results it seemed that the midwives in
our study perceived safety culture more positively. In a previous study conducted in family
practices in the Netherlands, using a slightly different version of the SCOPE-PC questionnaire,
means of the eight dimensions ranged between 3.8 and 4.1.20 This corresponds with our
results and indicates that primary care professionals, although not exposed as much as
general practitioners to the concept of patient safety yet, still experience patient safety
quite similarly. Other studies also found generally positive results in family practice.7, 8 In our
study the intention to report was perceived the least positive of all dimensions. This is in
line with other studies that found similarly low scores on the frequency of events reported9
and error management.7
CHAPTER 3
The strength of this study is that we have gained insight in patient safety culture in
several primary care professions that have not been examined before. Also, we used one
generic questionnaire to assess perceptions of patient safety culture in all professions.
Hereby, we could not only describe the current state of affairs but also make comparisons.
A limitation was the low response of 24%. Also, the responses varied across the professions.
Various causes may have contributed to the low response. Firstly, not all national
professional associations were able to provide up to date addresses of professionals nor
could they specify whether practices were single handed. This led to loss of participants.
Secondly, there was a selective non-response as the low response mainly occurred in
particular professions: dietetics, skin therapy, exercise therapy and occupational therapy.
These professions may perceive their practice as less likely to provoke harm and therefore
were less inclined to participate in our study. Thirdly, because we initially contacted one
professional, and in turn asked them to involve their colleagues, the degree of interest and
position of this contact person might have determined participation of the whole practice.
In line with this, some professions mainly consist of single handed or small practices. By
asking all staff to complete the questionnaire, in some professions this inherently will lead
to less questionnaires because practices are smaller. In hindsight, taking into account the
populations of different professions could have contributed to a larger response rate and
more individual questionnaires.
The low response rate constrains the generalizability of our results. In addition, we cannot
exclude that the responders were somewhat more positive about patient safety. However,
there is no reason to believe that selection was different for the different professions and
therefore did not affect the comparisons between them.
Also, there was a considerable amount of questionnaires that had missing values, which
might limit the strength of the study. However, we checked whether this might have caused

differences in outcome by imputing the mean dimension scores when <50% of dimension
items were missing. Original and imputed mean dimension scores and results of the
multilevel analyses gave only minor differences and none of relevance to the conclusions.
We chose to compare profession scores with a grand mean per dimension as comparing
nine professions with each other led to uninterpretable results because of the amount of
comparisons.
Interpretation of the results

Notable was the dimension ‘intention to report’ as it was perceived both the least positive
but also varied the most within the profession groups itself. This finding reflects the early
stage of development of incident reporting as part of safety management in primary care.
Nationally and within professions incident reporting procedures and forms have been
CHAPTER 3
developed. Whilst these tools are easily available, actual incident reporting has not landed
in daily practice yet. We see incident reporting as an important and logic tool in dealing with
the potential occurrence of incidents, however, not all professionals may share this view.
Implementation and adherence is dependent on personal motives and context.21 Therefore,
it could be that on the one hand practices are just gradually starting to implement patient
safety initiatives. In this process some practices are forerunners and some lag behind
which could explain the variation and possibly, in a few years incident reporting will be
more common in daily primary care practice. On the other hand, we should also consider
the possibility that incident reporting is of less relevance in some professions because of
the nature of the work. For example, occupational therapy which scored lowest, rarely
experiences an incident and, those that do occur are mostly without significant harm.
From the literature it is known that leadership is an important precondition for sustainable
patient safety implementation.22, 23 The relatively lower scores on ‘patient safety management’
might reflect the structure of the organisations in primary care that are generally small and
mostly do not have a clear hierarchy. Often there is no formal supervisor in these practices,
but professionals sharing the same responsibilities. Another plausible argument could be
that patient safety is relatively new and therefore is not managed explicitly.
Promising is the finding that ‘open communication and learning from error’ was perceived
positively by all professions, because it is an important condition for patient safety culture.
Three professions showed negative deviations on this dimension, i.e. occupational
therapists, dieticians and dentists. While the first two professions perceived their culture
more negatively overall, we considered the result of dental care on this dimension as more
striking. Whereas overall dental staff perceived most patient safety culture aspects as
54 CHAPTER 3
positive as the other professions, in open communication they seem to deviate negatively
from their primary care peers. This may indicate that open communication is a sensitive
subject in dental care that requires specific attention when targeting patient safety issues
in this profession.
Implications and conclusions

This study showed that primary care professionals in the Netherlands are rather
positive in their opinion of their patient safety culture. In addition, our study indicated
differences between professions which may demand a tailored approach of patient
safety management per profession. However, given the low response rate we cannot
draw firm conclusions nor specifically inform on directions of improvement strategies.
Primary care is a key factor in the whole healthcare system. Not only a large proportion
CHAPTER 3
of health care is provided in this setting, it also contributes to healthcare outcomes.24, 25
Furthermore, primary care professionals increasingly collaborate facing the challenge
of the aging population and more complex care. In this view it is of importance to
gain more insight in patient safety and improvement strategies in these settings.
Having said this, our study was an important first step in examining perceptions of different
professions. Hopefully this leads to more attention and research in this area of healthcare. We
believe it is necessary to conduct further research, desirably with mixed methods to further
explore attitudes towards patient safety and identify specific needs for improvements.

REFERENCES
2003;38(3):923-45.
Synthesis of the Measurement Tools. Journal of patient safety 2006;2:105-15.
medical record review study. Implement Science 2011;6:37.
CHAPTER 3
6. Singh G, Singh R, Thomas EJ, Fish R, Kee R, McLean-Plunkett E, Wisniewski A, Okazaki S, Anderson
D. Measuring Safety Climate in Primary Care Offices. In: Henriksen K, Battles JB, Keyes MA, Grady
ML, editors. Advances in Patient Safety: New directions and alternative approaches (Vol 2: Culture
and Redesign) Rockville: Agency for Healthcare Research and Quality; 2008. p. 59-72.
7. Hoffmann B, Miessner C, Albay Z, Schröber J, Weppler K, Gerlach FM, Güthlin C. Impact of
Individual and Team Features of Patient Safety Climate: A survey in Family Practices. Annals of
family practice 2013;11(4):355-62.
8. De Wet C, Johnson P, Mash R, McConnachie A, Bowie P. Measuring perceptions of safety climate
in primary care: A cross-sectional study. Journal of evaluation in clinical practice 2012;18:135-42.
9. Bodur S, Filiz E. A survey on patient safety culture in primary healthcare services in Turkey.
International journal for quality in health care 2009;21(5):348-55.
10. Nordén-Hägg A, Sexton JB, Kälvemark-Sporrong S, Ring L, Kettis-Lindblad Å. Assessing Safety
Culture in Pharmacies: The psychometric validation of the Safety Attitudes Questionnaire (SAQ) in
a national sample of community pharmacies in Sweden. BMC clinical pharmacology 2010;10(8).
11. Phipps D, De Bie J, Herborg H, Guerreiro M, Eickhoff C, Fernandez-Llimos F, Bouvy M, Rossing
C, Mueller U, Ashcroft D. Evaluation of the Pharmacy Safety Climate Questionnaire in European
community pharmacies. International journal for quality in health care 2012;24(1):16-22.
13. Ashcroft D, Morecroft C, Parker D, Noyce P. Safety culture assessment in community pharmacy:
development, face validity, and feasibility of the Manchester Patient Safety Assessment
Framework. Quality and safety in health care 2005;14:417-21.
14. Sorra J, Dyer N. Multilevel psychometric properties of the AHRQ hospital survey on patient safety
culture. BMC health services research 2010;10:199.
15. Smits M, Wagner C, Spreeuwenberg P, van der Wal G, Groenewegen PP. Measuring patient safety
culture: an assessment of the clustering of responses at unit level and hospital level. Quality and
safety in health care 2009;18:292-6.
56 CHAPTER 3
practice 2011;12:117.
17. Verbakel NJ, Zwart DLM, Langelaan M, Verheij TJM, Wagner C. Measuring safety culture in Dutch
primary care: psychometric characteristics of the SCOPE-PC questionnaire. BMC health services
research 2013;13:354.
2011.
19. Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life: the
remarkable universality of half a standard deviation. Medical care 2003;41(5):582-92.
20. Zwart DLM, Langelaan M, Kuyvenhoven MM, Rensen vELJ, Kalkman CJ, Verheij TJM. Exploration
of patient safety culture perceptions in Dutch general practice. A cross sectional survey. Incident
reporting in general practice [dissertation] Utrecht: University Medical Center Utrecht; 2011. p.
CHAPTER 3
47-61.
21. Bont de A, Bal R. Telemedicine in interdisciplinary work practices: On an IT system that met the
criteria for success set out by its sponsors, yet failed to become part of every-day clinical routines.
BMC medical informatics and decision making 2008;8:47.
2006.
23. Kaplan HC, Provost LP, Froehle CM, Margolis PA. The Model for Understanding Success in Quality
(MUSIQ): building a theory of context in healthcare quality improvement. BMJ Quality and Safety
2012;21:13-20.
24. Starfield B, Shi L, Macinko J. Contribution of Primary Care to Health Systems and Health. The The
Milbank quarterlyuarterly 2005;83(3):457-502.
25. Kringos DS, Boerma W, van der Zee J, Groenewegen P. Europe’s Strong Primary Care Systems
Are Linked To Better Population Health But Also To Higher Health Spending. Health affairs
2013;32(4):686-94.

PART II
Patient safety culture interventions in general

practice
CHAPTER 4
Improving patient safety culture in primary care:

a systematic review
Published as:
Verbakel NJ, Langelaan M, Verheij TJM, Wagner C, Zwart DLM.

Improving patient safety culture in primary care: a systematic review.
Journal of Patient Safety, published online.
ABSTRACT
Background
Patient safety culture, described as shared values, attitudes and behaviour of staff in a heal-
thcare organization, gained attention as a subject of study as it is believed to be related
to the impact of patient safety improvements. However, in primary care it is yet unknown
which effect interventions have on the safety culture.
Objectives
To review literature on the use of interventions that effect patient safety culture in primary
care.
Methods
Searches were performed in PubMed, EMBASE, CINAHL and PsychINFO on the 4th of March
2013. Terms defining safety culture were combined with terms identifying intervention and
terms indicating primary care. Inclusion followed if the intervention affected patient safety
culture and effect measures were reported.
Results
CHAPTER 4
The search yielded 214 articles from which two were eligible for inclusion. Both studies
were heterogeneous in their interventions and outcome, we present a qualitative summary.
One study described the implementation of an electronic medical record system in general
practices as part of patient safety improvements. The other study facilitated two workshops
for general practices, one on risk management and another on significant event audit. Re-
sults showed signs of improvement but the level of evidence was low due to the design and
methodological problems.
Conclusion
These studies in general practice provide a first understanding of improvement strategies
and their effect in primary care. As the level of evidence was low, no clear preference can
be determined. Further research is needed to help practices make an informed choice for
an intervention.
62 CHAPTER 4
INTRODUCTION
Patient safety has become a major topic in healthcare research and recently its scope has
been extended to primary care, as the role of primary care in healthcare increases in size
and spending.1,2 In the Netherlands, primary care covers a large part of healthcare: more
than 90% of healthcare is delivered in primary care against only four percent of the total
healthcare costs.3 Although incidents in primary care tend to be less harmful compared to
incidents that occur in hospitals, the impact on overall safety in healthcare is at least similar
due to the large number of patient contacts in primary care.2,4
In primary care, patient safety research initially focused on studying taxonomy5 and reporting
systems.6,7 These studies found that the majority of incidents can be categorised as process
incidents including administrative failures. Other important categories were communication,
knowledge and skills. Gaal et al.8 found patient safety incidents in 2.5 percent of patient
contacts by reviewing general practice medical records and communication was one of the
reported causes. Another study in 48 primary care centres in Spain, identified 773 adverse
events and stated that problems with communication and management were at the root
CHAPTER 4
of many of these events.9 Reviewing 75 error reports, Woolf et al. found that 77 percent
of the incidents were caused by a cascade of errors.10 This shows that collaboration and
communication are relevant issues to patient safety in primary care.
The way colleagues interact and collaborate in an organisation is part of their culture.
Safety culture is described as the shared values, attitudes and behaviour of all staff in
health facilities in regard to giving safety priority over efficiency, improving care provider
communication and collaboration, and creating a system that learns about and learns from
errors and problems.11 Furthermore, it is known that a safe and open culture is important
for patient safety improvement.12
Studies on patient safety culture were mostly conducted in hospitals. A systematic review
of patient safety improvement strategies indicated leadership walk rounds and multi-
faced unit-based programmes as having a positive impact on patient safety culture in
hospitals.13 Another review indicated multiple component strategies including team
training, communication and executive engagement in walk rounds to have the best
evidence.14 However, in both reviews the level of evidence moderates firm conclusions on
the effectiveness of patient safety culture in healthcare.
For primary care, however, it is not clear what the effect of patient safety interventions is on
patient safety culture. It is not self-evident that patient safety culture strategies conducted
in hospital care can be similarly applied in primary care, or that they will have similar effects.
The organisational structure differs as primary care practices have a smaller scale and are
generally less hierarchical than hospitals. In addition, hospitals mainly provide therapeutic
Improving patient safety culture in primary care: a systematic review 63

care whereas primary care practices also care for preventive and diagnostic care, which may
lead to different safety awareness and behaviour. Therefore, we conducted a systematic
review to assess the effectiveness of patient safety interventions on patient safety culture
in primary care.
METHOD
Search
A literature search of papers describing an intervention with patient safety culture
measurements in primary care was conducted in four databases: CINAHL, Embase, PubMed
and PsycINFO. We combined terms defining safety culture such as ‘organisational culture’,
‘safety management’, ‘patient safety’ with both terms identifying intervention, for example
‘improvement’, ‘change’, ‘effect’ and terms that indicated the setting of primary care. The
PubMed search strategy is enclosed in Appendix 1. No restrictions were set regarding
publication date. Language was restricted to English, Dutch and German. In addition, we
screened the webpages of the Institute for Healthcare Improvement, the National Patient
CHAPTER 4
Safety Agency and the Agency for Healthcare Research and Quality.15-17 Also the references
of included articles were checked for relevant literature. The search strategy was conducted
on the 4th of March 2013.
Eligibility and quality assessment

Studies were eligible for inclusion if they met three inclusion criteria. First, the research
had to be conducted in primary care. Second, a patient safety intervention affecting culture
had to be described. And third, the effect on patient safety culture had to be reported.
There were no requirements regarding the design of the study, provided that there had
to be more than one measurement. Opinion papers, editorials, reviews, interviews and
comments were excluded. For assessment of the study quality we examined the quality of
reporting, using five criteria for qualitative research used in previous intervention research,
and potential bias by using ‘the risk of bias tool’ from the Cochrane Collaboration, see Table
3.18,19 In addition, the GRADE approach from the Cochrane Collaboration was used to assess
the level of evidence ranging from high to very low.20
Selection process and data extraction

The title screening was conducted by one author (NV) followed by abstract and full-text
screening by two authors (NV; ML). Results were compared and discussed between both
authors. In case of disagreement a third author (DZ) was consulted. Data was extracted on
intervention characteristics, defined as aims, measurement tools, intervention description
64 CHAPTER 4
and effect on culture. We also extracted data on study characteristics, defined as country,
design and participants. For the extraction of data a beforehand composed form was used.
RESULTS
In total, 214 references were retrieved from the database search (Figure 1). After initial
screening, eighteen articles were selected for full text screening. No references were added
after searching the bibliographies of included studies. A list of excluded articles is enclosed
in Appendix 2. Two studies21,22 met our inclusion criteria, as these were both observational
and reported on different interventions with heterogeneous outcomes no meta-analysis
was done.
References identified by screening

CINAHL EMBASE PubMed PsychINFO
webpages and reference checking
(n=42) (n=81) (n=156) (n=11)
(n=0)
CHAPTER 4
After deduplication
(n=214)
Title screening
(n=214)
196 records removed
Abstract screening
(n=18) 11 records removed
Studies did not include an
intervention, setting was not
primary care or study was
Full-text screening
descriptive.
(n=7)
5 records removed
Studies did not report on
2 studies included culture effects or did not
include an intervention, one
protocol.
Figure 1: Flowchart search results

Both studies were conducted in general practice. Table 1 shows their study characteristics.
The study of McGuire et al. used a follow-up design with a total follow-up time of three
years.21 Wallace et al. used a pre-post design with an implementation time of eight months.22
Both studies did not use a control group.
Table 1 Study characteristics

Study Country Design Participants; response rate
McGuire et USA Follow-up 18 practices participated#, response rate per
al. 201321 measurement point:
Control group: no T1: 83.7% (103 of 123)
T2: 85.3% (122 of 143 )
Implementation T3: 78.5% (142 of 181)
time: 3 years
Wallace et al. UK Pre-post Risk management data, response rate per
200722 measurement point:
Control group: no Pre: 57% (43 from 75 practices)
Post: 33% (24 from 73 practices)*
CHAPTER 4
Implementation
time: 8 months Learning organization Culture Questionnaire:
Pre: 45 practices. Respondents: 41% (184/450)
Post: 36 practices. Respondents: 56%(125/225)
#
Respondents increased from 103 to 142 due to growing of the provider group
*
Two practices ceased to exist or amalgamated by T2
Table 2 shows details on the intervention and effect measurements. McGuire et al. described
the implementation of an electronic medical record (EMR) system.21 This was part of on-going
quality and safety improvement efforts. Additional efforts were made to facilitate the EMR
implementation such as identification of ‘change champions’, development of committees
to support implementation, reduction of work schedules during the first two weeks and on
site “super-user” support. Immediately prior to go-live, staff attended a training session.
The effect of the intervention was assessed with the Safety Attitudes Questionnaire (SAQ)23
directly after implementation and repeated after 1.5 and 2.5 years. Also, practices were asked
to indicate the most important safety issues specific for their practice. Five of seven domains
of the SAQ, ‘job satisfaction’, ‘perceptions of executive management’, ‘local management’,
‘safety climate’ and ‘teamwork climate’, showed significant improvements between T1 and
T3. ‘Working condition’ significantly improved between T2 and T3. Respondents reported
time constraints as the most significant concern, followed by communication problems.
66 CHAPTER 4
Table 2 Intervention characteristics
Study Aim Measurement tools Intervention description Effect on culture
McGuire Improving safety and evaluating changes - Safety Attitudes Implementation of an Changes in percentages for SAQ dimensions at T1,T2
et al. in perceptions of safety among the Questionnaire electronic medical record and T3:
Job satisfaction: 74.1 78.2 86.2
201221 primary care provider group after EMR - Practice-specific system.
Perceptions of Executive 59.1 66.7 72.6
implementation. needs assessment
Management:
Perceptions Of Local 76.2 84.6 86.0
Management:
Safety Climate: 76.4 84.2 87.8
Stress Recognition: 68.4 75.6 74.8
Teamwork Climate: 77.4 85.5 88.9
Working Conditions: 74.3 74.2 84.9
Agreement to question: “Our electronic medical

record has improved our ability to provide safe
patient care”: T2: 77.9% / T3: 85.4%
Wallace, To establish that practices were prepared - RM audit - Medical Defence Union RM competence score showed an overall significant
L.M. et al. to engage in risk management (RM) questionnaire RM workshops (single improvement at practice level.
200722 through: day)
- having the right skills, - Learning organization - Facilitation of significant At T1 there was no association found between the
- being supported by structures and Culture Questionnaire event analysis (SEA) (2 levels of competence and culture.
policies, (LCQ) with 4 domains: hours)
- having staff who believed their practice - creativity - Own development At T2 ‘task information’ was significant (p<0.01) in
has an open learning culture. - communication activities including a positive direction and ‘practice development’ was
- climate Quality Team significant (p<0.009) in a negative direction.
Secondary objective: - change Development.
Evaluation of the contribution of the RM
Improving patient safety culture in primary care: a systematic review

initiatives to the development of RM
competence and learning culture.
67
CHAPTER 4
A majority responded positively when asked whether implementation of the EMR enhanced
their ability to provide safe care to patients. Wallace et al. studied the effect of own patient
safety initiatives and two workshops, a Risk Management (RM) workshop and a significant
event audit (SEA)24 workshop, respectively.22 The RM workshop included a practice self-
assessment questionnaire and feedback against other training sites, use of protocols for
patient group directions and chaperones, and a lecture on how to conduct a SEA.
The second workshop consisted of the lecture on SEA solely. Effects were measured by a
RM competence score covering ‘the scope of Risk Management activity’, ‘staff involvement’,
‘documentation of RM activities’, ‘accessibility of RM records’, ‘existence of specific written
policies and an audit program’. In addition, the Learning organization Culture Questionnaire
(LCQ) was completed by practice staff. This survey measured eight dimensions: ‘personal
innovation’, ‘open communication’, ‘personal blame for errors’, ‘error awareness’, ‘team
problem-solving’, ‘task information’, ‘supportive climate’, and ‘practice development’,
distributed over four domains.
Seventy-five practices were invited to participate either in the workshop of preference
CHAPTER 4
(RM: n=40, SEA: n=2) or in both (n=9). There were twenty-four practices that chose
not to participate in one of the workshops as they undertook their own development
activities. Practices that responded at T1 and T2 were included in the analysis (n=20). The
authors reported an overall significant improvement of the RM competence score. Three
competences improved: there was a widening of the scope of RM activities, more staff
were involved and activities were increasingly documented in formal systems. Spent time
was indicated as main disadvantage of RM. The gains reported by most of the practices
was ‘better learning from events’, ‘fewer complaints’ and ‘a better atmosphere’. Results
from the LCQ were used to examine the association with the RM competence. At baseline
three subscales showed a positive relation with RM competence scores. At follow-up ‘task
information’ was positive and ‘practice development’ was negative correlated.
Quality appraisal
Both studies used an observational design without a control group. Following the GRADE
approach they are therefore graded as ‘low’ on the level of evidence rating. Though, an
observational design could be upgraded to a ‘moderate’ level of evidence when the
study is methodologically sound and yields large, consistent and precise estimates of the
intervention effect.20
We appraised the publications on methodological quality according to quality of reporting
and potential risk of bias (Table 3). As these studies did not use a control group, performance
and detection bias were not applicable.
68 CHAPTER 4
Table 3 Quality appraisal
Study Quality appraisal
McGuire et Quality of reporting:
al. 201321 Aims clearly reported +
Adequate description of context +
Adequate description of sample and methods of recruiting +
Adequate description of data collection +
Adequate description of data analysis +
Potential risk of bias:
Selection -
Attrition -
Reporting -
Other There was no baseline measurement.
No adjustment for possible within-person correlations (80% of respondents were similar in T1 and T3)
Other interventions (communication training, management processes and educational interventions) were simultaneously present.
Wallace et Quality of reporting:
al. 200722 Aims clearly reported +
Adequate description of context +
Adequate description of sample and methods of recruiting +/- Unclear which practices and corresponding demographics are included in the analysis.
Adequate description of data collection +
Adequate description of data analysis +/- Unclear what the significance and value of differences of the competency scales are.
Potential risk of bias:
Selection Possible bias due to asking practices to volunteer. Practice that declined the workshops were already undertaking their own development
activities.
Attrition There was selective drop-out of practices (from 43 to 24 for RM data) who chose not to participate anymore due to own initiatives or
priorities. Response rate was very low and for follow-up these were halved for the RM data. It was reported that a check for sample bias was
done.
Results on the scales of the LCQ are not reported and it was not reported which domains of the LCQ correlate with RM at baseline.
Improving patient safety culture in primary care: a systematic review

Reporting RM competence score at T2 was derived from audit of only the practice manager.
Only practices that delivered data at both T1 and T2 were included in analysis.
Other There was no distinction made between the three initiatives in the analysis and reporting.
69
CHAPTER 4
Quality of reporting was good in McGuire et al. Potential bias was possible as there was
no adjustment for possible within-person correlations and because of simultaneous
implementation of other interventions. The study of Wallace et al. had some limitations
regarding the reporting of the sample and the significance and value of the assessed risk
management competency scales. Bias could occur due to selection, attrition and reporting,
as there were half as many practices at T2 than at T1 and measurements of the LCQ were
not reported. Furthermore, the reporting of results was limited. In addition, only the twenty
practices from which data was available for both T1 and T2 were included. Also, in the
analyses all three initiatives, the RM workshop, the SEA workshop and own activities were
analysed together: no results were given for separate groups.
DISCUSSION
In our search for primary care studies that implemented patient safety strategies which
affect patient safety culture we found two studies, both conducted in general practice.
McGuire et al. implemented an Electronic Medical Record and measured improvement
CHAPTER 4
on safety climate and teamwork climate with the SAQ.21 Wallace et al. assessed the effect
of organizational initiatives; participation in a workshops on Risk Management (RM) or
Significant Event Analysis (SEA) or own activities.22 It showed increased risk management
activities on clinical or administrative issues. A learning culture seemed positive for the risk
competence score, although the size and content of this relation remained unclear. Overall,
both publications approached both their interventions as well as the evaluation of effect
differently. Whereas the study of McGuire et al. applied a culture questionnaire, Wallace
et al. more directly assessed patient safety behaviour and its relation to a learning culture,
in which aspects of a safe culture are incorporated. These varied approaches align with
observations that patient safety culture is a very versatile concept.13,25

This review revealed only two studies. This may be due to the strict application of the
inclusion criteria because we searched for intervention studies that both assessed and
reported on the cultural effect. Although we additionally searched websites of the IHI, NPSA
and AHRQ it may be that we missed publications in grey literature outside the mainstream.
In addition, it could be that studies have been conducted on the improvement of patient
safety culture in primary care but were not published.
We cannot draw any firm conclusions as the level of evidence of both studies was low.
This is largely due to the observational design but also because of the likelihood of bias.
70 CHAPTER 4
On the other hand, it is very difficult to rule out all influences as in a pragmatic study the
research environment cannot be standardised. Such complex interventions are inherently
conducted in existing systems and therefore raises the question of attribution of the effect
to the intervention.26 However, the strength of such observational studies is that they are
less intrusive in the usual course of affairs which is beneficial to the validity of the study
results.27
The validity may be enhanced by combining with a qualitative study, so called triangulation
in a mixed method.28 This could, for example, shed light on what respondents themselves
designate as most effective aspects for their organisation and why they perceive these as
such. To some extent Wallace et al. have done this by describing the disadvantages and
advantages of the intervention that were reported by practice managers.
Comparison with existing literature

Up to now, no review on patient safety interventions affecting culture was conducted with
a specific focus on primary care. In hospital setting, however, reviews showed leadership
walk rounds, multi-faceted unit-based programmes, teamwork and communication to
CHAPTER 4
have a positive effect on patient safety culture.13,14 Due to organisational differences and
size of the practices it is not clear whether these strategies are applicable or have a similar
effect in primary care. The study by McGuire et al. did use a broad approach by embedding
the intervention in facilitating activities such as training and installing committees. The
intervention was not a stand-alone: it was accompanied by communication, educational
and managerial interventions. Leadership walk rounds will be more difficult to apply in
primary care, as the small primary care practices often lack a clear hierarchical organisational
structure. However, audits or peer reviews by colleagues from other practices may
have similar beneficial effects. A tool to assess and discuss patient safety culture is the
Manchester Patient Safety Framework (MaPSaF).29 This framework was modified for use in
New Zealand’s general practice. During this qualitative study MaPSaF was observed to be a
helpful discussion tool which stimulated learning and enhanced communication.30
Implications for research and practice

In conclusion, this review indicates effect of interventions on patient safety culture in primary
care, but the size and external validity of the measured changes are unclear. Furthermore,
the range of interventions on patient safety culture in primary care is limited compared
to secondary care. In conclusion, it is not evident which intervention would help practices
most to improve their patient safety culture. Hence, practices should choose an intervention
close to their momentary needs of improvement and evaluate frequently to assess whether
the intervention leads to the desired effect.31

To support this choice, more research is needed to assess the effect of interventions on
safety culture in primary care practice. Studying facilitators and barriers aimed at clarifying
the mechanisms that underlie the dynamics of a patient safety culture interventions, would
add to patient safety improvement in primary care.
CHAPTER 4
72 CHAPTER 4
REFERENCES
1. Kringos DS. The strength of primary care in Europe [dissertation]. Utrecht: NIVEL; 2012.
2. Gandhi TK, Lee HL. Patient Safety beyond the Hospital. The New England Journal of Medicine
2010;363(11):1001-3.
3. Wiegers T, Hopman P, Kringos D, de Bakker D. Overzichtstudies. De eerste lijn. Utrecht: NIVEL;
2011.
4. Zwart DLM. Patient safety incidents in general practice: important needles in many haystacks.
Nederlands Tijdschrift voor Geneeskunde 2011;155:A4021.
5. Dovey SM, Meyers DS, Phillips RL, Green LA, Fryer GE, Galliher JM, Kappus J, Grob P. A preliminary
taxonomy of medical errors in family practice. Quality and safety in health care 2002;11(3):233-8.
6. Makeham M, Dovey S, Runiciman W, Larizgoitia I. Methods and measures used in primary care
patient safety research. Results of a literature review. World Health Organization; 2008.
2011;20(2):121-7.
medical record review study. Implementation Science 2011;6:37.
CHAPTER 4
9. Aranaz-Andres JM, Aibar C, Limon R, Mira JJ, Vitaller J, Agra Y, Terol E. A study of the prevalence of
adverse events in primary healthcare in Spain. European journal of public health 2012;22(6):921-
5.
10. Woolf S, Kuzel A, Dovey S, Phillips R. A string of mistakes: the importance of cascade analysis in
describing, counting, and preventing medical errors. Annals of family medicine 2004;2(4):317-26.
11. Blegen MA, Sehgal NL, Alldredge BK, Gearhart S, Auerbach AA, Wachter RM. Improving
safety culture on adult medical units through multidisciplinary teamwork and communication
interventions: the TOPS Project. Quality and safety in health care 2010;19(4):346-50.
2006.
15. Institute for Healthcare Improvement. Available at: http://www.ihi.org/knowledge/Pages/
Publications/default.aspx.
16. National Patient Safety Agency. Available at: http://www.npsa.nhs.uk/.
17. Agency for Healthcare Research and Quality. Available at: http://www.psnet.ahrq.gov/
collectionBrowse.aspx?taxonomyID=314.
18. Harden A, Brunton G, Fletcher A, Oakley A. Teenage pregnancy and social disadvantage: systematic
review integrating controlled trials and qualitative studies. BMJ 2009;:339:b4254.
19. Higgins J, Altman D, Sterne J. Chapter 8: Assessing risk of bias in included studies. In: Higgins
J, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0

ed.The Cochrane Collaboration; 2011.
20. Schünemann H, Oxman A, Vist G, Higgins J, Deeks J, Glasziou P, Guyatt G. Chapter 12: Interpreting
results and drawing conclusions. In: Higgins J, Green S, editors. Cochrane Handbook for Systematic
Reviews of Interventions. Version 5.1.0 ed.The Cochrane Collaboration; 2011.
21. McGuire MJ, Noronha G, Samal L, Yeh HC, Crocetti S, Kravet S. Patient safety perceptions of
primary care providers after implementation of an electronic medical record system. Journal of
general internal medicine 2013;28(2):184-92.
22. Wallace LM, Boxall M, Spurgeon P, Barwell F. Organizational interventions to promote risk
management in primary care: the experience in Warwickshire, England. Health services
management research 2007; 05;20(2):84-93.
23. Sexton J, Helmreich R, Neilands T, Rowan K, Vella K, Boyden J, Roberts P, Thomas E. The Safety
Attitudes Questionnaire: psychometric properties, benchmarking data, and emerging research.
BMC health services research 2006;6(44).
24. Pringle M. Significant event auditing. Scand J Prim Health Care 2000;18(4):200-2.
25. Guldenmund FW. The nature of safety culture: a review of theory and research. Safety Science
2000;34(1-3):215-57.
27. Grol R, Wensing M. Implementatie. Effectieve verbetering van de patiëntenzorg. [Implementation.
CHAPTER 4
Effective improvements of patient care]. Utrecht: De Tijdstroom; 2011.

28. Creswell JW, Fetters MD, Ivankova NV. Designing a mixed methods study in primary care. Annals
of family medicine 2004; January 01;2(1):7-12.
31. Zwart DLM, Bont de AA. Introducing incident reporting in primary care: A translation from safety
science into medical practice. Health, risk and society 2013;13(3):265-78.
74 CHAPTER 4
APPENDIX I Search strategy PubMed
((health care evaluation mechanisms[mesh]) OR (intervention*[tiab]) OR (improv*[tiab]) OR

(chang*[tiab]) OR (effect*[tiab])) AND (((((((((((((((((((((((“General Practice, Dental”[Mesh]))
OR (dental care[Mesh])) OR (dental care[tiab])) OR (“Dental Hygienists”[MeSH])) OR (“Dental
Hygienists”[tiab])) OR (“Dietetics”[Mesh])) OR (“Dietetics”[tiab])) OR (“Exercise Therapy”[Mesh]))
OR (“Exercise Therapy”[tiab])) OR (“Physical Therapy Specialty”[Mesh])) OR (physical therapy[tiab]))
OR (occupational therapy[MeSH Terms])) OR (occupational therapy[tiab])) OR (“Midwifery”[Mesh]))
OR (“Midwifery”[tiab])) OR (“General Practice, Dental”[Mesh])) OR (skin therapy[tiab])) OR
(“Speech Therapy”[Mesh])) OR (“Speech Therapy”[tiab])) OR (“Family Practice”[Mesh])) OR
(“Family Practice”[tiab])) OR (general practice[tiab]) OR (primary care[tiab]) OR (“Primary Health
Care”[Mesh])) AND (((((organizational culture[mesh]) OR (organizational culture[tiab]) OR
(organisational culture[tiab] OR (organizational climate[tiab]))) AND (safety[tiab])) OR ((safety
management[mesh]) AND (culture[tiab] OR (climate[tiab]))) OR ((patient safety[mesh]) AND
(culture[tiab] OR (climate[tiab]))) OR ((culture[tiab] OR (climate[tiab])) AND (safety[tiab]))))
CHAPTER 4
APPENDIX II Excluded articles
Study Reason for exclusion

Bowie (2010) No intervention
Evans (2012) No intervention
Gehring (2012) Descriptive study
Gonzalez (2011) Study protocol
Jacobs (2012) No report of culture effects
Linzer (2005) No intervention
Masotti (2009) No intervention
McKeon (2009) No report of culture effect, subject was nurse training
Milligan (2007) Descriptive study
Palinkas (2011) No intervention
Singh(2006) No intervention
Singh(2009) No intervention
Smith (2004) No intervention
Sorokin (2011) No intervention
Terol (2008) No intervention
Tham (2011) Setting was pediatric hospital

CHAPTER 5
Cluster randomized, controlled trial on patient

safety improvement in general practice: a study
protocol
Published as:
Verbakel NJ, Langelaan M, Verheij TJM, Wagner C, Zwart DLM.

Cluster randomized, controlled trial on patient safety improvement in general practice: a
study protocol.
BMC Family Practice 2013, 14:127
ABSTRACT
Background
An open, constructive safety culture is key in healthcare since it is seen as a main condition
for patient safety. Studies have examined culture improvement strategies in hospitals. In
primary care, however, not much is known about effective strategies to improve the safety
culture yet. The purpose of this study is to examine the effect of two patient safety culture
interventions: a patient safety culture questionnaire solely, the SCOPE, or the SCOPE questi-
onnaire combined with a patient safety workshop. The purpose of this paper is to describe
the rationale and design of this trial.
Methods/design
The SCOPE Intervention Study is a cluster randomized, three-armed controlled trial, that will
be conducted in 30 general practices in the Netherlands. Ten practices in the first interven-
tion arm will complete the SCOPE questionnaire and are expected to draw and implement
their own improvement initiatives based on a computerised feedback report. In the second
intervention arm, staff of the ten practices also will be asked to complete the SCOPE ques-
tionnaire and in addition will be given a complementary workshop. This workshop is theo-
retical and interactive, educating staff and facilitating discussion, leading to a practice spe-
cific action plan for patient safety improvement. The results of the SCOPE questionnaire are
incorporated in the workshop. The ten practices in the control arm continue care as usual.
CHAPTER 5
Baseline and follow-up measurements will be conducted with an implementation period of

one year. The primary outcome will include the number of incidents reported and seconda-
ry several quality and safety indicators and the patient safety culture. Moreover, interviews
will be conducted at follow-up to evaluate the implementation process of the intervention.
Discussion
Results of this study will give insight in the effect of administering a culture questionnaire
or the questionnaire with a complementary workshop. This knowledge will aid implemen-
tation of patient safety tools and future research. Attention has been given to the strengths
and limitations of the study.
Trial registration: Netherlands Trial Register: NTR3277.
78 CHAPTER 5
BACKGROUND
A main condition for patient safety is an open constructive safety culture. Patient safety
culture is described as the values, attitudes, norms, beliefs, practices, policies, and
behaviours regarding safety issues in daily practice.1 One of the main recommendations
in the Institute of Medicine report ‘to Err is Human’ was to support a safety culture. The
National Patient Safety Agency in the UK also recognizes the importance of an open culture.
In their developed “Seven steps to patient safety for primary care” the first step is to “build
a safety culture”.2 In a report about safety in healthcare in the Netherlands, the former
director of Shell, called an environment where acknowledging mistakes is taboo, one of the
main causes of safety-risks.3 Nonmedical industries have been working on safety for much
longer and showed that an open culture on error ameliorates business performance.4,5
Reports suggest a similar role of safety culture in healthcare.6,7
In hospital care, team training and communication, executive walkrounds and Comprehensive
Unit-Based Safety Program (CUSP) are well received interventions to improve patient
safety culture that have been studied. Although positive effects were reported, the level
of evidence moderates firm conclusions on the effectiveness of patient safety culture in
healthcare.8,9 Despite the fact that a large part of healthcare is delivered in primary care
where practice organisations are becoming larger scaled and more complex, leading to
CHAPTER 5
increasing importance of patient safety issues, the effectiveness of such improvement
strategies in primary care is underexposed.Often, the first step to initiate patient safety
culture improvements is to measure the current state of affairs. We have developed and
validated a patient safety culture questionnaire for general practice: the SCOPE.10 During this
former study we observed that this culture questionnaire raised awareness and stimulated
some professionals to change their practice. The conducting of a survey can be perceived
as a measurement tool and also as a vehicle for communication. It is stated that the actual
administration of a survey operates as an intervention. The survey affects people’s perceptions
and sends messages to employees about the importance of the topic it addresses.11 Also,
feedback of patient safety culture surveys, combined with benchmark data are found
highly informative.12 Others observed a possible intervention effect of conducting a culture
questionnaire.9,13,14 However, it is not clear what the magnitude and the sustainability of
the application of a single questionnaire is. It appears that professionals find it difficult to
shape actual improvement in practice.15 We expect that the effect of a single questionnaire
could only be temporarily, and subsequently, that the raised awareness will fade away and
thus will not lead to actual safety culture improvements. Sexton et. al developed a tool to
discuss results of a culture questionnaire as they state that without such a tool it would be
Patient safety improvement in general practice: a study protocol 79

unlikely that spontaneous discussions lead to meaningful improvements in cultures, given
the relative novelty of safety culture and its complexity.16 This corresponds to the reasoning
that “the process of reporting results is perhaps most important in determining a survey’s
effectiveness as a cultural change tool”.11 Hence, a more practical and comprehensive
intervention seems needed for obtaining profound and lasting results. Following this, we
develop a complementary workshop to the SCOPE questionnaire, based on the Manchester
Patient Safety Framework (MaPSaF).17 The MaPSaF is developed by the NHS specifically for
primary care. The tool aims at helping primary care practices to assess the current level of
maturity (pathological, reactive, bureaucratic, proactive, generative) of their approach to
patient safety. The tool’s output serves as a basis for discussions on how to improve the
practices’ patient safety. Our approach resembles the CUSP, an eight step programme to
improve safety culture, used in hospitals.18 The first step is to measure the culture, followed
by the science of safety, identification of safety concerns by staff, adopting of a unit by
senior executives, implementation of improvements, analysing and documentation of
efforts, sharing of results and last, reassessment of culture.
Objectives
The first objective of this study is to examine the effect of two interventions on patient safety
behaviour and patient safety culture in general practice: the SCOPE questionnaire solely,
and the SCOPE questionnaire combined with a safety culture workshop. We conduct a three
CHAPTER 5
armed trial instead of a two-armed trial. The purpose of the ‘questionnaire only’ arm is
twofold. Firstly, to assess whether administering a culture questionnaire with only a feedback
report has an effect on patient safety behaviour and culture, compared to the control arm.
Secondly, to be able to adjust for the possible intervention effect of the questionnaire in
the workshop arm. Our second objective is to evaluate the implementation process of both
interventions. Designing, implementing and evaluating a patient safety culture intervention
is complex. The direction of results will largely depend on the context.19-22 Therefore, evenly
important as the possible effect of the interventions is the process evaluation.
METHODS/DESIGN
Design and setting

The SCOPE Intervention Study is a cluster randomized, three-armed controlled trial
complemented with a qualitative study. The study will be conducted in thirty general
practices in the Netherlands. Practices selection and randomization All general practices (n
= 350) in Utrecht area received an invitation to participate in the study. Practices that consist
of at least three employees of whom at least one GP can participate in the study. In addition,
80 CHAPTER 5
the SCOPE questionnaire should not have been completed in the past two years. Stratified
randomization will be used to allocate the practices in the three trial arms (see Figure 1).
Stratification is based on practice size and whether a practice is accredited on the Dutch
GP Practice Accreditation system23, as we expect these parameters to possibly confound
the effects on patient safety culture. The randomisation will be performed by the Data
Management Unit of the Julius Center, independent of the research team. Because of the
nature of the intervention blinding is not feasible.
350 practices
Inclusion criteria: receive an
· the practice has at invitation
least three employees,
from which one
physician 30 practices to
· the practice has not enroll in the study
completed the SCOPE
questionnaire in the
past two years.
Small practices, Small practices, Large practices, Large practices,

no accreditation accreditated no accreditation accreditated
Intervention II
Intervention II
Intervention II
Intervention II
CHAPTER 5
Intervention I
Intervention I
Intervention I
Intervention I
Control
Control
Control
Control
Merging all practices in three arms:

Control (n=10)
Intervention I (n=10)
Intervention II (n=10)
Figure 1. Flowchart randomisation
Procedure
In Figure 2 an overview is given of the intervention procedure and timeframe. Practices
in the control arm continue work as usual. All practices are asked to complete a baseline
and follow-up questionnaire. At follow-up we administer the SCOPE questionnaire to all
participating practices and we will carry out interviews.

Intervention arm I
Practices allocated to intervention I receive access to the online SCOPE questionnaire
and simultaneously receive a key to download their results in a feedback report. Also, it
is communicated that they are expected to anticipate on these results themselves. After
one week a reminder is sent. Additionally, one month after the reminder an email is sent
reminding the practice of the feedback report and to inform them about the continuation of
the study. We interfere as little as possible in this research arm trying to mimic the normal
course of events, when a practice would choose for itself to use the SCOPE questionnaire as
improvement strategy.
Control Intervention I Intervention II

Baseline
Baseline Baseline Baseline
0 months
questionnaire questionnaire questionnaire
Intervention
SCOPE SCOPE
1 month questionnaire questionnaire
CHAPTER 5
Workshop
2-3 months
. . .
. . .
. . .
Follow-up Follow-up Follow-up Follow-up
12 months
SCOPE SCOPE SCOPE

13 months
Interviews Interviews Interviews

14-15 months
Figure 2 Flowchart of study procedure
82 CHAPTER 5
Intervention arm II
The practices in intervention II also receive access to the questionnaire. However, these
practices do not receive the key to download their results. Instead, they will be given a
patient safety workshop at their practice location. The feedback on the results of their
questionnaire is embedded in this workshop. The complete feedback report is handed out
at the end of the workshop.
Interventions
The intervention consists of the SCOPE questionnaire solely (intervention I) or the
SCOPE combined with a patient safety workshop (intervention II). We chose the SCOPE
questionnaire and the workshop for both practical and theoretical reasons. The European
Linneaus project recommends the AHRQ safety culture questionnaire, from which the
SCOPE has been derived, and the MaPSaF for primary care.17,24 The SCOPE questionnaire and
the Dutch translation of the MaPSaF were both readily available and translated in Dutch. In
addition, the tools combine well together as the dimensions largely correspond with each
other, facilitating the alignment of the workshop complementary to the questionnaire.
SCOPE questionnaire
The SCOPE questionnaire is a culture questionnaire for general practices. The SCOPE is
derived from the HSOPS and validated in Dutch general practice,10,25,26 Cronbach’s alpha
CHAPTER 5
ranged between 0.64 - 0.85. The questionnaire consists of 43 items divided over eight
dimensions:
1. Handover and teamwork (8 items);
2. Support and fellowship (5 items);
3. Communication openness (6 items);
4. Feedback about and learning from error (6 items);
5. Intention to report events (3 items);
6. Adequate procedures and adequate staffing (7 items);
7. Overall perceptions of patient safety management (4 items);
8. Expectations and actions of managers (4 items).
Items are answered using a five point scale varying from ‘strongly disagree’ to ‘strongly
agree’ or ‘never’ to ‘always’. In addition, respondents are asked to grade the patient safety
culture in their practice. Also, questions on demographics such as gender, age and years
of working experience are included. All staff of each practice are asked to complete this
questionnaire. For data collection and storage an online system will be used, managed by
the Dutch GP Practices Accreditation Organisation.23

SCOPE feedback report
Results of the completed questionnaires are presented in a computerized feedback report.
In this report it is presented how often the questionnaire is completed and by which
disciplines. Per dimension the percentage positive scores (four or five on the five point-scale)
are calculated. When this is 75 percent or higher it is perceived as a “strong” dimension.
When this percentage is lower than 50 percent it is perceived as a “weak” dimension and,
when scored in between it is “neutral”. The eight dimensions, the percentage positive scores
and their classification are presented in a table. Subsequently, all complete dimensions
with all questions are reported. Per question the average practice score is reported and
compared with an overall benchmark mean that is calculated from data of all practices that
have completed the SCOPE questionnaire in 2008 (n = 506-587). When a question scores
more than fifteen percent lower than this mean it is depicted red and when higher green.
By this, practices can inform themselves on their performance being average, below or
above, as compared to the benchmark average. The last page of the report gives general
suggestions to improve the patient safety in practice. Suggestions are not fully worked out,
but address some issues and refer to more reading material and organizations that can be
consulted, if desired. The feedback report is downloaded from the same webpage as where
the questionnaires are completed.
Workshop
CHAPTER 5
The workshop is based on the Dutch translation of the MaPSaF.27 The MaPSaF is a matrix of
nine dimensions in which for each dimension all five maturity stages (pathological, reactive,
bureaucratic, proactive, generative) of patient safety are described. We add items on theory
on patient safety, human factors engineering and safety culture (Figure 3). The workshops
are organized at each practice location to make it easier for staff to attend. It requires three
and a half hour and at least 75% of the staff should be present. The workshops are given
by both an educational scientist who also is a GP and one of the researchers (NV). We
intentionally chose for one of the trainers to be an outsider of the research project as well
as to be a GP. The first feature allows for questioning and interpreting independently of the
research project. The second feature, the trainer being a GP, may allow more rapidly gaining
a certain level of understanding and trust among the participants because of being familiar
with the GPs practice and context. We believe that the content of the workshop will be
better conveyed when explained by a GP. In addition, as the dialogues can contain intimate
content, for example when discussing an incident or flaws in communication between staff,
a GP as trainer may be easier to confide in and can also display more understanding of
the situation. The researcher attending the workshop gives the opportunity to observe and
gather research data. Being part of the research setting is linked with intimate knowledge of
84 CHAPTER 5
the situation, which is essential to develop an understanding ‘from within’.
The workshop is both theoretical and interactive, facilitating discussion among practice staff
about their own safety culture.
Workshop programme
● Introduction to patient safety
- Discussing patient safety terminology
- Data on number of incidents internationally and nationally
● Human factor engineering
- Why do people make mistakes
- Interactive examples
- System approach
● Classify organization according to the MaPSaF vignettes on two dimensions
(individually)
- Each respondent classified the maturity of their practice for two dimensions
without consultation
● Patient safety culture
- Theory on patient safety culture
● Feedback on SCOPE questionnaire
- Discussion about results
● Dialogue about own patient safety culture based on vignettes
- Vignettes are discussed in pairs (trying to align with each other)
- Vignettes are discussed with all staff
● Brainstorm on possible improvement actions
CHAPTER 5
● Drafting of practice improvement action plan
● Evaluation & take home message
Figure 3. Workshop programme
In consecutive order we cover an introduction to patient safety, including discussion about

terminology and international and national data about patient safety incident numbers.
Followed by theory and interactive examples of human factor engineering and a systems
approach to error. Subsequently, we ask all staff to classify the maturity stage, following the
MaPSaF matrix, on two vignettes. A vignette is an A4-paper with one dimension worked
out in five descriptions of this patient safety theme according to the five stadia. Staff are
asked to choose the description that resembles their own daily practice the most. Per
workshop we discuss two dimensions: the dimensions that scores the lowest on the SCOPE
questionnaire. To be sure the same vignettes will be used for the same SCOPE dimension
we made a compatibility table of the SCOPE dimensions and the MaPSaF dimensions (Table
1). The first two SCOPE dimensions have the same MaPSaF dimension. When these are the
weakest dimensions, we will use the three weakest SCOPE dimensions in order to have two

different vignettes to discuss. The two vignettes in each practice will be different dependent
on their SCOPE results. After scoring these vignettes we introduce theory on culture as an
important aspect of patient safety. Subsequently, we show the practice results on their
SCOPE questionnaires at dimension levels and ask whether these are recognized and discuss
these. Next, we will ask to discuss the MaPSaF vignettes in pairs and subsequently in the
whole group. During the workshop we facilitate discussion about the patient safety culture
of the practice using the results of the SCOPE, the vignettes and other themes that emerge,
leading to a brainstorm of possible improvements. At the end of the workshop the staff will
draw a practice specific improvement plan. The workshop ends with an evaluation and a
round with take home messages from everyone.
Table 1 Compatibility table of SCOPE questionnaire and MaPSaF dimensions

SCOPE10 MaPSaF27
Handover and teamwork Teamwork
Support and fellowship Teamwork
Communication openness Communication about patient safety
Feedback and learning from error Learning from errors and achievement of improvement
Intention to report events Registration and evaluation of errors
Adequate procedures and adequate staffing Personnel management and safety issues
(Resources)
CHAPTER 5
Overall perceptions of patients safety management Priority given to patient safety

(Staff education and training aimed at patient safety)
Expectations and actions of managers Errors and responsibility for patient safety
Pretesting workshop
The workshop has been piloted during a training day in six general practices. The aim of
this pilot was twofold, first to evaluate the workshop and to be able to customize possible
improvements. Second, to give the trainers a chance to get acquainted to the programme.
The workshop was well received, main adjustments were to print out a format for the action
plan to take home and to print out the feedback report and handing them out directly after
the workshop instead of e-mailing them afterwards.
Facultative workshops after ending the study

During the recruiting of the practices, we will communicate that a facultative workshop
will be offered for all practices allocated to the control and intervention I arm after ending
the study. Hereby, we aim to prevent selective drop-out of practices in the control and
intervention I arm.
86 CHAPTER 5
Measurements
a. Patient safety behaviour

The primary endpoint is the number of incident reports reported by staff in the practice.
This endpoint is chosen as incident reporting gives an indication of the patient safety
behaviour and culture in the practice. We hypothesize that an increase of reported incidents
corresponds to an open patient safety culture. Secondary endpoints are the presence of
quality and safety indicators in the previous year such as ‘the presence of a procedure for
complaints’, ‘how often safety was on the agenda and/or discussed during team meetings’,
‘whether a safety management policy was present’ and ‘whether safety was subject of staff
education’. These outcomes are measured using a practice questionnaire at baseline and
follow-up in all study arms.
b. Patient safety culture

As stated in the introduction, a culture questionnaire may have an effect on patient safety
culture. Therefore, we deploy the SCOPE questionnaire as an intervention. However,
as the SCOPE questionnaire will also provide information about the prevailing culture
simultaneously, we do not refrain from interpreting and using this data for analyses. Practices
in the intervention arms will complete the questionnaire at the beginning of the study as
a (part of the) intervention and at follow-up as a measurement tool. The practices in the
CHAPTER 5
control arm will only complete the SCOPE questionnaire at follow-up as measurement tool.
As such, data on the development of culture will be available for the intervention groups
and differences between groups will be available at follow-up.
c. Process evaluation
Besides the effect of the intervention we want to examine the implementation process. As
a complex intervention is dependent on contextual factors we want to study these in depth
to be able to address facilitators and barriers of the intervention. Therefore, we conduct
interviews with the physicians and other staff of the practice. Interviews are conducted by
a semi-structured format using a topic list. Topics will examine the patient safety behaviour
and culture. First the actual activities are assessed in reference to the research arm where
the practice is allocated. Subsequently, patient safety themes that come up during the
interview are scrutinized. For example, practices in intervention II will be questioned on
their follow-up of the action plan that has been drawn during the workshop. Which activities
are implemented and to which level? How did they approach this activity, and what were
barriers and facilitators? Practices in the control arm will be questioned on how they
perceive patient safety and on what they actually do in their practice around this theme.

Interviews are held in an iterative design, by two interviewers. Every week the interviewers
will discuss their data briefly to evaluate and, if necessary, to adjust the topic list accordingly.28
Statistical power
The power calculation for the effect of the interventions on patient safety behaviour is based
on the primary outcome, incident reporting, and resulted in a power of 0.90. The following
assumptions are used: 30 practices divided in three equal groups; an improvement of
reported incidents in a year (from 5013 to 70 (intervention I) to 100 (intervention II) incidents
per practice, standard deviation of 30 and an alpha of 0.05.
Ethical approval
The Medical Research Ethics Committee of the University Medical Center Utrecht concluded
that the Medical Research Involving Human Subjects Acts does not apply.
Implementation of study results

The results of the SCOPE Intervention Study will result in a self-employable product with
a guideline that can be used by professionals in primary care to improve their patient
safety and culture. To transfer the knowledge acquired during this study among general
practitioners we will organize a meeting with representatives of the primary care professional
associations other than of general practice. Here, we will present our results and discus
CHAPTER 5
possible adjustments of the tool and possible additional information needed to shape the
workshop so that these professions in primary care, such as physiotherapy and midwifery,
can use this tool as well.
Data analysis
To analyse the number of reported incidents we will use a poisson regression, if necessary,
the analysis will be adjusted for over- or underdispersion.29 Baseline characteristics such as
number of incidents, size of the practice and accreditation will be included as confounders.
Where possible we will adjust for baseline measurement of patient safety culture. We will
describe patient safety behaviour measured by complaints, meetings and other quality and
safety indicators and compare baseline with follow-up. The development of patient safety
culture in the two intervention groups and differences in culture between the three arms at
follow-up will be analysed by mean scores of the dimension using mixed linear models. All
analyses will be corrected for clustering within practices. If necessary a multiple imputation
technique will be used for missing data. Data collected from staff during the interviews will
be transcribed and analysed with thematic content analysis using software NVivo to code
and analyse the data.30
88 CHAPTER 5
DISCUSSION
The purpose of this paper is to outline the rationale and design of the SCOPE Intervention
Study. This study will provide insight in the effect of conducting a safety culture questionnaire
with a feedback report, on patient safety behaviour and culture in general practice. In
addition, this study will reveal whether a complementary workshop to a patient safety
culture questionnaire adds to the effect on safety behaviour in general practice. Lastly,
interviews will shed a light on the implementation process of the interventions.
This study has several strengths. The SCOPE Intervention Study is one of the first studies
that examines the effect of an intervention in primary care on patient safety behaviour and
culture. Moreover, the design, a controlled trial, will provide more trustworthy results than
previous studies which were observational. Another strength is that the second part of the
design is qualitative and will shed light on the implementation process of the interventions.
By conducting interviews with practice staff we will gain a deeper understanding on how
the interventions work. Several limitations have to be considered also. Firstly, we ask
practices to voluntary participate in our study. This may lead to selection bias. For instance,
it is likely that the most motivated practices will decide to participate. However, in daily
practice forerunners will also be the first to implement patient safety improvements. By
studying the effects and implementation of such interventions we hope to facilitate broader
implementation. Secondly, we realise that the number of incidents as outcome is ambiguous
CHAPTER 5
as both increasing and decreasing numbers could indicate an improvement in patient safety
culture and behaviour. However, we believe that reporting incidents is a good measurement
of the change in patient safety. Especially in an organisation where patient safety initiatives
are relatively new and the number of incident reports are likely to raise before they will
lessen.31 As reporting is still very uncommon in general practice, we will consider an increase
of reported incidents as an indicator of an ameliorating safety culture. Increased rates will
indicate the starting of safe reporting and raised awareness. Lastly, interventions in this study
are complex and may have a diffuse effect. This may be difficult to measure quantitatively.
Therefore, we designed a study with mixed methods to understand the potential effect. The
strength is that results will reflect daily practice and approximates the effect to be attained
when this intervention would be employed on a large scale.
This study will contribute to the body of knowledge concerning the effect of patient safety
interventions in general practice. This knowledge will enhance implementation of patient
safety tools in general practice and other primary care professions.

REFERENCES
2003;38(3):923-45.
2006.
you work safe, or you do not work here. Faster better - Safety in healthcare]. Den Haag: Shell
Nederland; 2007.
4. Clarke S. The relationship between safety climate and safety performance: a meta-analytic
review. Journal of occupational health psychology 2006;11(4):315-27.
5. Dyck van C, Frese M, Baer M, Sonnentag S. Organizational error management culture and its
impact on performance: a two study replication. Journal of applied psychology 2005;90(6):1228-
40.
6. Haynes A, Weiser T, Berry W, Lipsitz S, Breizat AH, Dellinger,,E,Patchen, Dziekan G, Herbosa T,
Kibatala P, M, Merry A, Reznick R, Taylor B, Vats A, Gawande A. Changes in safety attitude and
relationship to decreased postoperative morbidity and mortality following implementation of a
checklist-based surgical safety intervention. BMJ Quality and Safety 2011;20(1):102-7.
7. Mardon RE, Khanna K, Sorra J, Dyer N, Famolaro T. Exploring relationships between hospital
patient safety culture and adverse events. Journal of patient safety 2010;6(4):226-32.
CHAPTER 5
practices 2011;12:117.
11. Wagner DB, Spencer JL. The Role of Surveys in Transforming Culture. Data, Knowledge, and
Action. In: Kraut AI, editor. Organizational Surveys: Tools for Assessment and Change. 1st ed. San
Francisco, California: Jossey-Bass Inc.; 1996. p. 67-87.
12. Pronovost P, Sexton B. Assessing safety culture: guidelines and recommendations. Quality and
safety in health care 2005;14(4):231-3.
2011;20(2):121-7.
14. Shojania KG, Thomas EJ. Trends in adverse events over time: why are we not improving? Quality
and safety in health care 2013;22:273-7.
15. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in
patients’ care. The Lancet 2003;362:1225-30.
16. Sexton JB, Paine LA, Manfuso J, Holzmueller CG, Martinez EA, Moore D, Hunt DG, Pronovost PJ.
A check-up for safety culture in “my patient care area”. Joint Commission journal on quality and
90 CHAPTER 5
patient safety 2007;33(11):699-703.
17. Pronovost P, Weast B, Rosenstein B, Sexton JB, Holzmueller CG, Paine L, Davis R, Rubin HR.
Implementing and Validating a Comprehensive Unit-Based Safety Program. Journal of patient
safety 2005;1(1):33-40.
18. Bartholomew LK, Parcel GS, Kok G, Gottlieb NH, Fernandez ME. Planning Health Promotion
Programs: An Intervention Mapping Approach. 3rd ed. San Francisco: Jossey-Bass; 2011.
19. Stevens DP, Shojania KG. Tell me about the context, and more. BMJ Quality and Safety
2011;20(7):557-9.
20. Singer SJ, Vogus TJ. Safety climate research: Taking stock and looking forward. BMJ Quality and
Safety 2013;22(1):1-4.
2011.
23. Linneaus Euro-PC. Available at: www.linneaus-pc.eu. Accessed April 12th, 2013.
26. Struben V, Wagner C. Ontwikkeling van een Instrument voor Zelf Evaluatie van de
Patiëntveiligheidscultuur (IZEP) [Development of an instrument for self-evaluation of the patient
safety culture]. Utrecht: NIVEL; 2006.
CHAPTER 5
27. Boeije H. Analyseren in kwalitatief onderzoek. Denken en doen. [Analysis in qualitative research.
Thinking and doing]. Den Haag: Boom Lemma uitgevers; 2012.
28. Faraway JJ. Extending the Linear Model with R. Generalized Linear, Mixed Effects and
Nonparametric Regression Models. Boca Raton: Chapman & Hall/CRC Taylor & Francis Group;
2006.
29. QSR International Pty Ltd. NVivo qualitative data analysis software 2012;10.
30. Leistikow I. Patiëntveiligheid, de rol van de bestuurder. [Patient safety, the role of the director].
PhD thesis [dissertation]. Elsevier gezondheidszorg, Amsterdam: University of Technology Delft;
2010.

CHAPTER 6
A cluster randomized trial on the effects of

patient safety culture interventions in general
practice
Verbakel NJ, Langelaan M, Verheij TJM, Wagner, Zwart DLM
Submitted
ABSTRACT
Background
Having a constructive safety culture is essential for successful implementation of patient
safety improvements. Because it was unclear how to improve culture in general practice we
facilitated two culture tools for general practices.
Aim
To assess the effect of two patient safety culture interventions on safety culture.
Design and setting
A cluster randomised trial was conducted in a mixed method study, studying the effect of
administering a patient safety culture questionnaire (intervention I), the questionnaire com-
plemented with a practice-based workshop (intervention II) and no intervention (control) in
thirty general practices in the Netherlands.
Method
The primary outcome, the number of incidents reported, was measured with a question-
naire at baseline and a year after. Secondary outcomes were quality and safety indicators
and safety culture. Generalized linear models were used for analysis.
Results
The number of incidents increased in both intervention groups, to 81 and 214 in interven-
tion I and II respectively. Adjusted for baseline number of incidents, practice size and accre-
ditation status the study showed that practices that additionally participated in the work-
shop reported 40 times more incidents compared to the control group. Practices that only
completed the questionnaire reported 5 times more incidents. There were no statistically
significant differences in staff’s perception of patient safety culture at follow-up between
CHAPTER 6
the three study groups.

Conclusions
Educating staff and facilitating discussion about patient safety culture in their own practice
leads to increased reporting of incidents. It is beneficial to invest in a team-wise effort to
improve patient safety.
Trial registry number: Netherlands Trial Register: NTR3277.
94 CHAPTER 6
INTRODUCTION
At the very start of patient safety research the Institute of Medicine stated that “healthcare
organizations must develop a culture of safety to focus on improving the reliability and safety
of care for patients”.1 Originated from organizational culture, safety culture is described
as the product of individual and group values, attitudes, perceptions, competencies, and
patterns of behaviour that determine the commitment to, and the style and proficiency of,
an organization’s health and safety management.2
In the Netherlands, patient safety policy in general practice is now being developed and
among others there is a clear need for effective intervention tools on safety culture. Two
reviews examining culture improvement strategies in hospitals showed broad multipart
interventions and walk rounds with engaged leaders to be most successful.3, 4 Although a
large part of healthcare is delivered in primary care, a review resulted in only few studies on
interventions affecting its culture.5
Surveys, initially developed to measure existing culture,6-10 have been observed to possibly
affect aspects of safety culture.3, 11, 12 Administering a survey draws attention to the topic,
influencing staff and as such can be considered an intervention.13, 14 The advantage of a
survey is the usability and relatively low cost, however, when considering it as a safety
culture intervention, it is questionable whether it is strong enough to accomplish sustained
changes on its own. Indeed, the effectiveness of a survey as change tool is determined by
the process of digesting and reporting the data.13
In addition, educational activities, like workshops, showed positive results on risk
management and safety culture.15, 16 The Manchester Patient Safety Framework (MaPSaF) is
a discussion tool for assessing and improving the maturity of safety culture in primary care
CHAPTER 6
settings17 that is increasingly being used.18-20
The objective of our study was to assess the effect of administering a culture questionnaire
with digital feedback or the questionnaire combined with a practice-based workshop
including their feedback in general practice. We hypothesized that both interventions would
lead to improved patient safety culture relative to the control practices, and that practices
receiving the workshop would improve the most.
METHOD
Design and participants

We conducted a three-armed cluster randomised trial, in a mixed methods study. In this
paper quantitative results will be presented. Randomisation was stratified based on practice
size (small: < 8 employees, large: ≥ 8 employees), and accreditation status (Figure 1). The
Effects of patient safety culture interventions in general practice 95

minimisation technique was used taking into account the strata, performed by the data
management department independent of the research team. The practices were enrolled
and contacted by the first author (NV). Details of the study protocol were described
previously.21 Due to the nature of the interventions blinding was not possible.
Enrolment Invitation practices (n=350)
Enrolment (n=30)
Stratified
randomisation
Small, not accredited Small, accredited Large, not accredited Large, accredited
(n=17) (n=5) (n=5) (n=3)
Control (n=1)
Control (n=6)
Control (n=2)
Control (n=1)
Baseline
Control Intervention I Intervention II

(n=10) (n=10) (n=10)
Intervention Because 2 practices

dropped out before the
As usual Intervention: Intervention: intervention commenced 1
- SCOPE questionnaire - SCOPE questionnaire practice was moved
- Patient safety culture
workshop
CHAPTER 6
Follow-up
- Lost to follow-up (n=0) - Lost to follow-up (n=0) - Lost to follow-up (n=2)

- Change of intervention (n=1) - Change of intervention (n=0) - Change of intervention (n=0)
Analysis
- Analysis (n=9) - Analysis (n=10) - Analysis (n=9)

- Excluded from analysis (n=0) - Excluded from analysis (n=0)* - Excluded from analysis (n=0)
* None of the practices were excluded in the analysis of the primary outcome; 1 practice in intervention I was excluded in the analyses of the
SCOPE questionnaires.
Figure 1. Enrolment and randomisation
96 CHAPTER 6
Interventions
Three interventions were studied. The administering of and feedback on a patient safety
culture questionnaire, the administering of the questionnaire complemented with a patient
safety workshop and no intervention.
I. Patient safety culture questionnaire

The SCOPE questionnaire, translated and adapted originally from the Hospital Survey
on Patient Safety Culture (HSOPS)22 specifically for Dutch general practices23 was used as
intervention tool. Practices simultaneously received a login for the SCOPE questionnaire
and a key to download their results in a report. Practices were reminded twice to complete
the questionnaire and to download their report. An online system was used for collection
of the SCOPE data.24
II. Practice based patient safety workshop

The workshop, led by an independent GP trainer, provided education on the concept of
patient safety and culture, terminology and human factor engineering and was held at the
practice location. Discussion about the own culture was facilitated using their own SCOPE
results and Dutch translations of MaPSaF17 focusing on two SCOPE dimensions that scored
lowest. A brainstorm of possible improvements resulted in an action plan. Attendance
of 75% staff was required. To study the course of the workshop one of the authors (NV)
attended and kept observations. Also, participants were asked to complete an evaluation
form.
Outcome measurements
CHAPTER 6
Primary outcome
The primary outcome was the number of reported incidents per practice at follow-up,
measured with a questionnaire at baseline and one year hereafter. Actual reporting is a
prominent feature of a generative safety culture.25 Because reporting is just in its infancy in
general practice26, 27, we hypothesized that an increase of reports would reflect a ‘pattern of
behaviour’2 congruent to improvement of patient safety culture. Hence, we considered the
number of incidents reported as a proxy of actual patient safety culture.
Secondary outcome
Patient safety culture was additionally operationalized by quality and safety indicators
(e.g. the presence of complaints procedure, patient safety being an agenda item of team
meetings, see Appendix I).

Patient safety culture was measured at all practices at follow-up using the SCOPE questionnaire,
consisting of 43 items distributed over eight dimensions (Table 3). SCOPE has sound psychometric
properties: Cronbach’s alpha’s: 0.64 - 0.85.23 Two outcome questions were included: ‘Looking
back at the past 12 months, how many incidents reports did you fill-out’ and ‘How would you
grade the patient safety in your practice’ (PSG) (five-point scale from ‘failing’ to ‘excellent’).
Sample size and practice recruitment

The sample size was calculated based on the primary outcome, numbers of incident reports,
showing thirty practices were needed. Based on previous research11 we assumed an
increase from 50 to 70 (intervention I) and 100 (intervention II), respectively, with a standard
deviation of 30. Practices (n=350) were invited (February/March 2012) per mail. The first 30
practices that fulfilled inclusion criteria, ≥3 employees and not having completed the culture
questionnaire before, were enrolled and allocated to the three research groups (n=10).
Analysis
The number of incidents was analysed per practice with a generalized linear model based on
a negative binomial distribution. Intervention, number of reports at baseline, accreditation
status and practice size were included in the model. The model using a Poisson distribution
showed large overdispersion and minor violations of the assumptions of homoscedasticity
and normally distributed residuals. Therefore, we used the negative binomial distribution,
hereby deviating from the protocol.21
The quality and safety indicators were compared before and after using descriptives.
SCOPE questionnaires with >50% missing items were excluded. Multiple imputation (10
imputations) was performed on item level.28 Culture items were imputed and used as
CHAPTER 6
predictors while gender, discipline and age were used as predictors only. Because formal
management items could not be answered by everybody, these were not imputed. Therefore,
when calculating the mean scores of dimension 7 and 8 one missing was allowed. Percentages
positive scores were calculated per dimension. As described in the HSOPS manual, we adhered
to the cut off value of >75% positive scores to indicate practices’ strengths and by ≤50% positive
scores for weak dimensions.29 For two measurements in the same practice a 5% change was
considered meaningful.30 To analyse differences at follow up we calculated mean dimension
scores and performed generalized linear mixed analysis. Intervention type, practice size and
accreditation status were included in the model. All analyses were conducted in SPSS 20.0.
98 CHAPTER 6
RESULTS
Participants
After randomisation, two practices discontinued because of time-issues. Therefore one
control practice was moved to intervention II (this was the first practice allocated to the
control group). Table 1 gives an overview of practices and respondents characteristics.
Table 1 Practices and respondents characteristics

Control (n=9) Intervention I Intervention II
(n=10) (n=9)
Age* (m, sd) 44.6 (9.9) 44.3 (12.4) 41.08 (10.7)
Female gender* (percentage) 81.3 92.6 85.0
Composition of staff (n, %) n=67 n=87 n=81
GPs 35.8 35.6 33.3
Assistants 38.2 41.4 38.3
Nurses 17.7 19.5 27.2
Other 7.5 3.4 1.2
Years in current practice* (m,sd) 7.1 (6.3) 9.0 (8.6) 8.3 (6.2)
Accreditation status baseline
Yes 2 2 3
Accreditation status follow-up
yes 3 3 4
Working on 1 5 2
Formal reporting system 2 2 4
Baseline (y)
Formal reporting system 2 4 8
CHAPTER 6
Follow-up (y)
*Based on data of the SCOPE questionnaires at follow-up.
Number of incidents
Intervention I showed an increase of 66 incident reports (15 to 82), intervention II an increase
of 144 (70 to 224) and the control group a decrease from 18 to 4 (Figure 2). Appendix II
shows the distribution of incident reports, reporting procedure and accreditation status at
baseline and follow-up per practice. In intervention I there was one outlier with 57 reported
incidents at follow-up. An employee of this practice participated in a workshop on incident
reporting outside our study. An intention to treat analysis showed that intervention II
resulted in 42 times more reports than the control group, and intervention I reported 5
times as much when adjusted for baseline reports, accreditation status and practice size
(Table 2). Without the outlier mentioned earlier the effect of intervention I became non-

significant. The outcome question on number of reports in the SCOPE questionnaire showed
the same trend of increasing reports in intervention II and I, respectively (Appendix III).
Quality and safety management

Some indicators showed meaningful changes. Having a formal reporting system remained
the same in the control group, but doubled in both intervention groups (I: 2 to 4; II: 4
to 8). In intervention II more practices analysed incidents systematically (2 to 7), had an
orientation procedure for new employees (3 to 6) and patient safety was an agenda item
of practices’ meetings more often (2 to 8). During the study accreditation status of some
practices changed (control: 2 to 3, I: 2 to 3, II: 3 to 4). Particularly in intervention I practices
(n=5) reported at follow-up that they were working towards accreditation.
50
40
Number of incidents
30
20
10
0
co; baseline
co; follow−up
I; baseline
I; follow−up
II; baseline
II; follow−up
CHAPTER 6
Research arm at baseline and follow-up
Figure 2. Number of incidents by intervention at baseline and follow-up

As part of the intervention, 134 questionnaires were completed at baseline and 183 at follow-
up. One practice was excluded from analysis because only one questionnaire was completed
at both measurement moments. We included 131 and 166 questionnaires, respectively as
2 baseline and 14 follow-up questionnaires had <50% of safety items completed. A missing
analyses showed 2.6% missing items at baseline and 3% at follow-up.
Percentages positive scores ranged between 63 and 86 at follow-up (Table 3).
100 CHAPTER 6
Table 2 Effect of interventions on number of incidents at follow-up
Parameter Rate ratio1 p Rate ratio2 p Rate ratio3 p
(95% C.I.) (95% C.I.) (95% C.I.)
Intention to treat analysis with all 28 practices
Intervention I 18.45 (4.79-71.06) < 0.001 14.72 (3.72-58.20) < 0.001 5.45 (1.17-25.49) 0.031
Intervention II 56.00 (14.47-216.71) < 0.001 45.47 (11.56-178.93) < 0.001 41.72 (9.81-177.50) < 0.001
Number of incidents at 1.66 (1.02-2.70) 0.040 1.78 (1.02-3.10) 0.044
baseline (ln) - -
Accredited at baseline (y) 0.28 (0.08-1.03) 0.056

- - - -
Practice size in employees 1.22 (1.06-1.41) 0.005
- - - -
Analysis without outlier in intervention I

Intervention I 6.25 (1.54-25.42) 0.010 6.11 (1.49-25.00) 0.012 4.12 (0.92-18.44) 0.064
Intervention II 56.00 (14.47-216.71) < 0.001 46.50 (11.86-182.22) < 0.001 40.15 (9.88-163.10) < 0.001
Number of incidents at 1.37 (0.87-2.17) 0.175 1.46 (0.87-2.46) 0.151
baseline (ln) - -
Accredited at baseline (y) 0.52 (0.14-1.92) 0.327

- - - -
Practice size in employees 1.15 (0.99-1.33) 0.070

- - - -
1. Univariable analysis
Effects of patient safety culture interventions in general practice

2. Corrected for baseline number of incidents (ln)
3. Corrected for baseline number of incidents (ln), accreditation status and practice size.
101
CHAPTER 6
Several dimensions showed room for improvement (<75%), however, none were below
50%. In intervention I six dimensions improved ≥5%, in intervention II three dimensions
did. One dimension ‘support and fellowship’ decreased in intervention I. With regard to the
PSG, both intervention groups showed rather low scores at baseline. This increased with 8%
and 30% for intervention I and II, respectively. Multilevel analyses showed no differences
between groups at follow-up (Appendix IV).
Course of the workshop in Intervention II

All, but one practice met the minimal attendance, ranging from 4 to 10 caretakers (total 66).
Workshops proceeded in a pleasant atmosphere and we observed increasing willingness
to share opinions and experiences as the workshop progressed. Assigned maturity stages
of their own safety culture varied between the first four stages (pathological, reactive,
bureaucratic, proactive). All but one practice drew an action plan, predominantly about
introducing or activating a reporting procedure. Evaluation forms showed that, though
some staff was sceptic at start, responses after the workshop were fairly enthusiastic.
DISCUSSION
Summary
Aiming at contributing to the knowledge of culture interventions in general practice we
found that administering a culture questionnaire solely or integrated in a workshop both
increased reporting incidents. However, the effect was much larger in practices receiving
the workshop. Also, these practices were more active in analysing incidents and discussing
the subject during team meetings. These changes in handling incidents indicate patient
CHAPTER 6
safety culture improvement at the practices’ shop floor after a team-wise safety culture
intervention. Yet, safety culture measurements did not show large improvements nor
differences between the groups after one year of follow-up.

Our study is one of the first trials on culture improvement in general practice. Also, to our
knowledge it is the first trial that studied the possible effect of a questionnaire and indeed
found some effect. However, closer examination revealed that the increase of reporting in
this group largely occurred in three practices, of which one had participated in a workshop
outside the current study. Excluding this practice resulted in a non-significant effect. In the
practices that received the workshop the increase in incident reports was found in almost
all practices, indicating that the effect was not due to particular practices. Notably, also
practices that scored high at baseline showed improvement in Intervention II.
102 CHAPTER 6
Table 3. SCOPE dimension means, sd and percentage positive scores for the control and
both intervention groups
Dimensions Control Intervention I Intervention I Intervention Intervention
(scale 1-5) (follow-up) (baseline) (follow up) II II
m (sd) m (sd) m (sd) (baseline) (follow-up)
% positive % positive % positive m (sd) m (sd)
% positive % positive
1 Handover and 3.72 3.58 3.77 3.72 3.80
teamwork (0.46) (0.67) (0.49) (0.48) (0.37)
69.6% 63.4% 74.8% 71.8 74.8%
2 Support and 4.05 3.94 3.86 3.99 4.13
fellowship (0.50) (0.55) (0.73) (0.49) (0.55)
85.3% 82.4% 75.8% 82.8% 83.7%
3 Communication 4.16 3.91 4.06 4.13 4.22
openness (0.51) (0.70) (0.49) (0.57) (0.43)
85.6% 73.6% 80.9% 81.3% 85.6%
4 Feedback about 3.95 3.94 4.04 3.91 4.15
and learning from (0.84) (0.86) (0.65) (0.77) (0.61)
error 69.8% 69.8% 75.6% 69.5% 75.0%
5 Intention to 3.84 3.76 3.90 3.84 3.99
report events (0.88) (1.00) (0.89) (0.93) (0.71)
62.6% 62.7% 68.9% 64.7% 68.2%
6 Adequate pro- 3.83 3.73 3.96 3.91 3.92
cedures and ade- (0.49) (0.56) (0.45) (0.54) (0.54)
quate staffing 72.5% 70.1% 80.4% 75.2% 77.9%
7 Overall percep- 3.66 3.65 3.75 3.63 3.94
tions of patient (0.67) (0.62) (0.57) (0.63) (0.54)
safety 65.5% 64.0% 69.2% 61.5% 84.7%
CHAPTER 6
management
8 Expectations 3.71 3.68 3.78 3.67 3.84
and actions of (0.63) (0.61) (0.54) (0.59) (0.50)
managers 69.9% 66.5% 72.2% 70.2% 75.9%
Patient safety 3.63 3.57 3.65 3.57 3.84
grade (0.64) (0.83) (0.75) (0.79) (0.49)
58.3% 61.0% 69.1% 54.9% 85.2%
Percentages depicted in bold show differences ≥5%.
This study has some limitations. Firstly, during the study quality improvement initiatives
emerged, particularly in the questionnaire only group. Five practices appeared to be
working on the Dutch practice accreditation system (NHG Praktijk Accreditering®), which
requires an incident reporting system. Further examination showed that the number of
incident reports remained the same before and after the intervention indicating that the
accreditation process for these five practices did not change their reporting behaviour.
However, for future studies it would be advisable to include only fully accredited practices
to avoid this potential confounder.
Secondly, the absence of changes in culture measurements may be due to underpowering

because the sample size calculation was based on the number of incidents, a practice feature.
The culture questionnaires, however, were conducted at caregiver level, which resulted in
clustered data requiring higher numbers of participants for measuring a potential significant
effect. Also, we were not able to match individual questionnaires before and after.

The MaPSaF, when used in hospitalized setting, showed improvement in culture measurements
over a five year follow up.31 We did not find such improvement in our study. Interestingly,
using the MaPSaF in general practice Hoffmann et al. found effects on incident reporting
and not on self-reported culture improvement similar to our study.32 This lack of effect on
culture measurements may be explained either by the short intervention time of one year
or by the insufficient sensitivity of a survey for measuring perceptions of safety culture.33
Our workshop was an adapted version of the MaPSaF tool. An important asset of the
MaPSaF tool is stimulating participants to self-reflect on daily work within their team.18
Team effort seems crucial for patient safety.34-37 However, as the MaPSaF is extensive, we
integrated the SCOPE results in the workshop, both focusing the discussion and tailoring it to
the participating practice. Hereby, the workshop became comprehensive and manageable.
Furthermore, we added an educational part on safety science to the workshop. Education is
perceived important in quality improvement38 and as the most important factor to improve
CHAPTER 6
patient safety in primary care.39 The aim was to educate staff on safety science providing
them with a sense of urgency concerning safety in general practice in order to instigate
change. In addition, it supported participants’ understanding of the systems approach,
ensuring a safe atmosphere to discuss culture. With these consecutive elements we built
the workshop on the experiential learning principles of Kolb e.g. concrete experience,
reflection, conceptualization and experimentation.40 The subsequent order of the elements
of education and presentation of own practice results (what?), team based reflection on
own practice data (so what?) and team based development of action plan (now what?) is
in line with this experience-based learning cycle that mostly fits professionals because it
explicitly connects daily practice with the learning. Moreover, the workshop resulted in an
action plan made up by all staff, thus matching their practice with team based commitment,
increasing the feasibility of actual implementation.41 We belief that this format has added to
the workshop’s impact found in our study.
104 CHAPTER 6
Implications for practice
Applying a culture survey is a convenient way to enhance staff involvement in patient safety
culture improvements. However, discussing the results together as a team when embedded
in a workshop appeared to be more effective. For future research it is worthwhile to study
the sustainability of the results found and the need for repeated interventions. An additional
challenge hereby is to determine whether practices that changed their behaviour concerning
patient safety issues deliver better care than practices that do not invest in patient safety
culture change.
CHAPTER 6

REFERENCES
5. Verbakel NJ, Langelaan M, Verheij TJM, Wagner C, Zwart DLM. Improving patient safety culture in
primary care: a systematic review. Journal of patient safety 2014;epub ahead of print.
6. Singh G, Singh R, Thomas EJ, Fish R, Kee R, McLean-Plunkett E, Wisniewski A, Okazaki S, Anderson
D. Measuring Safety Climate in Primary Care Offices. In: Henriksen K, Battles JB, Keyes MA, Grady
ML, editors. Advances in Patient Safety: New directions and alternative approaches (Vol 2: Culture
and Redesign) Rockville: Agency for Healthcare Research and Quality; 2008. p. 59-72.
9. De Wet C, Johnson P, Mash R, McConnachie A, Bowie P. Measuring perceptions of safety climate
CHAPTER 6
2011;20(2):121-7.
12. Shojania KG, Thomas EJ. Trends in adverse events over time: why are we not improving?. Qual
and safety in health care 2013;22:273-7.
16. Ginsburg L, Norton PG, Casebeer A, Lewis S. An educational intervention to enhance nurse
leaders’ perceptions of patient safety culture. Health Services Research 2005;40(4):997-1020.
106 CHAPTER 6
20. Mannion R, Konteh FH, Davies HT. Assessing organisational culture for quality and safety
improvement: a national survey of tools and tool use. Quality and safety in health care
2009;18(2):153-6.
21. Verbakel NJ, Langelaan M, Verheij TJM, Wagner C, Zwart DLM. Cluster randomized, controlled
trial on patient safety improvement in general practice: a study protocol. BMC family practice
2013;14:127.
practice 2011;12:117.
24. NHG-praktijkaccreditering®. Dutch College of General Practice Quality Accreditation. 2011.
25. Waring J. Beyond blame: cultural barriers to medical incident reporting. Social science and
medicne 2005;60(9):1927-35.
26. Zwart DLM. Incident reporting in general practice [dissertation]. Utrecht: University Medical
Center Utrecht; 2011.
27. Gandhi TK, Lee HL. Patient Safety beyond the Hospital. New England Journal of Medicine
2010;363(11):1001-3.
28. Eekhout I, de Vet HCW, Twisk JWR, Brand JPL, de Boer MR, Heymans MW. Missing data in a multi-
item instrument were best handled by multiple imputation at the item score level. Journal of
CHAPTER 6
Clinical Epidemiology 2013;.
30. Sorra J, Famolaro T, Dyer N, Nelso D, Khanna K. Hospital Survey on Patient Safety Culture:2009
Comparative Database Report. Rockville: Agency for Healthcare Research and Quality; 2009.
31. Ohrn A, Rutberg H, Nilsen P. Patient safety dialogue: evaluation of an intervention aimed at
achieving an improved patient safety culture. Journal of patient safety 2011;7(4):185-92.
32. Hoffmann B, Müller V, Rochon J, Gondan M, Müller B, Albay Z, Weppler K, Leifermann M,
Mießner C, Güthlin C, Parker D, Hofinger G, Gerlach FM. Effects of a team-based assessment and
intervention on patient safety culture in general practice: an open randomised controlled trial.
BMJ Quality and Safety 2013; August 16;23(1):35-46.
2000;34(1-3):215-57.
34. Weaver SJ, Dy SM, Rosen MA. Team-training in healthcare: a narrative synthesis of the literature.

BMJ Quality and Safety 2014;23:359-72.
35. Pettker CM, Thung SF, Raab CA, Donohue KP, Copel JA, Lockwood CJ, Funai EF. A comprehensive
obstetrics patient safety program improves safety climate and culture. American journal of
obstetrics and gynecology 2011;204:216,e1-6.
36. Rabol LI, Ostergaard D, Mogensen T. Outcomes of classroom-based team training interventions for
multiprofessional hospital staff. A systematic review. Quality and safety in health care 2010;19(6).
37. Jones KJ, Skinner AM, High R, Reiter-Palmon R. A theory-driven, longitudinal evaluation of the
impact of team training on safety culture in 24 hospitals. BMJ Quality and Safety 2013;22(5):394-
404.
38. Conry MC, Humphries N, Morgan K, McGowan Y, Montgomery A, Vedhara K, Panagopoulou E, Mc
Gee H. A 10 year (2000–2010) systematic review of interventions to improve quality of care in
hospitals. BMC health services research 2012;12(275).
39. Gaal S, Verstappen W, Wensing M. What do primary care physicians and researchers consider
the most important patient safety improvement strategies?. BMC Health services research
2011;11:102.
40. Kolb DA. Experiential learning: experience as the source of learning and development. New
Jersey: Prentice-Hall; 1984.
CHAPTER 6
108 CHAPTER 6
APPENDIX I
Quality and safety management questions

The following questions were included in the quality and safety form conducted at
baseline and follow-up.
Incident reporting
1. How many incidents from your practice are known from 2011/2012? (primary
outcome)
2. In which way became these known by you?
3. If your practice has a formal reporting system, since when was this used?
4a. How many of these incidents have you analysed?
4b. Which method was used?
5. How many of these incidents caused harm to patients?
6. Of these incidents, how many were, to your opinion, possible avoidable?
7. Did you proactively searched for incidents in your practice? (for example by file
studies, audits, reporting weeks)
8a. Were there improvement actions implemented in response to (reported) incidents?
8b. If yes, did these improvement actions lead to the desired results?
Complaints procedure
1. How many complaints were received the past year (both from employees
and patients)?
2. Does your practice have an internal coordinator for complaints?
CHAPTER 6
3. Is there a formal procedure for handling of complaints?
Team meetings
1. Was the subject patient safety on the agenda for planned team meetings
the past year (2011/2012)?
2. If yes, please specify dates on which patient safety was on the agenda.
3. Was patient safety during these team meetings actually discussed?
4. Have there been team meetings in 2011/2012 where patient safety was
not on the agenda but was discussed?
5. Could you describe in catchwords the content of the discussed subject? (or sent
minutes)
6a. Were action points/improvement plans formulated during these mee
tings? If yes, could you describe these in catchwords.

6b. If yes, were these action points/improvement plans actually implemented and
evaluated? Which were and which were not? If no, why not?
Training
1. Was the subject “patient safety” subject of training the past year?
2. Which training was this?
3. Was this training for the whole practice or individual?
4. Did you notice the learned being implemented in practice? If no, why
not?
Safety management
1. Does your practice have a patients safety management plan or other
wise described safety management policy?
2a. Is this practice safety plan deployed the last year?
2b. If not, why not/which subparts were not?
Quality management
1. Does your practice have a protocols book?
2. Do you participate regularly in pharmacotherapeutic consultations?
3. Do you have a procedure/method for controlling the content of the GP
emergency bag? (inclusive medication)
4. Does your practice have an introduction procedure for new employees?
5. Does your practice have an emergency telephone?
6. Have you ever conducted a patient safety satisfaction survey?
CHAPTER 6
7. Does your practice use “ZorgDomein” (This is a referring aid for

physicians)
8. Does your practice have a procedure to check repeat prescriptions?
9. When was the last time the equipment in the practice was calibrated?
10. Did you implement a quality improvement project last years?
110 CHAPTER 6
APPENDIX II
Distribution of incident reports, presence of reporting procedure and accreditation status at

baseline and follow-up
Formal Formal
reporting reporting Accreditation
Interven- # incidents # incidents procedure procedure Accreditation status
tion baseline follow-up baseline follow-up status baseline follow-up
control 2 2 yes yes no yes

control 2 0 no - no no
control 0 0 no no no no
control 0 0 no yes yes yes
control 0 0 no no yes yes
control 3 1 yes - no in progress
SCOPE 0 0 no no no in progress
SCOPE 3 4 yes yes no in progress
SCOPE 0 0 no - no in progress
SCOPE 3 10 yes yes yes yes
SCOPE 4 57 no yes no yes
CHAPTER 6
SCOPE 0 0 no yes yes yes
SCOPE 1 10 no no no no
SCOPE 3 0 no no no no
workshop 0 5 no yes no in progress
workshop 5 20 yes yes yes yes
workshop 0 20 no yes no no
workshop 36 53 yes yes no yes
workshop 4 2 no no no in progress

APPENDIX III
Self-reported number of incident forms (outcome question included in the SCOPE

questionnaire)
Baseline Follow-up
Intervention Frequency Percentage Frequency Percentage
none 40 83.3
1 to 2 3 6.3
Control* 3 to 5 4 8.3
6 to 10 - -
11 to 20 - -
>20 1 2.1
none 49 84.5 34 61.8
1 to 2 7 12.1 7 12.7
3 to 5 1 1.7 4 7.3
SCOPE
6 to 10 1 1.7 6 10.9
11 to 20 - - 4 7.3
>20 - - - -
none 56 77.8 21 35.0
1 to 2 9 12.5 11 18.3
SCOPE + 3 to 5 7 9.7 17 28.3
workshop 6 to 10 - - 9 15.0
11 to 20 - - 2 3.3
>20 - - - -
*This question was included in the SCOPE questionnaire. This questionnaire was deployed as an intervention and
therefore baseline data was not available for practices in the control group.
CHAPTER 6
112 CHAPTER 6
APPENDIX IV
Effect of interventions on SCOPE safety culture dimensions after one year of follow-up
Regression coefficient (95% p Regression coefficient (95% p
C.I.)# C.I.)*
1. Handover and teamwork
Intervention I 0.031 (-0.20-0.27) 0.796 0.072 (-0.15-0.30) 0.531

Intervention II 0.087 (-0.14-0.31) 0.449 0.123 (-0.09-0.34) 0.254
Accredited (y) -0.039 (-0.25-0.17) 0.716
Practice size -0.024 (-0.05-0.00) 0.069
2. Support and fellowship
Intervention I -0.216 (-0.52-0.09) 0.164 -0.224 (-0.54-0.09) 0.160

Intervention II 0.065 (-0.24-0.37) 0.674 0.076 (-0.23-0.38) 0.625
Accredited (y) -0.201 (-0.50-0.10) 0.193
Practice size -0.002 (-0.04-0.03) 0.089
3. Communication openness
Intervention I -0.170 (-0.46-0.35) 0.245 -0.144 (-0.43-0.14) 0.323

Intervention II 0.064 (-0.22-0.35) 0.660 0.102 (-0.18-0.38) 0.479
Accredited (y) -0.113 (-0.39-0.16) 0.425
Practice size -0.019 (-0.05-0.02) 0.279
4. Feedback about and learning from error
Intervention I -0.024 (-0.41-0.36) 0.902 0.030 (-0.32-0.38) 0.866

Intervention II 0.102 (-0.28-0.48) 0.602 0.177 (-0.16-0.52) 0.308
Accredited (y) -0.317 (-0.65-0.02) 0.064
CHAPTER 6
Practice size -0.028 (-0.07-0.01) 0.177
5. Intention to report events
Intervention I 0.051 (-0.35-0.45) 0.800 0.064 (-0.34-0.45) 0.746

Intervention II 0.148 (-0.24-0.54) 0.455 0.168 (-0.20-0.54) 0.368
Accredited (y) -0.327 (-0.66-0.00) 0.053
Practice size -0.015 (-0.06-0.03) 0.712
6. Adequate procedures and adequate staffing
Intervention I 0.075 (-0.26-0.41) 0.663 0.083 (-0.27-0.44) 0.459

Intervention II 0.123 (-0.21-0.46) 0.474 0.135 (-0.22-0.49) 0.459
Accredited (y) -0.002 (-0.35-0.35) 0.992
Practice size -0.001 (-0.05-0.04) 0.766

7. Overall perceptions of patient safety management
Intervention I 0.017 (-0.33-0.37) 0.922 0.024 (-0.34-0.39) 0.895

Intervention II 0.218 (-0.13-0.57) 0.211 0.237 (-0.12-0.60) 0.189
Accredited (y) -0.122 (-0.48-0.23) 0.486
Practice size -0.007 (-0.05-0.04) 0.736
8. Expectations and actions of managers
Intervention I 0.001 (-0.37-0.37) 0.997 0.015 (-0.38-0.41) 0.936

Intervention II 0.133 (-0.24-0.50) 0.468 0.156 (-0.24-0.55) 0.418
Accredited (y) -0.050 (-0.44-0.34) 0.791
Practice size -0.011 (-0.06-0.04) 0.633
PSG
Intervention I -0.035 (-0.48-0.42) 0.875 -0.008 (-0.44-0.42) 0.970

Intervention II 0.161 (-0.29-0.61) 0.463 0.224 (-0.20-0.65) 0.287
Accredited (y) -0.344 (-0.76-0.08) 0.103
Practice size -0.022 (-0.07-0.03) 0.386
# Univariate analyses
* Corrected for accreditation status and practice size
CHAPTER 6
114 CHAPTER 6
CHAPTER 6

CHAPTER 7
How does it work?
An interview study on culture interventions in
general practice
Verbakel NJ, de Bont A, Verheij TJM, Wagner C, Zwart DLM
Submitted
ABSTRACT
Background
When improving patient safety a positive safety culture is key. As little is known on improving
patient safety culture in primary care we examined whether administering a culture
questionnaire with or without a complementary workshop could be used as interventions
for improving safety culture.
Aim
To gain insight into how two interventions affected patient safety culture in everyday
practice.
Design and setting
An interview study nested in a cluster randomized trial was conducted in Dutch general
practice.
Method
Interviews were conducted at practice locations (n=27). We spoke with 24 GPs and 24
practice nurses. The theory of Communities of Practice — in particular its concepts of a
domain, a community and a practice — was used to interpret our findings by examining
which elements were or were not present in the participating practices.
Results
We found that communal awareness of the problem was only raised after getting together
and discussing patient safety. The combination of a questionnaire and workshop enhanced
interaction of team members and nourished team-feelings. Also, this shared experience
helped them to understand and develop tools and language for daily practice.
Conclusions
In order for patient safety culture to improve, the safety culture questionnaire accompanied
with a practice workshop was more successful. Initial discussion and negotiation of shared
goals during the workshop fuelled feelings of coherence and belonging to a community that
wishes to learn about enhancing patient safety. Team meetings and day to day interactions
enhanced further liaison and sharing, making patient safety a common and conscious goal.
CHAPTER 7
118 CHAPTER 7
INTRODUCTION
A constructive safety culture is important for patient safety improvement efforts.1 Safety
culture reflects the values, competencies and behaviour that determine the commitment
to, and the proficiency of an organization’s safety management.2
Patient safety is a prominent issue in primary care as well.3 In 2008 a national collaboration
project was launched aiming to engage Dutch primary care professions in patient safety.4
And we developed a safety culture questionnaire applicable for all primary care professions.5
Previous research indicated raised awareness and possible intervention effects of culture
surveys.6, 7 A questionnaire can be deployed as an intervention as results can be reflected
and acted upon.8 However, it is unlikely that a questionnaire alone leads to meaningful
improvements.9-13
We conducted a randomised trial studying two culture interventions: (1) administering
a safety culture questionnaire and (2) the questionnaire combined with a practice based
workshop, compared to a control group.14 We found that the combination of a questionnaire
with a workshop led to an increased number of reported incidents. In contrast, the stand-
alone questionnaire was significantly less effective.15 In this paper, we aim to explain these
differences in effect using a qualitative approach.
Theoretical framework
The theory of Communities of Practice (CoP) was used to interpret the interviews and explain
the differences in intervention effect. A CoP is described as a set of people who “share a
concern, a set of problems or a passion about a topic, and who deepen their knowledge and
expertise in this area by interacting on an ongoing basis”.16 Central to learning is exchanging
experiences and reflecting upon everyday practice. Since the concept was introduced in
1991 by Lave and Wenger17 and further elaborated on by Wenger in 199818, it is picked up
as a tool for quality improvement, problem-solving and innovation. Three dimensions need
to be present in order to be a CoP; a joint enterprise (the domain), mutual engagement (the
community) and a shared repertoire (the practice).18-21 The interest that members share
CHAPTER 7
defines the domain, in our study this was patient safety. By sharing information and engaging
in activities and discussions, members build relationships that enable them to learn from each
other, thereby establishing a community.19,22 This mutual engagement, refers to the level of
communication and interaction with each other. By interrelating, the members are motivated
to give meaning to and negotiate about their practices. Members of a CoP develop “a shared
repertoire of resources: experiences, stories, tools, ways of addressing recurring problems”.19
How does it work? 119

SCOPE Intervention Study, the trial and intervention components
Details of the SCOPE Intervention Study design and its findings have been reported
elsewhere.14, 15 The first intervention was administering a patient safety culture questionnaire,
the SCOPE questionnaire22, further referred to as SCOPE group (Figure 1). The contact person
in each practice could download the results of the survey.
Intervention I: A safety culture questionnaire

• The contact person received a letter informing a) the allocation of their practice in the
three research arms and b) the request to fill-out the SCOPE with their whole practice
• Details of login procedures were given directly in this letter
• The feedback report could be downloaded by the contact person
• A reminder to complete the questionnaire and download the report was sent after 1
week and 1 month
SCOPE Questionnaire
• 43 culture items (five-point scales), 2 outcome questions
• Demographic variables
• 15-20 min to complete
Feedback report
• Oversight of the eight dimensions and percentages positive scores
• Mean score of the practice per item compared to a benchmark score
• Tips for improvement and further information
• The contact person was responsible for disseminating the report
Figure 1. Description of intervention I
The second intervention consisted of the SCOPE questionnaire complemented with a

workshop, further referred to as workshop group (Figure 2). Instead of downloading their
results, these were presented during the workshop and used for discussion.
METHOD
CHAPTER 7
Design and participants

Interviews were conducted by NV and MvM between May and July 2013 in all participating
practices, one practice (control group) refused to do interviews. We spoke with 24
general practitioners and 24 practice nurses (Table 1). Because interviews in the control
group revealed no change in patient safety activities, we only described results from the
intervention practices (36 interviews). In addition to the interviews, NV observed during the
workshops and staff was asked to complete an evaluation form afterwards.
120 CHAPTER 7
Intervention II: A practice-based patient safety workshop
• At practice location
• ≥ 75% of staff was required to attend
• 3,5 hours
• SCOPE questionnaire was completed a few weeks before the workshop
Workshop elements
• Education on safety science/human factor engineering/culture (systems approach)
• Filling-out and discussing two MaPSaF vignettes23
• Presentation and discussing the SCOPE results
• Guided discussion on own culture and possible improvement
• Drawing of an action plan to improve patient safety (culture)
Figure 2. Description of intervention II
Data collection and analysis

To direct the interviews we used a topic list. This addressed opinions on the prevailing
patient safety and culture and actual implementation of tools following their particular
intervention. In the control practices we additionally asked whether their focus on patient
safety was influenced by governance, insurance agencies or otherwise. All interviews were
audiotaped with consent from the interviewee, transcribed verbatim and transcripts were
presented to the interviewees for approval. Analysis of data was performed using NVivo
software for qualitative data for coding.24 Prior to coding important themes were discussed
within the research team to develop a coding template which focussed on safety culture,
behaviour and activities attributable to the intervention.25 After initial coding, the CoP
theory was used to analyse and interpret our findings.
RESULTS
The interviews showed the necessity of a joint follow-up of the questionnaire. When explicitly
asked at the end of the interviews, all interviewees of the workshop group stated that their
results would not have been achieved had they only completed the questionnaire. Moreover,
CHAPTER 7
the SCOPE practices felt that results would have been different had they participated in the
workshop. Below we explained why the workshop was regarded as necessary to intervene
in the patient safety culture.

Table 1 Characteristics of participants per intervention group
SCOPE group Workshop group
(10 practices) (9 practices)

GPs 8 8
Practice nurses 10 10
Female gender % 77.8 66.7
GPs 50 37.5
Practice nurses 100 90
Age (m, sd) 43.4 (9.7) 40.7 (14.1)
GPs 46.0 (8.8) 48.6 (9.4)
Practice nurses 41.4 (10.3) 33.4 (13.9)

Hours per week in practice 33.6 (15.8) 34.2 (10.7)
GP 39.6 (19.9) 35.8 (14.7)
Practice nurses 27.6 (7.4) 32.9 (6.6)

Qualified since (years) 15.2 (9.6) 13.1 (10.2)
GP 14.3 (9.0) 16.7 (8.9)
Practice nurses 16.0 (10.5) 10.3 (11.0)

Working in practice (years) 8.9 (7.1) 9.7 (6.8)
GP 10.3 (9.1) 12.5 (7.0)
Practice nurses 7.8 (5.3) 7.5 (6.0)

Duration interview (minutes) 38.9 (16.8) 34.2 (11.4)
GP 42.2 (10.2) 38.6 (7.1)
Practice nurses 36.4 (20.7) 30.7 (13.2)
Joint enterprise
CHAPTER 7
In our study the problem and thereby the possible identity of the CoP concerned patient
safety. The workshop showed to contribute to the awareness for patient safety in two ways.
Firstly, getting together and spending time on the topic sends the message of the subject
being important. Secondly, the workshop changed their view on patient safety. Discussing
international and national data about iatrogenic harm amazed and sometimes even shocked
the caregivers, creating a sense of urgency. The interviewees expressed how the workshop
changed their perception of the problem.
122 CHAPTER 7
[Talking about what was done about patient safety after the workshop]
“Anyway, we now have all become very alert. […] in any case, we all had our minds on the
job after the workshop. […] You really were facing the facts.”
Practice nurse, workshop group(20)
One of the assignments during the workshop illustrated the process of reaching agreement
on their own culture and the gaps. When assigning maturity stages to the MaPSaF vignettes
we noticed that the individually chosen stages were almost always relatively high and during
the process of discussing in pairs and subsequently the whole team negotiation of the best
fit arose and the stages chosen became lower.
Contrary, in the SCOPE practices patient safety mostly was not perceived to be an urgent
problem. Interviewees often stated that no action was undertaken because patient safety
was seen as adequate. It seemed that no risk-awareness for safety problems was generated
by the questionnaire.
Interviewer: “Do you think that the questionnaire had an impact on your practice?”
Interviewee: “I don’t think so. Since things are already going well. “
Practice nurse, SCOPE group (8)
Interviewee: “Until now, not one complaint and not one incident. That is perhaps also the
reason that until now, we haven’t put anything on paper.”
Interviewer: “Ok, are there no incidents or aren’t they noticed?”
Interviewee: “Yes… that could be. So, it is not reported as such … maybe also because it is
not noticed.”
GP, SCOPE group (9)
Mutual engagement
The key element in a community is learning from each other and discussing experiences,
i.e. knowledge sharing. Workshop participants stated that the workshop was experienced
CHAPTER 7
very positive also nourishing the team-feeling and mutual trust. The workshop initiated
discussions about patient safety (activities) during the day and during team meetings, for
instance by asking each other to write an incident report.
Interviewee: “Yes. I think that, since we had the workshop, we all improved, or at least
things are set up. We already had a sort of reporting procedure. But due to the workshop,
there came a sort of awareness in the whole team. “
[…]

Interviewer: “And what would you see as your success factor, why did it work so well for
you?”
Interviewee: “I do think the organization, thus, the workshop we had. Getting aware of
reporting incidents and the explanation on that, what the pros and cons are so to speak,
to discuss that among each other. As a result, we have started a reporting week. We
started reporting more. We give feedback during work-meetings. I mean during general
meetings, the practice assistants meetings, the nurses meetings. Allowing discussing
it together, become aware and learn about it. So I think that are major steps we made,
making it successful.”
Practice nurse, workshop group (5)
Interviewer: “Suppose that you only filled out the questionnaire. Would the effect have
been similar?”
Interviewee: “No, because in a certain way you have to be shown the facts and be made
more aware of the problem. And that is certainly what happened in this intervention [the
workshop], it more did get to us. So, that also the fear for reporting, that culture and the
usefulness of reporting was more clear than if we had not done it [the workshop]. I think
that, if I had only had the SCOPE questionnaire for the employees… nothing would have
been achieved regarding the reporting week. So, in that respect it has a clear effect and
added value to for the whole team… and its progression and improvement.”
GP, workshop group (5)
In the SCOPE group such impetus for change in daily practice lacked. The feedback report
bearing the results and benchmark (580 practices from previous research26) known to be
an incentive for improvement27, was mostly not read or, if read by the contact person, only
shared once with colleagues.
“Interviewer: Last year you have had a report with feedback. So, than you get a summary
of the whole practice results compared to other practices in the Netherlands. Do you
CHAPTER 7
remember that you have received it [the report]. Or that you have seen it?”
Interviewee: [silence] “That doesn’t ring a bell.”
Interviewer: “Or discussed during the team-meeting?”
Interviewee: “No, oh no, that would.. no, than I would have [remembered]… No, I dare not
say.”
124 CHAPTER 7
Though, there were two interviewees from different SCOPE practices that stated that the
SCOPE spurred them to thinking about the topic. One practice grabbed the opportunity
to put patient safety on the agenda. In this case the strong and weak points of the results
were discussed within the team and this made clear that there was no incident reporting
procedure. They decided that a nurse would participate in a course about incident reporting
(outside the study as they were in the questionnaire only group) and implemented this in
their practice. Hereby, the interaction and learning from each other was clearly established,
as was the enthusiasm.
Interviewee: “It [SCOPE] has certainly given a boost, because discussing, openly, the things
that don’t go well… That is something that clearly comes from the SCOPE, and that you
emphasize that again. […] It is a guide to discuss things and further elaborate on, ok, how
are we going to improve this further?”
GP, Intervention I (21)
In the other practice the contact person read the report but did not share it within the
team. The subject remained the responsibility of this one nurse. She stated that her
awareness upon the topic was raised and that this was indirectly the case for the other staff
as she broached it. However, no mutual relations were established in the sense of team
interaction about patient safety.
Interviewer: “Do you think that the results of the questionnaire raised awareness? Did it
foster your reflective thinking?”
Interviewee: “Yes, it did . The rest [of the team] indirectly. Because I bring it up. Somebody
has to take the lead. And that is what we are missing here, also due to the situation, that
nobody takes it on. If I only put it on the desk of the practice assistants, nothing will be
done. I really have to bring it up and then maybe something will be done with it.”
CHAPTER 7
In addition the workshop was perceived as a shared experience. Participants experienced

the success of the workshop to a large extent to be the ‘communality’, a team moment to
focus on patient safety. According an interviewee, the team often referred in particular to
the workshop when patient safety was discussed.

Interviewer: “Ok, and how did that go? [the workshop]? Or do you discuss these things
more often together? [incidents].”
Interviewee: “Eh, well not really. We do not… Yes, you know, you all do the workshop
together and so it is easier to refer to it like ‘gosh, how was that again, do you remember?’
So than, yes, it makes it easier to come back to it when you did it together so to say.”
Unlike these workshop practices, the SCOPE practices expressed precisely the opposite,
missing this ‘getting-together’. It was stated by several interviewees that they missed the
attention that was given to the workshop practices claiming that would have made a great
difference. An ‘event’ was thought to be key in involving staff and to make the subject
tangible.
[Talking about the lack of change following the SCOPE, what would be needed]
“Well, maybe such a workshop it will make us all more involved. It would probably help.
Now it stays all a bit theoretical [having only the SCOPE results].”
GP, SCOPE group (2)
In some SCOPE practices the questionnaire even became perceived as an exercise to the
fulfilment of the research obligations. From the interviews we learned that a few practices
even thought of themselves as a non-intervention practice.
“I hoped I would not be in the control group, but in the intervention group that was to
work systematically [on the topic during the intervention], but we weren’t, unfortunate…”
GP, SCOPE group (9)
Shared repertoire
The workshop and subsequent interactions around the topic contributed to the alignment
of terminology. We started the workshop by asking participants what they thought common
CHAPTER 7
definitions meant, which showed differences between staff. By discussing these terms the
team made them their own and negotiated a mutual understanding of the terms.
Also, interviews showed that without an action plan, nothing happened. We asked the
workshop group if they thought the results would be the same when they would only
had completed the questionnaire. They claimed that the questionnaire solely would not
have done enough, they needed this joint meeting to convert the message into action. An
action plan conveys commitment to change and, more important, how to address it. This
126 CHAPTER 7
point was also demonstrated by two practices, one in each intervention group. Only one
practice in the workshop group did not realize an action plan and only one practice in the
SCOPE group was able to discuss the matter and draw joint activities. Interviews indicated
that only the practices that did commonly agree on activities showed improvements. The
workshop helped the conversation and enhanced ideas for improvement and subsequent
implementation. The interviews in the SCOPE practices showed the conversion of ideas and
results to activities to be a bottleneck.
Interviewer: Can you explain this, what was the reason for it? [talking about already having
a reporting procedure but only after the workshop there was more attention for reporting]
Interviewee: Well, just the importance of it I guess, that due to such a workshop... Yes, and
it is more in your system. So you can so to say put flesh on the bones.
[Talking about what you pick-up on the subject and what you want to do in practice]
“You are right that if you read something, if you read articles on the subject, it makes you
more aware. But implementing it in daily practice is something else and that is where it
often falters.”
GP, SCOPE group (29)
Lastly, the start-up of activities in turn, also helped to reinforce the actual repertoire and
community. Workshop practices started to implement or revive an incident reporting
procedure. As this is an ongoing or a repetitive tool it also strengthened the interaction
around the subject of patient safety. Some practices installed reporting committees, forms
were downloaded or created and reporting weeks were organised. Interviewees told
they pointed out to each other to write a report after an incident had happened. Reports
were discussed during the day and during team meetings. This invigorative process again
conveyed the message of importance and also created a learning effect. In other words, the
shared repertoire of terms and tools in itself added to establishing knowledge sharing and
interaction.
CHAPTER 7

DISCUSSION
Summary
We examined how administering a patient safety culture questionnaire solely or combined
with a workshop affected patient safety culture in general practice. Trial data showed that
incident reporting increased significantly in the workshop group, compared to the control
and SCOPE group. The latter showed some increase, but the effect was not significant

when discarding an outlier. We used the concept of Communities of Practice – and its
key concepts of joint enterprise, mutual engagement, shared repertoire – to explain the
differences between the two interventions.
Due to the workshop, risk-awareness arose. In addition, it contributed to team feelings
regarding patient safety and helped to align terminology and negotiate subsequent
activities. Patient safety was more often discussed and the atmosphere was more open to
discuss incidents. In the practices in the SCOPE group, contrary, the questionnaire did raise
some awareness, but s did not lead to actual changes. Almost none of our contact persons
in this group had read the report with their findings and only once they were discussed
with the remainder of the team. Our study, therefore, showed that a workshop is a valuable
addition to a questionnaire to improve patient safety culture.

This qualitative study has been conducted alongside an RCT. By studying observations
and perceptions of participants the trial results are supported and better understood. In
addition, we believe the intervention itself to be a strength. Though a workshop itself is
a one-time event our study showed that it can be used to set the focus on patient safety
and to set behavioural change in motion. Moreover, with one workshop the whole team
can be reached. The advantage of one team meeting, reaching all disciplines in one time,
enables the intervention to be unifying. This reduces the risk of ‘in-silo’ behaviour which
interferes with collaboration negatively.28 Limitations of this study could be the recall bias
of the interviewees. The interviews were held a year after the intervention and most of the
interviewees found it hard to remember details of the workshop or the action plan they
drafted.

Through the years, patient safety culture questionnaires were developed, modified and
validated in primary care, adding to an evidence-based means of assessing culture in
practice.29-34 A survey as change instrument is comprehensible, usable and affordable
CHAPTER 7
for practices. Though, Sexton et al. stated it to be unlikely that questionnaire results
and spontaneous discussion would lead to meaningful improvements and developed a
discussion tool.11 Analogous, an assessment and discussion tool, the Manchester Patient
Safety Framework, was found to be a meaningful instrument in primary care.35, 36 In addition,
there have been numerous improvement approaches based on team wise efforts such as
CRM, TeamSTEPPS and variable safety workshops that showed promising results.37-41 A
successful team that strives for a clear common goal and that regularly discusses how to
achieve this is seen as valuable in improvement programmes.42 In the current study we
128 CHAPTER 7
found that combining the assessment and the team wise approach has added value. This is
in line with reviews showing the multifaceted or multi component interventions to be most
successful in improving patient safety culture.6, 43
Implications for research and practice

Our findings exposed the interaction between targeting safety culture and implementing a
structure simultaneously, as these reinforced each other.44 A generative culture is needed to
raise risk-awareness. Also it is a prerequisite for conducting activities that require openness
and trust, such as reporting.45 In turn, incident reporting is a recurrent activity, which leads
to regularly discussing safety issues each time reports are analysed. These discussions
contribute to openness and trust.46,47 In our study, indeed, interviewees indicated that
discussions around reports helped them to remain aware of incidents and to address each
other around safety issues during daily practice. In this way, safety management was
embedded in everyday work. The structure with recurrent features provided improvement
of safety culture, and vice versa. A frequently heard, almost unanimous response during the
interviews in the SCOPE practices was that it certainly would have made a difference when
more serious attention, as was done with the workshop, was paid to the subject. For future
research, therefore, it would be interesting to conduct a likewise trial in a stepped wedge
design. Yet, we can confidently encourage general practitioners to invest time in a team wise
event in their practice to effectively put patient safety on the map in primary care.
CHAPTER 7

REFERENCES
2006.
3. Gandhi TK, Lee HL. Patient Safety beyond the Hospital. N Eng J Med 2010;363(11):1001-3.
4. ZorgvoorVeilig. Available at: www.zorgvoorveilig.nl. Accessed 5,1, 2014.
research 2013;13:354.
2011;20(2):121-7.
10. Grol R, Wensing M. Implementatie. Effectieve verbetering van de patiëntenzorg. [Implementation.
Effective improvements of patient care]. Utrecht: De Tijdstroom; 2011.
11. Sexton JB, Paine LA, Manfuso J, Holzmueller CG, Martinez EA, Moore D, Hunt DG, Pronovost PJ.
A check-up for safety culture in “my patient care area”. Joint Commission journal on quality and
patient safety 2007;33(11):699-703.
12. Saint S, Howell JD, Krein SL. Implementations science: How to jumpstart infection prevention.
Infect control and hospital epidemiology 2010;31(Suppl 1):S14-7.
13. Auerbach AD, Landefeld CS, Shojania KG. The tension between needing to improve care and
knowing how to do it. The New England Journal of Medicine 2007;357(6):608-13.
14. Verbakel NJ, Langelaan M, Verheij TJM, Wagner C, Zwart DLM. Cluster randomized, controlled
trial on patient safety improvement in general practice: a study protocol. BMC family practice
2013;14:127.
CHAPTER 7
15. Verbakel NJ, Langelaan M, Verheij TJM, Wagner C, Zwart DLM. A cluster randomised trial on the
effects of patient safety culture interventions in general practice. Submitted.
16. Wenger E, McDermott RA, Snyder WM. Cultivating communities of practice: A guide to managing
knowledge. Boston: Harvard Business School Press; 2002.
17. Lave J, Wenger E. Situated learning: Legitimate peripheral participation. Cambridge: Cambridge
University Press; 1991.
18. Wenger E. Communities of Practice: Learning, meaning and identity. Cambridge: Cambridge
University Press; 1998.
19. Etienne Wenger’s Introduction to Communities of Practice. Available at: http://wenger-trayner.
130 CHAPTER 7
com/theory/. Accessed 06/07, 2014.
20. Iverson JO, McPhee RD. Knowledge management in Communities of Practice: Being true to the
communicative character of knowledge. Management Communication Quarterly 2002;16(2):259-
66.
21. Iverson JO, McPhee RD. Communicating knowing throug communities of practice: Exploring
internal communicative processes and differences among CoPs. Journal of applied communication
research 2008;36(2):176-99.
practice 2011;12:117.
24. QSR International Pty Ltd. NVivo qualitative data analysis software 2012;10.
25. Available at: http://www.hud.ac.uk/hhs/research/template-analysis/. Accessed 5/14, 2014.
26. Zwart DLM, Langelaan M, Kuyvenhoven MM, Rensen vELJ, Kalkman CJ, Verheij TJM. Exploration
of patient safety culture perceptions in Dutch general practice. A cross sectional survey. Incident
reporting in general practice [dissertation] Utrecht: University Medical Center Utrecht; 2011. p.
47-61.
27. Dixon-Woods M, Bosk CL, Aveling EL, Goeschel CA, Pronovost PJ. Explaining Michigan: Developing
an ex post theory of a quality improvement program. The The Milbank quarterlyuarterly
2011;89(2):167-205.
28. Waring J, Currie G. The politics of learning: the dilemma for patient safety. In: Rowley E, Waring
J, editors. A socio-cultural perspective on patient safety Farnham: Ashgate Publishing Limited;
2011. p. 133-50.
29. Hoffmann B, Domanska OM, Albay Z, Mueller V, Guethlin C, Thomas EJ, Gerlach FM. The Frankfurt
Patient Safety Climate Questionnaire for General Practices (FraSiK): analysis of psychometric
properties. BMJ Quality and Safety 2011;20:797-805.
30. de Wet C, Johnson P, Mash R, McConnachie A, Bowie P. Measuring perceptions of safety climate
CHAPTER 7

BMJ Quality and Safety 2013; August 16;23(1):35-46.
37. Stead K, Kumar S, Schultz TJ, Tiver S, Pirone CJ, Adams RJ, Wareham CA. Teams communicating
through STEPPS. The Medical journal of Australia 2009;190(11 Suppl):S128-32.
38. Lisbon D, Allin D, Cleek C, Roop L, Brimacombe M, Downes C, Pingleton SK. Improved Knowledge,
Attitudes, and Behaviors After Implementation of TeamSTEPPS Training in an Academic Emergency
Department: A Pilot Report. American journal of medical quality 2014;pii.
40. Kemper PE, van Dyck C, Wagner C, Wouda L, de Bruijne M. Barriers and facilitators for taking
action after classroom-based crew resource management training at three ICUs. Joint Commission
journal on quality and patient safety 2014;40(7):311-8.
41. Thomas L, Galla C. Building a culture of safety through team training and engagement. BMJ
Quality and Safety 2013;22:425-34.
42. Grol R, Wensing M. Implementatie. Effectieve verbetering van de patiëntenzorg [Implementation.
Effective improvement of patient care]. Amsterdam: Reed Business,; 2011.
44. McCulloch P, Catchpole K. A three-dimensional model of error and safety in surgical health care
microsystems. Rationale, development and initial testing. BMC surgery 2011;11(23).
45. Snijders C, Kollen BJ, van Lingen RA, Fetter WPF, Molendijk H. Which aspects of safety culture
predict incident reporting behavior in neonatal intensive care units? A multilevel analysis. Critical
care medicine 2009;37(1):61-7.
46. Braithwaite J, Westbrook MT, Mallock NA, Travaglia JF, Iedema RA. Experiences of health
professionals who conducted root cause analyses after undergoing a safety improvement
programme. Quality and safety in health care 2006;15(6):393-9.
47. Iedema RA, Jorm C, Long D, Braithwaite J, Travaglia J, Westbrook M. Turning the medical gaze in
CHAPTER 7
upon itself: root cause analysis and the investigation of clinical error. Social science and medicine
2006;62(7):1605-15.
132 CHAPTER 7
CHAPTER 7

CHAPTER 8
General discussion
The aim of this thesis was to measure patient safety culture and to assess the effect of
safety culture interventions in primary care. Assessment of patient safety culture can be
used for positioning and tailoring (improvement) activities and evaluation. For this, a valid
and reliable tool was needed. We validated a generic questionnaire to assess patient safety
culture in primary care professions and subsequently examined the prevailing culture. In
addition to measurement, we intended to add knowledge on how to improve patient safety
culture in primary care. Therefore, a trial was conducted incorporating two interventions:
the administering of a patient safety culture questionnaire solely or combined with a
practice based workshop. Five research questions were formulated:
1. What is the validity of the patient safety culture questionnaire SCOPE-PC?

2. What is the prevailing patient safety culture in Dutch primary care proefessions and are
there differences between professional groups?
3. Which tools are available for patient safety culture improvement in the primary care
setting and what is their effectiveness?
4. What is the effect of two culture interventions in general practice?
5. How can the effect of the two culture interventions be explained?
The trial focussed on question 4 and 5 was conducted in the general practice setting, but
was made transferable to other primary care practices.
MAIN FINDINGS
The patient safety culture questionnaire – SCOPE-PC –

Validation of the SCOPE-PC questionnaire showed excellent internal consistency and good
construct validity. Seven dimensions were found: ‘open communication and learning from
error’ (8 items); ‘handover and teamwork’ (7 items); ‘adequate procedures and working
conditions’ (9 items); ‘patient safety management’ (5 items); ‘support and fellowship’ (5
items); ‘intention to report events’ (3 items); ‘organisational learning’ (3 items).
Prevailing culture in Dutch primary care

Exploring patient safety culture in nine types of Dutch primary care professions (dental care,
dietetics, exercise therapy, physiotherapy, occupational therapy, midwifery, anticoagulation
CHAPTER 8
clinics, skin therapy and speech therapy) showed that perceptions were positive, as were
their grades on patient safety in their practice. ‘Intention to report events’ scored lowest of
all dimensions. Moreover, this dimension showed the largest variation in culture within the
professional groups itself. Differences between professions were small.
136 CHAPTER 8
Tools affecting patient safety culture
A systematic literature review showed a lack of research regarding interventions affecting
patient safety culture in the area of primary care. Two studies were found. One study
described the implementation of an electronic medical record system and the other study
reported on the effect of two workshops. Both studies were performed in general practice.
Studying two culture interventions in general practice

The SCOPE Intervention Study was designed as a cluster randomized trial where practices
were allocated to either the control group or one of the two intervention groups: (I) the
administering of a patient safety culture questionnaire or (II) the same questionnaire
complemented with a practice based patient safety workshop. Practices in the first
intervention group (the questionnaire) were to download their results themselves whereas
practices in the second intervention group (workshop) were given their results during
the workshop. The workshop was designed to educate and facilitate discussion i.e. using
vignettes of the Manchester Patient Safety Framework (MaPSaF). A mixed-methods design
was chosen to both allow for quantitative effect measuring as well as in-depth qualitative
evaluation.
The effect of two culture interventions in general practice

Combining the patient safety culture questionnaire with a practice-based workshop showed
larger effect in terms of increased number of incident reports compared to the questionnaire
solely or the control practices. In addition, the workshop practices analysed incidents more
systematically and the topic was more often on the agenda of practice meetings. Although
these changes in behaviour concerning patient safety issues strongly suggested safety
culture changes in the workshop practices, no intervention effect was found on the safety
culture dimensions of the SCOPE questionnaire.
Understanding the implementation of two culture interventions

The interviews supported the quantitative findings of our trial study. Analysis showed that the
components of a ‘community of practice’ (CoP) regarding patient safety were more present
in the practices that received the workshop. These practices showed a shared understanding
of the patient safety concept, increased interaction on the topic and shared terminology
and tools. Contrary, all practices except one that had completed the questionnaire solely,
CHAPTER 8
showed no shared problem definition, minimal patient safety activities, scarce dialogue and
difficulty in applying tools.
General discussion 137

METHODOLOGICAL REFLECTIONS
The SCOPE questionnaire as measuring instrument

In the first part of this thesis the SCOPE questionnaire was used to measure patient safety
culture. Without getting entirely entangled in the clinimetric debate on the measurement of
culture a few words are in order. First, there is the issue whether it is possible to capture the
concept ‘culture’ with a questionnaire. It is known that a survey is not a perfect instrument
as it will only provide a snapshot, the surface of culture.1 Marshall et al. described this
problem as ‘the inability to capture the heart of culture’.2 They suggested to use participant
observation, interviews and focus groups combined with attitudinal surveys and established
cultural assessment tools. Still, the use of a questionnaire is one of the most convenient
and efficient ways to collect information on a subject. It can be done at relatively low cost
and it is easy deployable. Moreover, with regard to the delicate subject on hand it should
not be overseen that in contrast to more observational methods, a questionnaire can be
completed anonymously. In addition, a survey can be used to get the views of a relatively
large population at once, for example a hospital ward or, in primary care, a whole health
care centre.
Second, when used for evaluation purposes, features regarding responsiveness and
interpretability are important.3 Responsiveness is described as the ability to detect change
over time and interpretability refers to how well we can assign a qualitative judgement to
the quantitative results found. Measures that can be applied are the Minimal Detectable
Change, which shows the change larger than the measurement error, also called the ‘real
change’,4 and the Minimal Important Change (MIC), which is the amount of change that is
considered important or beneficial. These criteria are not yet established for patient safety
culture questionnaires. The current best alternative is to use a well validated questionnaire
and predefine meaningful changes for a study. Because culture questionnaires are frequently
used for monitoring or evaluation purposes, for the future, it is recommendable to study
and determine the characteristics of culture measurement more thoroughly. Awaiting these
further studies, we, in our study, used an adaptation of the well-developed and widely
used HSOPS questionnaire (Hospital Survey on Patient Safety Culture)5-9 and adhered to the
guideline of this questionnaire.10
In our trial we found no significant differences in the dimensions of patient safety culture
CHAPTER 8
survey between the three research groups at follow-up nor did we find any significant
improvements before and after the intervention in the intervention practices. Firstly, it could
be that there were indeed no actual differences or changes in culture in the participating
practices. Secondly, as it is not formally studied yet, it is possible that the SCOPE questionnaire
138 CHAPTER 8
is not sensitive enough to detect changes. Thirdly, it may also very well be that the time after
the intervention till the measurement (1 year) was too short to measure changes. Culture
changes generally are slow processes.11 A study that measured safety culture by a similar
questionnaire five years after an intervention programme did find improvements on their
measurement scale (HSOPS).12 Lastly, it may be that staff filled-out the SCOPE questionnaire
too optimistic in the first round. We found that risk-awareness was less present in practices
that did not participate in the workshops, which suggests that practices without education
and discussion are too optimistic about their own culture. Moreover, it is possible that staff
was more critical on their culture after the workshop. A similar phenomenon was found in
the study by Hoffmann et al. who conducted an intervention study also using the MaPSaF.13
Presumably, all these factors influenced our findings, however, it is most likely that the short
time frame in particular contributed to the a lack of effect found at the SCOPE questionnaire.
Qualitative study
In the second part of the thesis we consciously chose a mixed method approach, applying
qualitative methods during our trial. Qualitative research focuses on ‘why’ and ‘how’ allowing
for more profound understanding of the findings as it can discuss perspectives, experiences
and contexts.14 Despite these important contributions, qualitative research is not commonly
conducted as part of RCTs.15 We intended to explore factors that would hinder or contribute
to the success of changing safety culture in the practices and to explain found differences in
effectiveness between groups. Interestingly, where the culture questionnaire results failed
to show significant improvements on culture dimensions, the increase in incident reports as
well as data from the interviews suggested otherwise. These latter indicated that the culture
was more open after the workshops. Though most interviewees of all three research groups
stated that their culture was fine before the intervention, at follow up, interviewees that
had the workshop intervention stated that discussing patient safety as well as discussing
incidents was easier than before.
Nonetheless, when asking the professionals more in detail about how this worked in
daily practice, we encountered that many interviewees had difficulties remembering the
workshop in detail and more particularly the action plan they drew. This hindered in depth
exploration of the interviewees’ perceived changes. Apparently, these kind of qualitative
observations have to balance between the risk of recall bias and influencing trial results
when observations are performed during or closely after interventions.
CHAPTER 8

PATIENT SAFETY CUTLURE IN PRIMARY CARE: RISK-AWERENESS IS KEY
The validation study and subsequent examination of the prevailing culture in nine primary
care professions gave a first impression of patient safety culture in primary care. Patient
safety research in primary care is emerging but until now mostly focussing on general
practice and midwifery. Our exploration of safety culture in other professions incites some
considerations that may add to further directing practice and research. Examination of
the SCOPE dimensions showed that culture is perceived highly positive by the primary
care professions. At the same time, the SCOPE dimension ‘intention to report’ scored the
lowest compared to the other dimensions, and moreover, it showed the largest variations in
opinions of professionals within the profession groups itself.16
We believe three issues are relevant to elaborate on here: possible overestimation of the
level of the own safety culture, underestimation of possible risks and number of incidents
in the own profession and possible essential differences between primary care professions
regarding the relevance of patient safety thinking.
Firstly, the intervention study showed that the workshop contributed highly to a sense
of urgency. Participants stated that the numbers of adverse events presented in the
educational part of the workshop ‘opened their eyes’. In other words, their risk-awareness
enhanced. Although this study was conducted in general practice, it is likely that also in the
other professions the initial risk-awareness, i.e. without additional education, is low and
that safety culture is overestimated.
Second, examining the dimension ‘intention to report’ we found on the one hand, incident
reporting not being common yet. But on the other hand, incidents and reporting did seem
to play an important role in the concept of patient safety. Participants in our validation
study were enabled to give open comments at the end of the questionnaire. These remarks
indicated that having incidents or not was connected to how patient safety was perceived.
More specifically, it was often stated that because they perceived that no incidents
occurred, they felt that patient safety was fine in their practice and not an issue that had
to be improved; “I would surely discuss patient safety if there were incidents”. There is not
much research on the occurrence of incidents in primary care professions other than in
general practice. One study indeed reported only low percentages: 0.8% in dental care, 2,5%
in midwifery and 1.0% in paramedical practices based on 1000 patient medical records.17 So,
although incidents are perceived less likely, they do occur. Therefore, it is likely that primary
CHAPTER 8
care professions are not sufficiently aware of potential risks in their practice.
Thirdly, remarks were made about the relevance of the subject patient safety to their
practice or profession: “I do not see much risk with respect to patient safety” (speech
therapy), “My profession is not very risky” (employee anticoagulation care), which supports
140 CHAPTER 8
the assumption of lack of risk-awareness. Similar findings were reported in a study on the
meaning of patient safety in primary care, stating that professionals do care about patient
safety but they do not recognize any safety problems in the current approach of their work.18
Adding to this, there was a fairly low response rate in our study (38%) which could indicate
a lack of interest in the subject. Primary care consists of a broad array of professions and
though they share many similarities such as the educational level, structure of the practice
and patient population, with respect to medical orientation of the care provided there are
gradations. It is quite conceivable that professions with a less medical orientation, such as
speech therapy, dietetics, skin therapy and occupational therapy are less inclined to see
adverse events or other safety issues in their practice than their colleagues from midwifery,
dental care and general practice where incidents generally may cause actual harm. Hence,
primary care professions may deservedly have different levels of risk awareness.
Taking this all together, for future research and practice it is important to firstly search for
a dialogue with each of the professions. Above raised questions should be addressed to
investigate to which extent overestimation of the culture and underestimation of risks exist.
Single-handed practices should be included in this dialogue because these are a large part
of primary care. Also, professions evidently differ from each other concerning risk of patient
harm. These differences should be taken into consideration when addressing patient safety
(culture) in a particular profession. Still, educating professionals remains an important first
step as otherwise potential risks are not assessed. Quality circles, described as “small groups
of 6 to 12 professionals from a similar background who meet at regular intervals to discuss
and review their clinical practice”19 could provide a platform to discuss these risks and tailor
safety interventions.
A SAFETY CUTLURE QUESTIONNAIRE AS INTERVENTION?
A central issue in the intervention study was the question whether a safety questionnaire
can be deployed as a culture intervention to change perspective and eventually behaviour.
In fact, Dixon-Woods et al. in their evaluation of the Michigan study referred to Heisenberg’s
uncertainty principle that states that measurement cannot be performed without influencing
the system being measured. In case of this Michigan study, disclosure of data to other
IC’s also boosted action.20 Morello et al. in their review of safety culture interventions in
hospitals posited the same question.21 And, during the validation of the SCOPE questionnaire
CHAPTER 8
our research group also found that it raised awareness.22 Remarks during the SCOPE-PC
validation similarly showed that questionnaire items in theory could contribute to changes
in practice, for example:

• ‘We would have to make a protocol we think now and secure its inclusion in the weekly
meetings, good point. Thank you for this’ (occupational therapist)
• ‘Do you have an example of a reporting procedure? It cannot hurt to have one’
(occupational therapist)
• ‘We report to each other, not formally but during a meeting, though it is no fixed
feature. By this questionnaire I get the idea to do so. Thanks for that!’ (midwife)
Theoretically, a survey could be a suitable and effective instrument for change. Though,
measurement is only one thing, additional important steps are feeding back the results and
involvement of employees in planning and action. Five steps are distinguished in a well-
designed survey process:
1. Measuring and assessing;
2. Understand issues (what the survey is really telling us);
3. Prioritizing the issues most importantly;
4. Plan actions (who is accountable?);
5. Implement plans and follow-through.23
Culture change is a complicated process, among others because culture concerns the
identity of its members. By involving staff, awareness can be raised and through discussion
they will become more open to new ideas and become motivated to search for solutions.
As Wagner et al. stated: “The process of reporting the results is perhaps most important
in determining a survey’s effectiveness as a cultural change tool. It is the process that is
used to understand and act on the results which turns data into actionable information.”24
Pringle et al. found that using a questionnaire in patient safety improvement efforts
helped to adjust regional patient safety initiatives to specific needs. Moreover, 60% of
the participating hospitals used the survey to address their patient safety culture.25 In our
interview study we found that precisely the feedback and discussion part was lacking in the
SCOPE group, nullifying the effect of the questionnaire. When asked, interviewees stated
that they missed ‘getting-together’ and additional attention given to the subject. Contrary,
interviewees from the workshop group stated explicitly that the intervention would not
have been as successful without the workshop. From research on implementation of
evidence based practice or guideline adherence it is already known that it is laborious to
convert new knowledge and procedures into real practice.26-28 When these new ideas and
CHAPTER 8
procedures are only made available but no additional activity is conducted, it will not be
adopted by most professionals. In our study we found that when the above described steps
were addressed during a workshop where staff had room for discussion, it helped to convey
the message to the caregivers and subsequently led to behavioural change. During the
142 CHAPTER 8
workshop, besides education, results of the SCOPE were presented and discussed whether
they were recognizable (step 2). Together, the staff brainstormed on possible improvements
that could be made (step 3) and at the end of the workshop an action plan was made (step
4). Moreover, we found that after the workshop these practices implemented the activities
they chose for their action plan (step 5). The workshop helped these practices to digest their
results and convert their thoughts and ideas for improvement into concrete action.
CULTURE AND STRUCTURE
So far, there was no clear approach for improving patient safety culture in primary care.
Our study showed that a patient safety culture questionnaire combined with a patient
safety workshop enhanced the risk-awareness and spurred dialogue upon the subject.
Interestingly, all practices chose incident reporting as their activity to improve patient
safety in their practice. Incident reporting can be perceived as the most practical step
to start improving patient safety, as the goals are clear and the anticipated results are
understandable. In fact, it is conceivable that choosing this particular activity in their action
plan contributed to the success of the intervention. As incident reporting is not a one-time
event, but a cyclic process, it provides a structure for patient safety behaviour in everyday
practice. Analogous to any ‘plan, do, check, act’ cycle awareness is raised each time an
incident is addressed, also motivating staff when they perceive their reports were useful.
To understand error but also to improve safety, interventions should target system, culture
and technology simultaneously.29 Approaching culture and structure simultaneously has a
reinforcing effect on each other. By providing a structure, like incident reporting, culture
is positively affected. And vice versa, by enhancing a positive culture during discussions of
incidents, barriers can be broken down. From the perspective of diffusion of innovation,
incident reporting as a safety intervention fulfils the criteria proposed to be successfully
implemented (perceived benefit of the change, compatibility with the culture, manageable
complexity).30 It is plausible, as culture and structure reinforces each other, that a positive
culture will mature during the pursue of safety activities and implementing safety structures,
like incident reporting. Nevertheless, it will still be crucial to foster an open and positive
culture by discussing underlying values and building trust.31 Our trial showed that the team-
effort helped to emphasize both the importance of safety and having a positive culture.
Just talking is not sufficient. Both culture and structure need to be addressed for managing
CHAPTER 8
safety.

SAFETY MANAGEMENT SYSTEM
Is implementing a safety management system attainable to provide a solid structure for primary
care? In 2004, the president director of Shell in The Netherlands, wrote recommendations
to improve hospital safety.32 The first recommendation was to install certified safety
management systems. These systems were to constitute of three components: a risk
inventory, incident analysis and a management system to plan and monitor improvement
actions. A risk inventory calls for a proactive approach to walk through processes in order
to address potential safety risks adequately. Contrary to incident analysis, this inventory is
undertaken while an incident has not yet occurred. The idea is to predict where problems
could occur in order to optimise the safety of the process. Several tools are available to
support this process analysis. The Healthcare Failure Mode and Effect Analysis (HFMEA)
is probably the best known tool.33 This tool was translated in Dutch and adapted to SAFER
(Scenario Analyse van Faalwijzen, Effecten en Risico’s).34 In addition, the platform for patient
safety in primary care (ZorgvoorVeilig) has developed and provided a risk scan to recognize
and address patient safety issues in practice.35 For incident analysis two methods are mostly
applied in The Netherlands. The SIRE (Systematische Incident Reconstructie en Evaluatie)
method36 and the PRISMA (Prevention and Recovery Information System for Monitoring
and Analysis).37 Both components, the proactive inventory and the incident analysis, are
well suitable to be used by professionals in the primary care setting. The third component
requires a safety management system comparable to quality management systems which is
needed to manage the safety processes. Emphasis should be placed on both planning and
implementing as well as evaluating and making adjustments in order to let the system be
more than a bureaucratic thing. A generative safety culture overarching all components is
key condition for such safety management systems.32, 38
CERTIFICATION AND ACCREDITATION
Certification and accreditation may be helpful to achieve and maintain safety management
systems in primary care practices. It is available for all primary care professions and
incorporates safety management. General practice has a professional accreditation system
which requires a formal incident reporting procedure.39 Stichting HKZ (Foundation for
Harmonisation of Quality of Care ) provides two options for other primary care professions.
CHAPTER 8
Firstly, a patient safety management certification scheme that focuses explicitly on patient
safety management. This scheme includes all three mentioned components.40 Secondly,
there is a certification scheme that is designed specifically for small(er) organisations. This
scheme addresses a quality management system, and also includes above described safety
144 CHAPTER 8
components.41 Participants in our interview study indicated that patient safety management
was perceived as ‘something extra’ that had to be done. This was also found in other
professions.18 Certification or accreditation can well be used to secure safety elements
into daily policy and practice. Moreover, the recurrent aspect of these elements contribute
to the frequency of discussing the topic which enhances the liveliness and incorporates
a safety view by caregivers. In addition, a safety management system contributes to the
sustainability of safety improvements and allows for evaluation and learning.
EXTERNAL INCENTIVES?
As argued earlier, implementing innovations and involving professionals in this process

is very laborious. Is external pressure needed? External pressure could be used to oblige
practices to implement certain safety components, such as incident reporting, in their
practice. This could be enforced by certification or accreditation and subsequently by
demands of insurance companies. We believe that external pressure can certainly be very
helpful to get practices started. For some practices this is needed to cross the threshold and
just start. Once started they often experience the benefits themselves which may boost
motivation and finally contribute to a constructive safety culture. Moreover, regulation
enables the possibilities for compensations similar to the funding practices receive from
insurance companies when meeting targets of quality indicators of care.
Of course, regulating safety management in such a manner will only be helpful when it is
focused on the whole process. Having an incident reporting system is not enough, incidents
should be reported and analysed too. Otherwise, it will only create paperwork and a
burden for caregivers. Evidently, this is the drawback of external versus internal incentives.
When activities are only implemented because they are obligatory there will never be real
commitment. Particularly in matters of culture, commitment and a real team-effort are
needed. Lastly, safety management should not be confined to the own profession. Primary
care professions are increasingly collaborating in the care of patients, for example in disease
management. By building in safety components such as incident reporting, prospective
risk analysis and explicit focus on a constructive culture, professions that lag behind can be
towed along by other professions that are forerunners. Other studies used intermediate
meetings to allow for professionals to share thoughts and ideas. Moreover, it is known that
comparisons help practices to initiate change.20, 42 Integrated healthcare chains or other
CHAPTER 8
collaborative care structures offer great chances for a patient safety dialogue among primary
care professions. Hence, safety management initiatives within the current healthcare chains
of chronic diseases will create new opportunities to share lessons on quality and safety as
well as on multidisciplinary collaboration.

SHIFTING PERSPECTIVES AND THE IMPORTANCE OF CULTURE
Through incident reporting a practice can learn from their mistakes and anticipate by
adjusting procedures. The notion of proactive risk analysis is also based on the principle to
search for potential risks and incorporate barriers. These approaches assume a practically
linear relationship between risk management and safety. However, healthcare cannot be
fully standardized in procedures and policy, it consists of complex systems. Safety research
and management, therefore, should not only focus on incidents and risks but also on all
the things that go right. Mesman proposed “a research perspective that focuses on the
presence of safety and explores its texture”. She calls it ‘exnovation’, referring to “the
attempt to foreground what is already present”.43 It is stated that this will not only offer new
understanding of the vigour of healthcare, but also perspectives on caregivers’ competencies
and inventiveness. Based on similar motives, Hollnagel et al. described a necessary shift
from a ‘Safety One”- to a ”Safety Two”-perspective.44, 45 Safety One is “a state where as few
things as possible go wrong”, where incident analysis can be used to identify causes and
contributing factors, and risk assessment to determine their likelihood. The Safety One state
focuses on eliminating these causes and improving safety barriers, a so called ‘find and fix’
(reactive) approach. However, it assumes that systems are decomposable and that systems
function bimodal (successful or unsuccessful). But, healthcare does not always fulfil these
assumptions and another approach is needed (too). Instead, Hollnagel et al. describe Safety
Two as “as much things as possible go right”. It is about the system’s ability to succeed
under varying conditions as is applicable to healthcare. Safety management should be more
proactive, ‘we need to know how they go right’ and should invest in examining theoretical
foundations, underlying mechanisms and their manifestations.
Regarding safety culture, the resilient ( i.e. Safety Two) instead of reliable (i.e. Safety One)
approach calls for a high level of safety culture where adequate risk-awareness is crucial. The
Safety Two view naturally aligns with daily handling uncertainties as health professionals
do.46 In this view, good safety management leaves room for proactive handling degrading
situations. It assumes that health care is not suitable to be fully standardized, but in fact
it needs to be resilient. Moreover, sometimes it is even better to decide not to follow the
protocol. However, while primary caregivers generally are used to handle uncertainties,
they are not infallible nor always aware of risks in daily routine proceedings. Indeed, both
approaches should be considered as complementary. And a debate should be started on
CHAPTER 8
which part of daily primary care practice should be handled in a Safety One way and which
part needs a Safety Two approach. For balancing both approaches , safety management
should encourage a learning attitude, open communication, understanding of human
error and possible preventive measures and knowledge on potential risks in primary care
146 CHAPTER 8
practice. Hence, the shifting perspectives on safety a fortiori require a solid, constructive
safety culture. Evidently, health care leaders should keep in mind that risk-awareness is
not a matter of course and that full attention is needed in order to build a constructive
safety culture. The study described in this thesis showed that with relatively small effort the
necessary attention can be given and that this actually leads to safety culture improvement.
CHAPTER 8

REFERENCES
2000;34(1-3):215-57.
2. Marshall M, Parker D, Esmail A, Kirk S, Claridge T. Culture of safety. Quality and safety in health
care 2003;12:318.
3. COSMIN checklist. Available at: www.cosmin.com. 2014.
4. Beckerman H, Roebroeck ME, Lankhorst GJ, Becher JG, Bezemer PD, Verbeek AL. Smallest
real difference, a link between reproducibility and responsiveness. Quality of life research
2001;10(7):571-8.
5. Etchegaray JM, Thomas EJ. Comparing two safety culture surveys: safety attitudes questionnaire
and hospital survey on patient safety. BMJ Quality and Safety 2012; 21(6):490-8.
practice 2011;12:117.
research 2013;13:354.
10. Sorra J, Famolaro T, Dyer N, Nelso D, Khanna K. Hospital Survey on Patient Safety Culture:2009
Comparative Database Report. Rockville: Agency for Healthcare Research and Quality; 2009.
11. Guldenmund FW. (Mis)understanding safety culture and its relationship to safety management.
Risk analysis 2010;30(10):1466-80.
12. Noord Iv, Zwijnenberg N, Wagner C. Patiëntveiligheidscultuur in Nederlandse Ziekenhuizen. Een
stap in de goede richting [Patient safety culture in Dutch hospitals. A step in the right direction].
Utrecht: Nivel en EMGO+ Instituut; 2013.
BMJ Quality and Safety 2013; (1):35-46.
14. Kuper A, Reeves S, Levinson W. An introduction to reading and appraising qualitative research.
BMJ 2008;337:a288.
15. Lewin S, Glenton C, Oxman AD. Use of qualitative methods alongside randomised controlled trials
CHAPTER 8
of complex healthcare interventions: methodological study. BMJ 2009;339:b3496.

16. Verbakel NJ, van Melle M, Langelaan M, Verheij TJM, Wagner C, Zwart DLM. Exploring patient
safety culture in primary care. International journal for quality in health care 2014;26(6):585-91.
17. Harmsen M, Giesen P, Martijn L, Mettes T, Verstappen W, Nijhuis-van der Sanden R, Wensing M.
Patiëntveiligheid in de Nederlandse eerstelijnszorg anno 2009. [Patient safety in Dutch primary
148 CHAPTER 8
care anno 2009]. Nijmegen: Scientific Institute for Quality of Healthcare, UMC St Radboud; 2009.
18. de Bont A, Jerak S, de Mul M, Zwart DLM. Vragen voor veiligheid. De betekenis van patiëntveiligheid
in de eerste lijn [Questions for safety. The meaning of patient safety in primary care]. Rotterdam:
Institute of health policy & management. Erasmus University Rotterdam; 2009.
19. Rohrbasser A, Mickan S, Harris J. Exploring why quality circles work in primary health care: a
realist review protocol. Systematic Reviews 2013;2:110.
20. Dixon-Woods M, Bosk CL, Aveling EL, Goeschel CA, Pronovost PJ. Explaining Michigan: Developing
an ex post theory of a quality improvement program. The The Milbank quarterlyuarterly
2011;89(2):167-205.
2011;20(2):121-7.
23. Hinrichs JR. Feedback, action planning, and follow-through. In: Kraut AI, editor. Organizational
surveys: Tools for assessment and change San Francisco: Jossey-Bass; 1996. p. 255-79.
25. Pringle J, Weber RJ, Rice K, Kirisci L, Sirio C. Examination of how a survey can spur culture changes
using a quality improvement approach: A region-wide approach to determining a patient safety
culture. American Journal of Medical Quality 2009;24:374-84.
26. Powell CVE. How to implement change in clinical practice. Paediatric Respiratory Reviews
2003;4:340-6.
27. Grol R, Wensing M. What drives change? Barriers to and incentives for achieving evidence-based
practice. The Medical journal of Australia 2004;180(Supplement):57-60.
28. Wallace M, Shorten A. The challenge of implementing clinical guidelines. Evidence-Based
Healthcare & Public Health 2005;9:276-7.
29. McCulloch P, Catchpole K. A three-dimensional model of error and safety in surgical health care
microsystems. Rationale, development and initial testing. BMC Surgery 2011;11(23).
30. Greenhalgh T, Robert G, Macfarlene F, Bate P, Kyriakidou O. Diffusion of innovations in service
organizations: systematic review and recommendations. The The Milbank quarterlyuarterly
2004;82(4):581-629.
31. Firth-Cozens J. Organisational trust: the keystone to patient safety. Quality and safety in health
care 2004;13(1):56-61.
CHAPTER 8
you work safe, or you do not work here. Faster better - Safety in healthcare]. Den Haag: Shell
Nederland; 2007.
33. Habraken MM, van der Schaaf TW, Leistikow IP, Reijnders-Thijssen PM. Prospective risk analysis
of health care processes: a systematic evaluation of the use of HFMEA in Dutch health care.
Ergonomics 2009;52(7):809-19.

34. UMC Utrecht, TU Eindhoven, MAASTRO clinics. SAFER. Scenario Analyse van Faalwijzen, Effecten
en Risico’s. Instructie DVD en werkboek. Utrecht: 2006.
35. ZorgvoorVeilig. Risico scan. Available at: http://zorgvoorveilig.nl/risicoscan-basic. 2014.
36. Leistikow IP, den Ridder K, de Vries B. Patiëntveiligheid. Systematische Incident Reconstrucite en
Evaluatie [Patient safety. Systematic Incident Reconstruction and Evaluation]. 2nd ed. Maarssen:
Elsevier gezondheidszorg; 2009.
37. Van der Schaaf FW, Habraken MMP. PRISMA-Medical. A brief description. Eindhoven: TU
Eindhoven; 2005.
2006.
2011.
40. Certificatieschema Cliënt-/Patiëntveiligheid. Utrecht: Stichting HKZ; 2009.
41. Certificatieschema Kleine organisaties. Utrecht: Stichting HKZ; 2010.
42. Benning A, Dixon-Woods M, Nwulu U, Ghaleb M, Dawson J, Barber N, Franklin BD, Girling A,
Hemming K, Carmalt M, Rudge G, Naicker T, Kotecha A, Derrington MC, Lilford R. Multiple
component patient safety intervention in English hospitals: controlled evaluation of second
phase. BMJ 2011;342:d199.
43. Mesman J. Resources of strength: An exnovation of hidden competences to preserve patient
safety. In: Rowley E, Waring J, editors. A socio-cultural perspective on patient safety Farnham:
Ashgate Publishing Limited; 2011. p. 71-92.
44. Hollnagel E, Leonhardt J, Licua T, Shorrock S. From safety I to safety II: A white paper. European
Organisation for the Safety of Air Navigation (EUROCONTROL); 2013.
45. SafetySynthesis The future of safety management. Available at: http://safetysynthesis.com.
Accessed 10-10, 2014.
46. Jerak-Zuiderent S. Certain uncertainties: Modes of patient safety in healthcare. Social studies of
science 2012;42(5):732-52.
CHAPTER 8
150 CHAPTER 8
CHAPTER 8

English summary
Patient safety is key in health care. Though, as a subject for policy and research it came into
focus after the publication of the report ‘To Err is Human’. Patient safety is defined as: “The
(almost) absence of (the risk of) patient harm (physical/mental) that is caused by not acting
according to the professional standard of care providers and/or failure of the health system”.
An important element of patient safety is patient safety culture – shared values, attitudes
and perceptions regarding patient safety – or in short “the way we do things around here”.
Patient safety culture covers topics as: collaboration, (daring to ) speak up to each other,
reporting and learning from incidents and discussing patient safety in the practice. Research
showed that a positive, open safety culture contributes to or even is a precondition for the
success of safety interventions. Creating a constructive culture, therefore, is often the first
step for patient safety improvement.
In 2008 a platform, ‘ZorgvoorVeilig’ (Care for Safety) was launched to engage all primary care
professions in the development of policy on patient safety in the Netherlands. In this project,
the need emerged for tools that could be used to visualise and improve patient safety in
the whole primary care setting. At that moment, only a validated Dutch questionnaire was
available to measure patient safety culture in general practice. The first part of this thesis
describes the measurement and exploration of patient safety culture in Dutch primary care
(chapter 2 &3). Next, we performed a literature search for interventions affecting safety
culture conducted in primary care and developed our own intervention (chapter 4 & 5).
Subsequently, we tested these interventions in a randomised trial in 30 general practices
(chapter 6 & 7).
Chapter 2 describes the development and validation of a questionnaire to measure patient

safety culture in primary care, the SCOPE-PC. SCOPE is a Dutch acronym for systematic
culture inquiry on patient safety. Because the first (general practice) questionnaire was
named SCOPE, PC (Primary Care) was added to the questionnaire for the other primary care
professions. The culture questionnaire for general practice was used as a starting point. In
a written feedback round professionals of eleven professions – dentists, dental hygienists,
dietetic, exercise therapy, physiotherapy, occupational therapy, midwifery, anticoagulation
therapy, general practice, skin therapy, speech therapy – were asked to give feedback on the
adjusted questionnaire. Adjustments were made mainly in the terminology and wording.
The questionnaire was subsequently conducted in March and April 2011. A confirmative
factor analysis showed that the factor structure of the original SCOPE questionnaire did
not match well with the factor structure of the SCOPE-PC questionnaire. A subsequent
explorative factor analysis showed seven dimensions:
1. Open communication and learning from error;
154
2. Handover and teamwork;
3. Adequate procedures and working conditions;
4. Patient safety management;
5. Support and fellowship;
6. Intention to report events;
7. Organisational learning.
The reliability of the dimensions was satisfactory and the coherence was good.
Chapter 3 presents the state of affairs regarding patient safety culture in the different
primary care professions. Nine hundred and six individual questionnaires (519 practices)
were completed by professionals from nine different professions: dental care (dentists
and dental hygienists together), dietetic, exercise therapy, physiotherapy, midwifery,
anticoagulation care, skin therapy and speech therapy. Six hundred and twenty-five
questionnaires were eligible for analysis. Overall, safety culture was perceived positively.
The dimension ‘intention to report events’ and ‘patient safety management’ were scored
lowest. ‘Open communication and learning from error’ and ‘support and fellowship’ scored
highest. The dimension ‘intention to report events’ showed the largest variation within the
professions itself, possibly indicating the early developmental stage of incident reporting.
It could also be that the reporting of incidents in some professions is perceived as less
relevant. Differences between professions were however small.
Chapter 4 describes a review of patient safety interventions conducted in primary care that
affected safety culture. This showed a lack of published interventions in primary care. In
total, 214 articles were retrieved, but only two articles met the inclusion criteria: 1) the
research was conducted in primary care, 2) a patient safety intervention affecting culture
had to be described, and 3) the effect on patient safety culture had to be reported. The
first article described the implementation of an electronic medical record as part of on-
going quality and safety improvement efforts. Safety culture measurement showed
significant improvement. The second article described an intervention in which practices
were given the opportunity to participate in two workshops, a risk management workshop
and a workshop about incident analysis (Significant Event Audit). The authors reported an
overall improvement on risk management after participating in one of the workshops. Both
studies were conducted in general practice. The level of evidence of both studies was low,
both did not include a control group and there were other methodological problems. The
most important finding of the review was the lack of well-designed research of culture
interventions in primary care.
English summary 155

Chapter 5 is a detailed description of the study protocol. The aim of the SCOPE Intervention
Study was to test two culture interventions in general practice. To this end, we designed a
three-armed cluster randomised trial with ten general practices in each research arm. In
the control group no intervention was conducted. Practices in intervention I were asked
to fill-out the SCOPE questionnaire with the whole team. Contact persons of each practice
were able to download reports with their results and benchmark scores. The practices could
then decide what to do with their results. In the intervention II group practices were also
asked to complete the SCOPE questionnaire, however, these contact persons could not
download their report. Instead, these practices participated in a practice-based workshop.
The workshop consisted of two parts. The first concerned education about safety science, in
which the concept of patient safety and the size of the problem, terminology and causes of
human error (human factors engineering) were discussed. In the second part we facilitated
discussion about patient safety culture using the SCOPE results of their own practice and
parts of the Manchester Patient Safety Framework (MaPSaF). The MaPSaF is a matrix of nine
safety dimensions described in ascending maturity stages of safety culture. This instrument
was used to start-up a discussion about the practices’ safety culture. The workshop ended
with a brainstorm on possible improvements and the drawing of an action plan.
The primary outcome measure consisted of the number of reported incidents in the year
before the trial and one year after the intervention. Because incident reporting was still at
its infancy in general practice we expected an increase in reports. It was assumed that this
would indicate improved openness and communication and thus an improvement in safety
culture. Secondly, we assessed safety culture using the SCOPE questionnaire a year after
the intervention at all participating practices, including the control practices. In addition,
we collected data on quality- and safety indicators such as analysing incidents, having a
complaint procedure, discussing patient safety during team meetings and having a formal
safety policy.
Chapter 6 presents the results of the trial. After the baseline measurement of number
of reported incidents, the two interventions were deployed, and a year after follow-up
measurements were conducted. Statistical analysis of the number of incidents showed that
the practices in the workshop group reported 42 times more incidents than the control
group during follow-up when adjusted for the number of incidents known at baseline,
accreditation status and size of the practice. The SCOPE group reported 5 times more incidents
than the control group. Both results were statistically significant. Closer examination of the
results showed that one practice in the SCOPE group was an outlier. This practice chose to
participate in a workshop on incident reporting outside our study after reading their SCOPE
results. This practice reported 57 of the 82 incidents in the follow-up measurements in this
156
group. Repeating the analysis without this practice resulted in a non-significant result for
the SCOPE group compared to the control group. A few of the quality- and safety indicators
showed meaningful changes. Having a formal reporting system remained unchanged in
the control group, though, doubled in both intervention groups. In the workshop group
incidents were also more often analysed, more often there were orientation procedures for
new employees and patient safety was more often an agenda item of practice meetings.
Patient safety culture showed little improvement in the intervention practices, however,
statistical analysis showed that the three groups did not differ significantly at follow-up. The
assessment of patient safety (patient safety grade) showed a major improvement in the
practices that participated in the workshop.
Chapter 7 describes the findings of the qualitative study of the trial. Next to the quantitative
analysis we conducted interviews with the caregivers in the participating practices. Twenty
four general practitioners and 24 assistants and practice nurses from the three groups were
interviewed. The interviews from the control practices were in line with the quantitative
results indicating little to no change had taken place. Thirty six interviews from the
intervention practices were analysed to explain the found effect and differences in effect. We
used the theoretical framework of ‘Communities of Practice’ (CoP) to interpret the results.
A CoP is a group of people who share a purpose, problem or interest, and who deepen their
knowledge and expertise by an ongoing interaction. Three elements: a domain, a community
and a ‘practice’ (meaning ‘actual daily practice ’) have to be present. In this study, the
domain was patient safety. The interviews showed that more risk-awareness was present in
practices that participated in the workshop. In addition, there was more interaction on the
topic, incidents were discussed and analysed, in other terms: a community was formed. In
the SCOPE practices we found that a dialogue about the topic was established only scarcely.
Eight of the ten contact persons within the SCOPE practices had not read nor shared the
feedback report containing their results and the benchmark scores. During the interviews it
was stated that they missed the ‘getting together’ and the attentional awareness given to the
subject, in contrast with the workshop practices where the workshop was experienced as a
shared experience, a shared starting point. The workshop also seemed to have contributed
to the practice of patient safety in everyday practice – instruments, jargon, experiences – in
these practices.
Combining the measurement of culture and the mutual approach had a positive effect on
both behaviour and culture. With the qualitative analysis we showed that actual changes
occurred and that patient safety as a subject of activity and discussion came to life in the
teams.
English summary 157

Chapter 8 is an overall reflection on the thesis and describes recommendations for future
research and policy. We performed a first inventory of patient safety culture in Dutch
primary care. In this first measurement we found indications that caregivers overestimate
their own culture and underestimate risks. Therefore, it is important to search for dialogue
with professionals their needs and their opinion on relevance of patient safety interventions
in their setting.
The trial results with respect to the SCOPE questionnaire showed no significant changes in
time nor between the three groups at follow-up. It could be that there were no changes, that
the questionnaire was completed too optimistic in the first round, that the one-year period
was too short to change culture or it could be that the questionnaire was not sensitive
enough to measure the difference. Relevant characteristics of the questionnaire, such as
responsivity and interpretation of the results are important to examine further in the future.
Notable was that the interviews indicated that there was an improvement in patient safety
culture in the workshop practices in contrast to the other two groups.
The interviews further showed that education is an important first step when improving
patient safety and culture. Participants in the workshop practices had more risk-awareness
and this was among others ascribed to the theoretical part of the workshop where issues
such as incidents and how these develop were discussed. In addition to that, the workshop
seemed important to translate the SCOPE results to daily routine. Contrary, in the SCOPE
practices this important process of feedback and interpretation virtually was not established.
In sum, the trial showed that approaching patient safety as a team was most effective.
Patient safety and its maintenance initially focused on mistakes and trying to prevent these.
Instead, recently the focus shifted towards learning from the things that are going right.
Mainly in an environment such as health care, which cannot easily be standardised and
with a high degree of uncertainty, resiliency next to reliability is considered as an fruitful
approach. Particularly in such a resilient environment the professionals themselves are
the key players in sustaining patient safety where indeed, a generative culture, clear risk-
awareness and having a good judgement about when to intervene are crucial.
The study in this thesis showed that with relatively little effort, the necessary attention can
be given to patient safety culture and that this leads to positive changes.
158
English summary 159
Nederlandse samenvatting
Patiëntveiligheid is een kernwaarde van de gezondheidszorg. Toch is het als onderwerp voor
beleid en onderzoek pas echt onder de aandacht gekomen na de publicatie van het rapport
‘To Err is Human’. Patiëntveiligheid wordt gedefinieerd als “Het (nagenoeg) ontbreken van
(de kans op) aan de patiënt toegebrachte schade (lichamelijk/psychisch) die is ontstaan door
het niet volgens de professionele standaard handelen van hulpverleners en/of door tekort-
koming van het zorgsysteem”. Een belangrijk onderdeel van patiëntveiligheid is patiëntvei-
ligheidscultuur – de gedeelde waarden, attitudes en opvattingen over patiëntveiligheid – of
in het kort “de manier waarop we de dingen hier doen”. Patiëntveiligheidscultuur omvat
onderwerpen als; samenwerken, elkaar (durven) aanspreken, het melden en leren van in-
cidenten en het bespreken van patiëntveiligheid in de praktijk. Uit onderzoek is gebleken
dat een positieve, open veiligheidscultuur bijdraagt of zelfs een voorwaarde is voor het suc-
ces van patiëntveiligheidsinterventies. Het creëren van een constructieve cultuur is daarom
vaak de eerste stap bij het verbeteren van de patiëntveiligheid.
In 2008 is het platform ZorgvoorVeilig gelanceerd, om alle eerstelijns beroepsgroepen te

betrekken bij het ontwikkelen van beleid rond patiëntveiligheid in Nederland. Binnen dit
project ontstond de vraag naar middelen waarmee veiligheidscultuur in de brede eerstelijns
setting zichtbaar gemaakt kon worden en kon worden verbeterd. Er was op dat moment
alleen een gevalideerde Nederlandse vragenlijst beschikbaar voor het meten van patiënt-
veiligheidscultuur in huisartsenpraktijken. Het eerste deel van deze thesis beschrijft dan ook
het meten en exploreren van patiëntveiligheidscultuur in de Nederlandse eerste lijn (hoofd-
stuk 2 & 3). Daarnaast hebben we in de literatuur gezocht naar interventies in de eerste lijn
die effect hadden op de cultuur en hebben we in dit onderzoek een interventie ontwikkeld
(hoofdstuk 4 & 5). Vervolgens hebben we deze interventies in een gerandomiseerde trial
getoetst binnen 30 huisartspraktijken (hoofdstuk 6 & 7).
Hoofdstuk 2 beschrijft de ontwikkeling en validatie van een vragenlijst om patiëntveilig-

heidscultuur te meten in de eerste lijn, de SCOPE-PC. SCOPE is een afkorting voor Systema-
tisch Cultuur Onderzoek Patiëntveiligheid Eerstelijn. Omdat de eerste (huisartsen) vragen-
lijst SCOPE genoemd is, is PC (Primary Care) toegevoegd voor de vragenlijst voor de andere
eerstelijns beroepsgroepen. De cultuurvragenlijst voor huisartsen is als uitgangspunt geno-
men. In een schriftelijke feedbackronde zijn professionals van elf disciplines – tandartsen,
mondhygiënisten, diëtetiek, oefentherapie, fysiotherapie, ergotherapie, verloskunde, anti-
coagulantia, huisartsgeneeskunde, huidtherapie, logopedie – gevraagd om feedback te ge-
ven op de aangepaste vragenlijst. Aanpassingen werden met name gedaan in de gebruikte
terminologie en formulering. De vragenlijst is vervolgens van maart en april 2011 in deze
professies uitgezet. Een confirmatieve factoranalyse liet zien dat de factorstructuur van de
162
SCOPE-PC niet overeenkwam met de structuur zoals vastgesteld voor de SCOPE vragenlijst
voor huisartsen. Een exploratieve factoranalyse resulteerde vervolgens in zeven dimensies:
1. Communicatie over en leren van incidenten;
2. Overdracht en samenwerking;
3. Adequate procedures en werkomstandigheden;
4. Patiëntveiligheidsmanagement;
5. Steun en collegialiteit;
6. Meldingsbereidheid;
7. Lerende organisatie.
De betrouwbaarheid van de dimensies bleek bevredigend en ook de onderlinge samenhang
was goed.
Hoofdstuk 3 presenteert de stand van zaken van de veiligheidscultuur in de verschillende

eerstelijns beroepsgroepen. Negenhonderd zes individuele vragenlijsten (519 praktijken)
werden ingevuld door professionals vanuit negen verschillende professies: mondzorg (tand-
artsen en mondhygiënisten samen), diëtetiek, oefentherapie, fysiotherapie, verloskunde,
anticoagulantia, huidtherapie en logopedie. Zeshonderdvijfentwintig vragenlijsten waren
geschikt voor analyse. Over het algemeen werd cultuur positief gepercipieerd. De dimen-
sies ‘meldingsbereidheid’ en ‘patiëntveiligheidsmanagement’ werden het laagst gescoord.
‘Communicatie over en leren van incidenten’ en ‘steun en collegialiteit’ scoorden het hoogst.
De dimensie ‘meldingsbereidheid’ liet de meeste variatie binnen de professies zelf zien, wat
een indicatie zou kunnen zijn van het vroege ontwikkelingsstadium waarin het melden zich
bevond. Mogelijk wordt het melden van incidenten in sommige professies gepercipieerd
als minder relevant voor de dagelijkse praktijk. De verschillen onderling tussen de professies
waren niet groot.
Hoofdstuk 4 beschrijft de review van patiëntveiligheidsinterventies in de eerste lijn die een

effect hebben op de cultuur. De review liet een gebrek aan gepubliceerde interventies in de
eerste lijn zien. In totaal leverde de zoekstrategie 214 referenties op, echter, slechts twee
studies voldeden aan de inclusiecriteria: 1) het onderzoek moest uitgevoerd zijn in de eerste
lijn, 2) een patiëntveiligheidscultuur interventie met effect op cultuur moest beschreven
zijn, 3) het effect op patiëntveiligheidscultuur moest beschreven zijn. Het ene artikel be-
schreef de implementatie van een elektronisch medisch systeem als onderdeel van lopende
kwaliteits- en veiligheidsinitiatieven. Patiëntveiligheidscultuur liet een significante verbete-
ring zien. Het andere artikel beschreef een interventie waarbij praktijken de mogelijkheid
geboden werd deel te nemen aan twee workshops, een risico management workshop en
een workshop over incidenten analyse (Significant Event Audit). De auteurs rapporteerden
Nederlandse samenvatting 163

een algemene verbetering in risico management na deelname aan één van de interventies.
Beide onderzoeken zijn uitgevoerd in huisartspraktijken. De kwaliteit van beide studies was
laag, beide hadden geen controle groep en er waren andere methodologische problemen.
De belangrijkste bevinding van het review was een gebrek aan goed opgezet onderzoek naar
cultuurinterventies in de eerste lijn.
Hoofdstuk 5 is een gedetailleerde beschrijving van het studieprotocol. Het doel van de
SCOPE Interventie Studie was het testen van twee cultuurinterventies in de huisartsgenees-
kunde. Hiertoe hebben we een driearmige cluster gerandomiseerde trial opgezet met tien
huisartspraktijken in elke onderzoeksarm. In de controlegroep werd geen interventie uitge-
voerd. Praktijken in interventie I werden gevraagd om met alle medewerkers van de praktijk
de SCOPE patiëntveiligheidscultuurvragenlijst in te vullen. De contactpersoon van de prak-
tijk kon de feedbackrapportage met resultaten en benchmark scores downloaden. Vervol-
gens was het aan de praktijk om te besluiten wat te doen met de resultaten. In interventie
II werden praktijken ook gevraagd de SCOPE in te vullen, echter, voor deze contactpersonen
was het niet mogelijk de feedbackrapportage te downloaden. In plaats daarvan namen zij
met zoveel mogelijk medewerkers deel aan een praktijkgerichte workshop. De workshop
bestond uit twee onderdelen. Ten eerste, educatie over veiligheidskunde waarin het con-
cept van patiëntveiligheid en de grootte van het probleem, de terminologie en oorzaken van
menselijk falen (human factor engineering) behandeld werden. Ten tweede, een gefacili-
teerde discussie over de eigen veiligheidscultuur met behulp van de SCOPE resultaten uit de
eigen praktijk en onderdelen van het Manchester Patient Safety Framework (MaPSaF). De
MaPSaF is een matrix van negen veiligheidsdimensies uitgewerkt in vijf oplopend volwas-
senheidsstadia van veiligheidscultuur. Dit instrument werd gebruikt om de discussie over
de eigen veiligheidscultuur op gang te brengen. De workshop eindigde met een brainstorm
over mogelijke verbeteringen en het opstellen van een actieplan.
Als primaire uitkomstmaat is gekozen voor het aantal gerapporteerde incidenten in het jaar
voor de start van de trial en een jaar later. Omdat incidentmelden nog in de beginfase staat
binnen de huisartsenzorg verwachtten we een stijging van het aantal gemelde incidenten.
Dit werd gezien als een toename van openheid en communicatie en dus een verbetering
van de patiëntveiligheidscultuur. Secundair hebben we veiligheidscultuur gemeten met de
SCOPE vragenlijst een jaar na de interventie bij alle deelnemende praktijken, inclusief de
controle praktijken. Daarnaast hebben we data verzameld over een aantal kwaliteits- en
veiligheidsindicatoren zoals het analyseren van incidenten, het hebben van een klachten-
procedure, het bespreken van patiëntveiligheid tijdens teambijeenkomsten en het hebben
van een formeel veiligheidsbeleid.
164
Hoofdstuk 6 presenteert de resultaten van de trial. Na de baselinemeting van het aantal in-
cidentmeldingen per praktijk werden de twee interventies uitgevoerd waarna een jaar later
de follow-up metingen zijn gedaan. Statistische analyse van het aantal incidenten liet zien
dat praktijken in de workshop groep 42x meer incidenten meldden dan de controle groep
tijdens de follow-up, wanneer er gecorrigeerd werd voor het aantal incidenten bekend bij
de voormeting, accreditatie status en grootte van de praktijk. De SCOPE groep rapporteerde
5x meer incidenten dan de controle groep. Beide resultaten waren statistisch significant. Uit
nadere bestudering van de resultaten bleek dat één van de praktijken in de SCOPE groep
een outlier was. Deze praktijk besloot n.a.v. de SCOPE rapportage zelf een cursus over veilig
incidenten melden te volgen buiten onze studie. Deze praktijk rapporteerde 57 van de 82
incidenten in de nameting van de SCOPE groep. Het herhalen van de analyse zonder deze
praktijk resulteerde in een niet-significant resultaat voor de SCOPE groep ten opzichte van
de controle groep.
Van de kwaliteits- en veiligheidsmanagement indicatoren lieten enkele betekenisvolle ver-
anderingen zien. Het aantal praktijken met een formeel meldingssysteem bleef onveran-
derd voor de controle groep, echter verdubbelde in beide interventiegroepen. Daarbij wer-
den In de workshop groep de meldingen vaker systematisch geanalyseerd , waren er vaker
inwerkprogramma’s aanwezig en stond patiëntveiligheid vaker op de agenda van werkbe-
sprekingen.
Patiëntveiligheidscultuur liet een kleine verbetering zien in beide interventiegroepen, ech-
ter, statistische analyse liet geen significante verschillen zien tussen de drie groepen bij de
nameting. De beoordeling van de patiëntveiligheid liet een grote verbetering zien in de
praktijken die hadden deelgenomen aan de workshops.
Hoofdstuk 7 beschrijft de bevindingen van de kwalitatieve studie van de trial. Naast de

kwantitatieve analyse in het voorgaande hoofdstuk hebben we interviews gehouden met
zorgverleners in de deelnemende praktijken. Vierentwintig huisartsen en 24 assistenten en
praktijkondersteuners uit alle drie de groepen zijn geïnterviewd. De interviews uit de con-
trole praktijken waren in lijn met de kwantitatieve data die indiceerden dat er binnen deze
praktijken weinig tot geen veranderingen hadden plaatsgevonden. Zesendertig interviews
uit de interventiepraktijken zijn geanalyseerd om het gevonden effect en verschil in effect
te begrijpen. We hebben het theoretisch raamwerk van ‘Communities of Practice’ (CoP)
gekozen om de resultaten te interpreteren. Met de term CoP wordt een groep mensen aan-
geduid die een belang, een probleem of interesse delen en vervolgens hun kennis en exper-
tise verdiepen door een continue interactie met elkaar. Drie elementen: een domein, een
‘community’ en een ‘praktijk’ (met praktijk wordt hier bedoeld de wijze van praktijkvoering)
moeten aanwezig zijn. In deze studie was het domein patiëntveiligheid. De interviews lieten

zien dat er meer risico-bewustzijn aanwezig was in de praktijken die de workshop hadden
gevolgd. Daarnaast was er meer interactie rondom het onderwerp, incidenten werden be-
sproken en geanalyseerd, met andere woorden: er ontstond een community. In de SCOPE
praktijken vonden we dat slechts in beperkte mate dialoog was ontstaan over het onder-
werp. Acht van de tien contactpersonen binnen de SCOPE praktijken hadden het feedback-
rapport met de SCOPE resultaten van de eigenlijk praktijk en een benchmark niet gelezen
noch gedeeld met de andere zorgverleners. Tijdens het interview werd aangegeven dat ze
het ‘samenkomen’ en de bewuste aandacht voor het onderwerp hadden gemist, in tegen-
stelling tot de workshop praktijken waar de workshop werd gezien als een gedeelde erva-
ring, een gedeeld startpunt. De workshop bleek ook bij te hebben gedragen aan de concrete
uitvoering rond patiëntveiligheid in de dagelijkse gang– instrumenten, jargon, ervaringen
– binnen deze praktijken.
Het combineren van het meten van cultuur en de gemeenschappelijke aanpak heeft een
positief effect op zowel gedrag als cultuur. Met de kwalitatieve analyse hebben we laten zien
dat er daadwerkelijk verandering optreedt en dat patiëntveiligheid als onderwerp van actie
en gesprek meer gaat leven binnen de teams.
Hoofdstuk 8 is een algemene reflectie op het proefschrift en beschrijft aanbevelingen voor

toekomstig beleid en onderzoek. We hebben een eerste inventarisatie gemaakt van patiënt-
veiligheidscultuur in de Nederlandse eerste lijn. In deze eerste meting vonden we aanwijzin-
gen dat zorgverleners de eigen cultuur mogelijk overschatten en risico’s onderschatten. Het
is daarom van belang om de dialoog te zoeken met de zorgverleners: waar is nu behoefte
aan en welke aspecten van patiëntveiligheid zijn belangrijk voor iedere beroepsgroep afzon-
derlijk?
De trial resultaten met betrekking tot de SCOPE vragenlijst lieten geen significante veran-
deringen zien over de tijd of tussen de drie groepen in de nameting. Dit zou kunnen komen
omdat er geen verschil is, omdat de vragenlijst in de eerste ronde te positief is ingevuld,
omdat een tijdsperiode van 1 jaar te kort is om cultuur te veranderen of omdat de vragen-
lijst niet sensitief genoeg is om het verschil te meten. Relevante karakteristieken van de
vragenlijst zoals responsiviteit en interpretatie van de resultaten zijn belangrijk om in de
toekomst verder te onderzoeken. Opvallend was dat de interviews suggereerden dat er wel
een verbetering in de veiligheidscultuur was binnen de workshop praktijken in tegenstelling
tot de andere twee groepen.
De interviews lieten verder zien dat educatie een belangrijke eerste stap is bij het werken
aan patiëntveiligheid en cultuur. Participanten die de workshop hadden gevolgd hadden
daarna meer risico-bewustzijn, dit werd onder andere toegeschreven aan het informatieve
deel van de workshop waar zaken besproken werden als het aantal incidenten en hoe deze
166
ontstaan. Daarnaast bleek dat de workshop een belangrijke rol heeft gespeeld in het verta-
len van de SCOPE resultaten naar de praktijkvloer. In tegenstelling, in de SCOPE praktijken
kwam dit belangrijke proces van feedback en interpretatie vrijwel niet tot stand. Samen-
vattend heeft de trial laten zien dat het benaderen van patiëntveiligheid als een team het
meest effectief was.
Patiëntveiligheid en de waarborging ervan was voornamelijk gericht op fouten en het voor-
komen daarvan. Echter, recentelijk is de focus verschoven naar het kunnen leren van dingen
die goed gaan. Vooral in een omgeving zoals de gezondheidszorg, die niet te standaardise-
ren is en waar een hoge mate van onzekerheid bestaat, wordt veerkracht (resilience) naast
betrouwbaarheid (reliability) gezien als een vruchtbare aanpak. Juist in zo’n ‘veerkrachtige
omgeving’ zijn de professionals zelf sleutel figuren in het behouden van patiëntveiligheid,
waar een generatieve cultuur, een duidelijk risico-bewustzijn en het hebben van een goed
inschattingsvermogen cruciaal zijn.
De studie in deze thesis heeft laten zien dat met een relatief kleine inspanning, de nodige
aandacht gegeven wordt aan patiëntveiligheid en dat dit positieve verandering leidt.

Dankwoord
Dit proefschrift was niet tot stand gekomen zonder de hulp van vele anderen.
Allereerst wil ik graag mijn promotoren en copromotoren bedanken. Dorien, ik kan

wel zeggen dat zonder jou dit proefschrift er nu niet zou liggen. Het was fijn om jou als
copromotor te hebben. Je bent zo enthousiast over patiëntveiligheid en jou kennis over de
‘insides’ van de huisartspraktijk was onmisbaar. Ik weet niet hoe vaak je hebt gezegd: ‘Het
komt goed’ :) en gelukkig is het ook goed gekomen.
Maaike, jou ervaring met het patiëntveiligheidsonderzoek gaf mij vaak nieuwe inzichten en
nieuwe artikelen om te lezen. Bedankt voor al je praktische tips en je input voor de artikelen.
Cordula en Theo bedankt voor het meedenken over de lijn en inhoud van de studies. Theo,
jou kant vanuit de huisartsenpraktijk was erg informatief en praktisch bruikbaar in de trial.
Bedankt ook voor de kritische blik op mijn Engelse teksten. Cordula, jou ervaring met en
kennis van patiëntveiligheid was zeer bruikbaar. Het was fijn om met jullie samen te werken.
Ik wil graag de leescommissie bedanken voor het beoordelen van mijn proefschrift; Dinny de
Baker, Roger Damoiseaux, Job Metsemaker, Kit Roes en Marleen Smits.
Natuurlijk ook iedereen die heeft deelgenomen aan onze studies. Alle respondenten die
aan de start van het traject de SCOPE-PC vragenlijst hebben ingevuld, hartelijk bedankt voor
jullie tijd en moeite! Ook wil ik graag de praktijken bedanken die hebben deelgenomen
aan de trial studie. Fijn dat jullie tijd wilden investeren om vragenlijsten in te vullen en de
workshops te doen. Ik vond het ontzettend leuk en leerzaam om de workshops te doen en
met jullie te discussiëren over patiëntveiligheid en cultuur.
En dan moest natuurlijk alle data ook geannalyseerd worden. Peter Zuithoff, bedankt voor
het uitleggen van de analyses en als ik de draad weer kwijt was om alles nóg een keer door
te nemen, inclusief het review commentaar ;). Naast alle statistiek kwam er ook nog een
kwalitatief artikel om de hoek. Antoinette de Bont, bedankt dat je ons wilde inwijden in de
ins- en outs van het kwalitatieve onderzoek en voor het meedenken en meeschrijven van
het artikel.
Mijn paranimfen, Marja en Anneke, bedankt dat jullie mijn paranimfen willen zijn! Marja,
ik ben blij jou als vriendin te hebben. Het is fijn dat we altijd alles kunnen bespreken, zowel
gezellige dingen en serieuze dingen. Ik hoop dat we dat nog lang blijven doen! Anneke,
bedankt voor al je steun en motivatie tijdens mijn promotie en de master. Het was niet altijd
makkelijk, maar we hebben het gered! We waren lang kamergenoten en ook daarna was het
fijn dat we elkaar konden steunen bij de laatste loodjes.
170
Tessa, Loan en Marije, het was gezellig jullie als collega’s en kamergenoten te hebben.
Bedankt voor alle gezelligheid, het meedenken, het vieren van de verjaardagen en de
etentjes.
Ankie, bij HKZ was je al een fijne collega om mijn hart te luchten. Ik vind het leuk dat we ook
na die tijd contact hebben gehouden en ik hoop dat we dat in de toekomst blijven doen.
Ingeborg, tijdens onze master Sociologie deden we de meeste opdrachten samen. Ik vond
het altijd knap van je hoe gestructureerd jij kan werken. Tijdens alle uren dat ik naar mijn
beeldscherm staarde heb ik nog wel eens aan jou gedacht en wilde ik dat ik ook wat minder
uitstelgedrag had! Corinne, we hebben elkaar leren kennen tijdens vakantiewerk in het
zorgcentrum. Omdat jij in Groningen studeerde (en daar een kamer had) en ik in Groningen
ging studeren leek het ons handig contact te houden... wie weet kon het nog eens handig
zijn. Gelukkig maar dat we toen onze mailadressen hebben uitgewisseld want daarna is er
een leuke vriendschap tot stand gekomen. De afstand maakt het lastig, maar ik hoop dat we
nog lang vriendinnen blijven. Ik wil jullie bedanken voor jullie steun op afstand, ook via de
mail (en Whatsapp) was het erg fijn als jullie informeerden naar mijn voortgang en ik mijn
frustraties kon uiten.
Maria Fraters, bedankt voor het mooie ontwerp en het schilderij voor de voorkant. Fijn dat
je steeds een nieuwe versie wilde maken... andere ‘poppetjes’... toch weer andere kleuren…
ander materiaal…maar hij is mooi geworden! Leontine bedankt voor de bewerkingen (en je
geduld) van het schilderij zodat hij mooi op de voorkant zou komen.
Ben verbakel, bedankt voor de gezelligheid tussendoor, je leuke gitaarfilmpjes en het helpen
met onze verhuizingen.
Jacintha en Hans, ik wil jullie bedanken voor alle steun tijdens het schrijven van de artikelen
maar ook daarbuiten. Fijn dat jullie er altijd voor mij en Steven zijn! Bedankt voor alle keren
dat jullie geholpen hebben met verhuizen en alle keren dat ik bij jullie kon logeren, het
lekkere eten, de gezelligheid en natuurlijk niet te vergeten de houtkachel :) niets is zo leuk
als (gecontroleerd) met vuur spelen. Fijn om een plek te hebben om uit te rusten en te
relaxen.
Steven, mijn rots in de branding. Ik ben blij dat je zoveel geduld hebt met mij en zoveel
begrip kan opbrengen. Bedankt voor alles, samen staan we sterk!
Dankwoord 171
Curriculum Vitae
174
Natasha Verbakel was born in Exeter, Great Brittain, on the 18th of April 1984 and was raised
in the Netherlands.
In 2005 she obtained her Bachelor Degree in Nursing after which she studied Sociology
(Policy & Consultancy) from 2005 to 2007 at the University of Groningen. After graduation
she worked three years as a policy worker at the Foundation for Harmonisation of Quality
in Care (Stichting HKZ). Here, she worked on the development and maintanance of certifi-
cation schemes for quality management and later safety management systems in several
professions.
Late 2010 she started working at the Julius Center, University Medical Center Utrecht on the
research described in this thesis. In May 2014 she obtained her Masters of Science degree
in Clinical Epidemiology at the University of Utrecht.
Curriculum Vitae 175

Patient Safety Culture in Primary Care: Natasha Verbakel

Uploaded by

Copyright:

Available Formats

Patient Safety Culture in Primary Care: Natasha Verbakel

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Patient Safety Culture in Primary Care: Natasha Verbakel

Uploaded by

Copyright:

Available Formats

Patient safety culture in

ISBN: 978 - 94 - 6203 - 800 - 4

Patiëntveiligheidscultuur in de eerste lijn

ter verkrijging van de graad van doctor aan de Universiteit Utrecht

Natasha Johanna Verbakel

geboren op 18 april 1984 te Exeter, Verenigd Koninkrijk

Copromotoren: Dr. D.L.M. Zwart

Chapter 1 General introduction 7

PART I Patient safety culture in primary care 21

Chapter 2 Measuring safety culture in Dutch primary care: psychometric 23

Chapter 3 Exploring patient safety culture in primary care 43

PART II Patient safety culture interventions in general practice 59

Chapter 4 Improving patient safety culture in primary care: a systematic 61

Chapter 5 Cluster randomized, controlled trial on patient safety improvement in 77

Chapter 6 A cluster randomized trial on the effects of patient safety culture 93

Chapter 8 General discussion 135

Curriculum Vitae 173

(GP; argumentation not to participate in our study)

PRIMARY CARE AND PATIENT SAFETY

PATIENT SAFETY CULTURE

ASSESSING AND IMPROVING PATIENT SAFETY CULTURE

Figure 1. Culture ladder42

1. What is the validity of the patient safety culture questionnaire SCOPE-PC?

Patient safety culture

Assessing Improving Experimenting

Chapter 2 Validating a Chapter 4 Searching Chapter 6 Studying the

Chapter 3 Exploring the Chapter 7 Examining

Patient safety culture in primary care

Measuring safety culture in Dutch primary care:

Verbakel NJ, Zwart DLM, Langelaan M, Verheij TJM, Wagner C.

First adjustment to the questionnaire

Psychometric characteristics of the SCOPE-PC questionnaire 25

8. Expectations and actions of managers (4 items).

Psychometric characteristics of the SCOPE-PC questionnaire 27

values >0.1 are regarded as not acceptable.20,21

rate Mean (SD)

Occupational therapy practices 34 17% 41 (8.5) 93.5

Speech therapy practices 67 33.5% 37 (10) 100

Confirmatory factor analysis

Psychometric characteristics of the SCOPE-PC questionnaire 29

score and standard deviation (SD) per dimension.

Reliability and construct validity

1. Open communication and learning from error

Psychometric characteristics of the SCOPE-PC questionnaire

A1 People support one another in this practice 4.56 0.62 0.90

A2 We have enough staff to handle the workload 3.93 0.94 0.60

Psychometric characteristics of the SCOPE-PC questionnaire

A9 Mistakes have led to positive changes here 3.97 0.68 0.57

Dimensions n (SD) safety 1 2 3 4 5 6

and learning from error (0.64)

and working conditions (0.54)

become stronger as safety management activities become common practice.

Strengths and limitations

Conclusions and implications

Psychometric characteristics of the SCOPE-PC questionnaire 37

Psychometric characteristics of the SCOPE-PC questionnaire 39

Square Error of Approximation (RMSEA).

Psychometric characteristics of the SCOPE-PC questionnaire 41

Exploring patient safety culture in primary care

Anticoagulation therapy -0.25* -0.57 -0.35 -0.23 -0.27** 0.10 0.08

((health care evaluation mechanisms[mesh]) OR (intervention[tiab]) OR (improv[tiab]) OR