International Journal of Nursing Studies 50 (2013) 552–568

Contents lists available at SciVerse ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

Do implementation strategies increase adherence to pain assessment

in hospitals? A systematic review
Erwin Ista a,*, Monique van Dijk a, Theo van Achterberg b
a
Intensive Care Unit, Department Paediatric Surgery, Erasmus MC – Sophia Children’s Hospital, Rotterdam, The Netherlands
b
Scientific Institute for Quality of Healthcare, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

A R T I C L E I N F O A B S T R A C T

Article history: Objectives: Pain assessment and reassessment is an essential part of the treatment of
Received 30 March 2012 hospitalised patients and must be integrated in pain management protocols. Yet nurses’
Received in revised form 7 November 2012 adherence to pain assessment recommendations is problematic. We sought to review the
Accepted 7 November 2012 comparative evidence for implementation strategies aiming to improve nurses’ adherence
to pain assessment recommendations in hospitalised patients.
Keywords: Design: Systematic review using the narrative method.
Implementation Data sources: PubMed (MEDLINE), CINAHL, Cochrane library and hand searching.
Pain assessment
Review methods: Studies published since 1990, reporting implementation strategies that
Adherence
aimed to improve nurses’ adherence to pain assessment recommendations in hospitalised
Education
Feedback
patients were included. According to the Cochrane Effective Practice and Organization of
Implementation strategies Care group (EPOC) classification system, strategies were categorized as directed at: health
professionals, organizations, financing, or regulations. Given the heterogeneity in
strategies, samples, outcomes and settings, evidence from the studies was synthesized
using a narrative approach.
Results: From 743 initial citations, 23 studies were included. They reported a variety of
implementation strategies, but only directed at health professionals and/or organizations.
In seven studies, a single strategy was applied (e.g. education or feedback). The remaining
16 studies used multifaceted approaches. The effectiveness of the implementation
strategies varied. In all studies but one, adherence rates had improved after
implementation compared to the before measurement, by 9% up to 49%. These effects
were measured at different time points after completion of the implementation, ranging
from 2 weeks to 6 months. Half of the reviewed studies reported an adherence rate of 80%
or higher after implementation activities; other reported rates ranging from 24 to 80%. In
two controlled studies the adherence to pain assessment recommendations increased
significantly when feedback was provided compared to no feedback. Sustained effects
were reported in three studies.
Conclusions: Based on this systematic review we conclude that implementation strategies
to improve nurses’ adherence to pain assessment recommendations vary but generally
address professionals and organizational aspects. Educational and feedback strategies are
often used and seem largely effective.
Due to the heterogeneity of the implementation strategies it is not possible to
recommend one preferred strategy. The level of evidence for strategies to improve pain
assessment recommendations is limited however, as well-conducted studies are lacking.
ß 2012 Elsevier Ltd. All rights reserved.

* Corresponding author at: Intensive Care Unit, Erasmus MC – Sophia Children’s Hospital, Office Sh-4010, P.O. Box 2060, 3000 CB Rotterdam,
The Netherlands. Tel.: +31 10 7037028; fax: +31 10 7036796.
E-mail address: [email protected] (E. Ista).

0020-7489/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijnurstu.2012.11.003
 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568 553

What is already known about the topic? hospitalised patients in the USA (Joint Commission on
Accreditation of Healthcare Organizations, 2001). As a
 Repeated pain assessment is essential to effective pain result, pain assessment or management quality improve-
management. ment initiatives and studies have been widely set up. This
 Adherence to guidelines or protocols for the assessment resulted in an increased presence of structural elements
of pain in hospitalised patients remains problematic. (e.g. interdisciplinary work group, standards for assess-
ment and documentation, policies or procedures for
What this paper adds
treatment) that are critical to improving pain manage-
ment in practice (e.g. increased use of pain rating scales,
 Most strategies to promote pain assessment are multi-
decreased use of intramuscular opioids and increase use of
faceted and target determinants at the levels of
non-pharmacological strategies) (Dahl et al., 2003;
professionals and organizations.
Gordon et al., 2002).
 Strategies can be successful in view of the reported post-
A variety of implementation strategies are advised to
implementation adherence rates of 80% in half of the
facilitate the use of pain assessment guidelines (American
studies.
Pain Society, 1995; Gordon et al., 2005). Implementing the
 The level of evidence for strategies directed for improv-
recommendations from pain guidelines requires the use of
ing pain assessment recommendations is limited how-
theoretically underpinned strategies that reduce barriers
ever, due to a lack of well-conducted studies.
and stimulate facilitators (Grol et al., 2007). Continuous
education and training in pain assessment and manage-
1. Introduction ment is widely recommended. However, it is unclear if
training will suffice, and which interventions are most
Many hospitalised patients suffer from pain; when left effective in improving pain assessment adherence rates. A
untreated, pain may delay recovery and prolong hospital recent systematic review described the effectiveness of
stay (Ferrell et al., 1989; Jakobsson et al., 2007; Kehlet and standardized pain assessment tools on improving patient
Holte, 2001; Morrison et al., 2003; Perkins and Kehlet, (e.g. pain scores) and process outcomes (e.g. adherence) in
2000). However, patients’ pain is still under treated, not paediatrics (Franck and Bruce, 2009). The authors selected
only due to pain relief policies but also due to inadequate fourteen studies that evaluated the effect of pain assess-
and infrequent assessment, reassessment and documen- ment on patient outcomes (n = 10) or process outcomes
tation (Dalton et al., 2001; Dihle et al., 2006; Gelinas et al., (n = 12). All studies markedly differed in interventions
2004; Herr et al., 2004). Better pain detection is a applied, pain assessment instruments used, implementa-
fundamental first step in improving pain treatment, and tion strategies applied, and reported levels of adherence
one that has been difficult to achieve (Gordon et al., 2002). with the interventions. The implementation strategy
Good pain treatment includes regular assessment and education was widely used. Nevertheless, the effects on
reassessment, and these elements must be integrated in adherence must be interpreted with extreme caution due
pain management protocols or guidelines (Carlson, 2010). to the preponderance of combined strategies and multiple
They usually fall to the nursing domain as nurses have major methodological problems (Franck and Bruce, 2009).
more patient contacts than other health professionals. Therefore, it was difficult to draw firm conclusions. The
Studies on pain assessment in postoperative, cancer, only conclusion made was that pain assessment is not yet
paediatric and emergency patients identified barriers evidence-based.
related to health care professionals, such as knowledge The overall aim of this study was to systematically
deficits, misconceptions about assessment, lack of experi- review empirical evidence about the effectiveness of
ence, resistance to the use of validated tools, limited implementations strategies for the improvement of nurses’
competences or assessment skills, inadequate commu- adherence to pain assessment recommendations in
nication, and not accepting patients’ descriptions of pain hospitalised patients. We addressed two main questions:
as the gold standard (Berben et al., 2012; Carlson, 2010;
Czarnecki et al., 2011; Horbury et al., 2005; Van Niekerk 1. What implementation strategies for promoting nurses’
and Martin, 2001; Yildirim et al., 2008). Organizational adherence to pain assessment recommendations are
barriers identified are the lack of assessment protocols used?
and insufficient feedback on adherence to and quality of 2. What is the effectiveness of these implementation
pain assessment (Berben et al., 2012). Finally, patients are strategies measured in terms of adherence rates?
often reluctant to complain about pain. Efforts to improve
pain management have been made through the develop-
2. Methods
ment and dissemination of clinical guidelines. The
updated recommendations of the American Pain Society 2.1. Search strategy
focus on pain assessment, patient education, pain treat-
ment, an interdisciplinary approach, effective commu- Relevant studies were sought via the databases PubMed
nication and coordination of services (Miaskowski et al., (Medline), Embase, CINAHL and Cochrane library, from
2005). Another major development is the increased 1990 to May 2011 using the following search terms or
attention for monitoring quality of care. In the 1990s, equivalent index terms and free-text words for each of the
the Joint Commission for the Accreditation of Healthcare different databases: nurs* AND (compliance OR adherence
Organizations has mandated routine pain assessment in OR nursing evaluation) AND (pain assessment, pain
554 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

measurement), AND (implementation OR knowledge obtained. These were then screened by two authors (EI,
transfer OR quality improvement) (see Appendix A for TvA) to decide on in- or exclusion. In both rounds, the
extended search list). A librarian helped devise the search reviewer’s selections were compared and discrepancies
strategy. The authors are qualified researchers with were resolved through discussion.
expertise in performing systematic reviews. The search
was not limited by language of publication. 2.4. Data extraction

2.2. Criteria for inclusion One member of the review team (EI) independently
extracted data from all included studies; the two other
We included randomized controlled trials (RCTs); members (MvD, TvA) extracted data from half of the
cluster randomized controlled trials (C-RCTs), non-rando- studies. Again, results were compared and resolved
mized cluster controlled trials or quasi experimental through discussion in case of differences. Data on design,
studies, controlled before and after studies (CBAs), setting, subjects, implementation strategies, implementa-
interrupted time series (ITS), before–after studies without tion theory or model and outcomes were extracted. The
a control group and comparative studies with historical following outcomes were considered: adherence rates
controls. (percentages) for all types of pain assessment (e.g.
The following inclusion criteria enabling us to address assessment, reassessment after treatment intervention),
the specific research questions were set (see Fig. 1): and pain intensity or treatment effects before and after
implementation.
1. Descriptions of implementation strategies or interven- To facilitate comparison and discussion, the imple-
tions to promote adherence to pain assessment mentation strategies were classified using the Cochrane
recommendations (e.g. guidelines, protocols or proce- Effective Practice and Organization of Care group (EPOC)
dures for pain assessment, new or revised assessment classification system (EPOC, 2002) as being directed at
tool). Healthcare professionals, Organizations, Financing, or
2. Rates of adherence to (compliance with) pain assess- Regulations. Subcategories within the four main categories
ment recommendations (pain assessment or reassess- were also considered (Table 2).
ment after interventions) reported before and after
implementation activities and/or for intervention and 2.5. Methodological quality
control groups.
3. Pain assessment rates for nurses or nurse practitioners Methodological quality of the studies was evaluated
reported. with three EPOC tools; the RCT, CBA, and the interrupted
4. Pain assessment rates in hospitalised patients (children times series (ITS) tools (EPOC, 2002). The RCT tool was
and adults) or patients admitted to an emergency used for (C)RCT and controlled clinical trail designs; it
department reported. contains items related to unit of analysis, power, baseline
Studies evaluating non-behavioural outcomes only (e.g. measure, concealment of allocation, blinded or objective
nurses’ knowledge or attitudes) were excluded. Further- assessment of outcome(s) protection against contamina-
more, narrative reviews and opinion papers were tion, reliable outcome(s), and completeness of follow-up.
excluded. The CBA tool was used for CBA and before–after study
designs without control groups. The ITS tool contains
2.3. Identification of studies items related to unit of analysis, power, baseline measure,
comparability of groups, blinded or objective assessment
One author (EI) assessed all and each of the other of outcome(s), protection against contamination, reliable
authors (MvD, TvA) reviewed half of the retrieved titles outcome(s), and completeness of follow-up. Each item
and abstracts on relevance for this review. At this stage, was scored as: done, not done, or: not clear. Two reviewers
articles were selected if they clearly or possibly met the assessed each study and discrepancies were resolved
inclusion criteria. For these publications, full texts were through discussion.

Fig. 1. Conceptual model.

 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568 555

2.6. Data analysis eligibility. Then, 34 articles were excluded because the
setting was outside the hospital, adherence rates were
Due to different definitions of adherence rates, pooling not reported before and after implementation activities
of effects sizes was impossible. Furthermore, the hetero- and/or for intervention and control groups. Eventually
geneity of implementation strategies, outcomes and we included 23 studies meeting all criteria. These are
participants precluded a meta-analysis; a narrative synth- ranked in Table 1, hierarchically for type of design, and
esis was thought more appropriate. alphabetically.

3.2. Methodological quality

3. Results

3.1. Search results Overall, the quality of the studies was low (Appendix B).
Only two studies were RCTs (Dalton et al., 2001; Johnston
The searches resulted in 816 hits and after removing et al., 2007), one study was a controlled clinical trail
duplicates 743 abstracts were reviewed. Fig. 2 shows the (Morrison et al., 2006) and the remaining 19 studies were
search strategy and the results. Based on abstract and before–after studies without control group or time series
title selection, 57 full-text articles were assessed for comparisons (Narasimhaswamy et al., 2006; Oakes et al.,

Fig. 2. Search strategy.

 556
Table 1
Characteristics of the included studies.

Author (country) Design (data collection) Setting, sample Intervention Implementation Process outcome Model/theory Results – adherence
(focus)a strategiesb (adherence, criteria) driven rates [before (%)
vs. after (%)]

Dalton Cluster-RCT Six community G Educational Perc. of: NRS 0–10; pain ND NRS pain
et al. (2001) hospitals (100–500 programme (3 intensity, quality, duration,
USA beds) (3 Exp; 3 Contr), sessions) behaviour. (pain
PACU and surgical floor. documented yes/no)
Chart audit: Pre (T0); Exp: (n = 368; T1:126; Exp.: 43% (T0) vs. 49%

E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

programme completed T2:131; T3:111) Contr: (T1); 73% (T2)
(T1); 6 mths (T2) n = 418 (T1:134; Contr: 34% (T0) vs. 34%
T2:142: T3:142) (T1); 42% (T2)
(x2 = 12.37, p < 0.001)
Johnston Cluster RCT Multiple (unspecified) PPA One-to-one Perc. of pain PARIHS Exp group: 15 to 58%
et al. (2007) paediatric wards, 6 coaching with A&F documentation model
Canada hospitals (3 children’s (exp) vs. Audit
hospitals) (contr)
Pretest–posttest Hospitals randomized, (x2 = 138.34,
comparison p < 0.0001)
Prospective review of patient chart sampling
patient charts 2 weeks unspecified (n = 464, Cont group: 24% to 9%,
pre and (interval 306 [985 charts] pre, (x2 = 34.86, p < 0.001)
unspecified) post 158 [617 charts] post)
Morrison Controlled clinical trial 1172 beds teaching PPA Ph. 1: Education Perc. pain assessments ND Ph. II: 32% vs. 64%
et al. (2006) hospital, 9 wards (p < 0.001)
USA (surgical, internal)
2 blocks (matched) Ph. 2: 1-item vs. (at least 1 assessment per Ph. III: 64% vs. 85%
enhanced pain shift, for first 5 days) (p < 0.001)
scale (A vs. B)
4 phases: Wards randomized, Ph. 3: both Ph. IV: 79% vs. 64%
I: 0–4 mths; II: 5–11 enhanced scale, (p < 0.001)
mths; III: 12–19 mths; +A&F (block B)
IV: 20–25 mths. 25% of all admissions Ph. 4: enhanced
screened scale, CDSS (A + B);
+A&F (B)
Ang and Pre-post-test (without Tertiary hospital, 24 PPA Education 1. Initial pain assessment GRIP 1. T0: 50% vs. T1: 83%
Chow (2010) control gr) beds oncology ward. programme on admission programme (x2 = 6.207, p = 0.03)
Singapore (workshop)
Prospective (3-weeks) n = 24 A&F 2. Patients reassessed to (barriers 2. T0: 29% vs. T1: 75%
audit: before (T0) determine treatment effect identified) (x2 = 10.492, p = 0.003)
(3-weeks); 3 mths Revised pain
after (T1) management
(flowchart)
Material support
Bach (1995) Retrospective Post operative patients, G, PM Education 1. Perc. standard RU 1. T0: 90%; T1: 73%; T2:
USA Pre-post-test hospital professionals assessment (before: a 4 h; 84%
(without control gr) after: a 2 h)
Chart audit: Pre Patients: n = 30 (T0), Modified guideline 2. Perc. pain intensity 2. T0: 3%; T1: 45%; T2:
(T0); post: 6 mths n = 22 (T1), n = 25 (T2) 84%
(T1), 10 mths (T2) Patient material 3. Perc. of documentation 3. T0: 83%; T1: 91%; T2:
updated after intervention 92%
Material support (no p value reported)
Baumann Pre-post-test Emergency PM Templated chart on Perc. pain assessments at ND Overall: 41% vs. 57%
et al. (2007) Department, urban documentation of triage (physicians/NPs) (p < 0.001) (non
USA tertiary University pain assessment templated vs.
hospital. template)
Prospective chart Patients: n = 768 (T0), Physicians: 39% vs. 59%
review: pre (T0), n = 471 (T1) (p < 0.001)
2 mths post (T1)

E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

NP: 41% vs. 54% (ns)
Bookbinder Pre-post-test Cancer Centre, Tertiary PM Interdisciplinary Perc. twice daily screening QI, framework, T0: 55% vs. T1: 91%
et al. (1996) (without control hospital (12 units, 565- pain team intensity PDSA, barriers
USA group); T0 – beds) identified.
baseline; T1 one Patients: n = 398 (T0), Structure (no p value reported)
year later n = 298 (T1)
Education
Campbell Pre-post-test Emergency department PM Education Perc. of pain assessed ND 1. T0: 25% vs. T1:
et al. (2004) (55 beds), Hospital. (NRS): 43% (a)
USA Chart audit: Pre Patients: n = 100 (T0), New pain protocol 1. Admission ED 2. T0: 63% vs. T1:
(T0); post: n = 30 (T1) 90% (a)
4 mths (T1). Feedback 2. During ED stay 3. T0: 43% vs. T1:
88% (a)
Leadership 3. At discharge (no p value reported)
Dulko Pre-post-test Hospital, nurse G NP weekly 1. Perc. initial pain ND 1. T0: 1% vs. T1: 43%
et al. (2010) practitioners (NP) feedback assessment (overall)
USA (n = 8) thoracic
medicine (4),
neurology (4).
Pre (T0): 3 mths; Patients: n = 96 (T0); Educational session 2. Perc. pain reassessment 1% (1–10) vs. 45% (11–
post (T1): 4 mths. n = 96 (T1) within timeframe 75)
Laminated pocket (p = 0.008) [mean NP
guideline version rate (range)]
2. T0: 5% vs. T1: 87%
[overall]
7% (0–20) vs. 87% (73–
100) (p = 0.008) [mean
NP rate (range)]
Duncan and Quasi-experimental, Orthopaedic unit of a G Individual 1. Perc. of pain assessment ND 1. T0: 72% vs. T1:83%
Pozehl (2001) Pre-post-test mid-size acute care performance every 4 h.
USA hospital (n = 34 nurses) feedback
Retrospective Patients: n = 119 (T0); 2. Perc of reassessment
review of medical n = 122 (T1) after analgesics
patient record, (over 4 days of patient 2. 43% vs. 64%
pre (T0) 15 wks.; admission) (no p value reported)
post 34 wks.
(T1) -17 wks.

557
 558
Table 1 (Continued )

Author (country) Design (data collection) Setting, sample Intervention Implementation Process outcome Model/theory Results – adherence
(focus)a strategiesb (adherence, criteria) driven rates [before (%)
vs. after (%)]

Ellis Pre-post-test 5 paediatric wards, PM Education Perc. of pain scales used Ottawa T0: 5%
et al. (2007) general hospital workshop Model of
Canada Research Use
Clinical audit, Patients (random Champions T1: 43% (T0 vs. T1,
3 times: pre (T0), selected): n = 75 (T0), (leaders) p < 0.001)
6 mths post 1 n = 44 (T1), n = 50 (T2) Feedback T2: 24% (T0 vs. T2,
(T1), post 2 (T2) p = 0.005)
Others (supportive)
Erdek and Pre-post-test 2 surgical ICUs PPA, PM In-service sessions, Perc. of number of VAS PDSA QI, T0: 42%
Pronovost (general/oncologic) importance of scores

E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

(2004) documentation
USA Clinical audit, Patients: random Modified scale Barrier T1: 54%
5 wks: pre wk. sample n = 10–15 per identification
1 (T0), post wk. ICU.
2–5 (T1–T4) Daily discussion T2: 70%
T4: 75%
T5: 71% (no p value
reported)
Gelinas Pre-post-test Intensive Care Unit, PAT Education Perc. of reassessment ND T0: 10%,
et al. (2011) Academic hospital. workshop
Canada (including video)
Pre (T0), post Patients: n = 30 T0, Pocket pain cards T1: 43%,
3 mnts (T1), n = 30 T1, n = 30 T2 Reminders – T2: 59%
12 mnts (T2). posters
Support of (T1–T2: U = 298.00,
clinicians p = 0.97)
Idell Pre-post-test Medical/surgical G, PPA Individual Perc. of reassessment CURN model T0 61% vs. T1 78%
et al. (2007) oncology units, Cancer performance (within 1 h.)
USA hospital. feedback
Chart audit, Patients: unspecified Reminders – (t-test, p = 0.004)
5 charts among posters
patients selected
for each nurse
(n = 42)
Case discussions
Innis Pre-post-test Internal medicine unit PM Pain education Perc. of pain assessment ND T0 52% vs. T1 100%
et al. (2004) (74-beds), teaching (interval unspecified)
Canada Hospital (urban).
Chart audit, Patients: n = 50 (T0), Reminders – poster (x2 = 60.554, p < 0.001)
medical records: n = 50 (T1)
pre (T0), post (T1)
Pocket pain
management cards
Jordan-Marsh Pre-post-test Paediatric ward PM Pain tool 1. Perc. of documentation Social 1. T0: 54% vs. T1: 93%,
et al. (2004) (serial) comparison (number unspecified), pain intensity ecological
USA during and after general hospital. approach
4-phase:
T0: Q1 (pre); T1: Education 2. Perc. evaluation T3 (sustain) 84%
Q2-8; T2: Q 9-13; treatment effect
T3: Q14.
Chart audit: random Leadership
selection of 10% Rounds 2. T0: 80% vs. T1: 97%,
of patients’ charts. T3 89%
Revised pain policy (no p value reported)
Michaels Quasi experimental. 16 units (Medical/ PM, PPA Educational 1. Perc. pain assessment ND 1. No (statistical)
et al. (2007) surgery, adult critical programme (Exp. during 24 h differences were found
USA care, geriatrics, units) between exp. vs. co
paediatrics) of units

E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

Academic medical
centre (841-beds).
Chart audit Exp. n = 8 (units); Contr. 2. Perc. reassessment 2. 43% at baseline
n=8 (within 1 h) versus 52%
Patients: n = 911 (no p value reported)
O’Connor Pre-post-test Surgical, orthopaedic PPA Documentation 1. Perc. of presence pain ND 1. 43% vs. 83%
(2003) inpatients, General tool documentation on
Hospital. admission
Retrospective chart Patients: n = 30 (T0), 2. Perc. of pain assessment 2. 50% vs. 53%
audit: random n = 30 (T1) every 4 h (during first 24
selection pre, (T0) after surgery)
1 month post (T1)
(no p value reported)
Phelan (2010) Pre-post-test + Neuro/neurosurgery PM Multidisciplinary 1. Perc. pain assessment on ND 1. 25% vs. 65%
Australia mixed methods unit (32 beds), Large education admission
tertiary referral programme
hospital.
Chart audit Patients (random): Reminders 2. Perc. pain assessment 2. 5% vs. 35%
(random): pre n = 20 (T0), n = 20 (T1) during admission
(T0), 8 wks Promotional 3. Perc. of reassessment 3. 10% vs. 35%
post (T1) material
(no p value reported)
Treadwell Pre-Post-test Haematology/oncology PPA Education Adherence rate pain QI programme 30% vs. 59%
et al. (2002) (1-year interval) ward (26-beds), large programme assessment (in accordance
USA regional tertiary care with unit protocol)
children’s hospital.
Chart audit: Patients (random Reminders PDSA (no p value reported)
pre (T0), post (T1) sample): n = 36 (T0), (posters)
n = 49 (T1)
Leadership
White (1999) Pre-post-test Neurological hospital, PM Education 1. Perc. of pain assessment ND 1. T0: 40%; T1: 90%, T2
Canada postoperative after programme (first 3 h after surgery) 80% (x2 = 11.31,
spinal surgery. p < 0.01)
Chart audit: pre Patients: n = 15 (T0), 2. Perc. of reassessment 2. T0: 0%, T1: 25%, T2:
(T0), 3 mths n = 20 (T1), n = 15 (T2) 33% (x2 = 5.73, p > 0.05
post (T1), 2 yrs. ns)
post (T2)

559
 560
Table 1 (Continued )

Author (country) Design (data collection) Setting, sample Intervention Implementation Process outcome Model/theory Results – adherence
(focus)a strategiesb (adherence, criteria) driven rates [before (%)
vs. after (%)]

E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

Narasimhaswamy Time series Suburban teaching PM Education sessions Compliance pain ND Median perc. increased
et al. (2006) (retrospective) hospital (550-beds) assessment during first over time: 77% (January
USA 72 h 2001) to 92% (January
2004) (p < 0.001)
2001–‘02 Patients: unspecified Feedback
(individual nurse)
2003-implementation Clinical case
discussion
2003–‘04 New forms
Oakes Time series Tertiary Children’s PM Continuous QI 1. Perc. of pain assessments ND 1. Overall mean
et al. (2008) cancer hospital (60- every 4 h adherence 87%. Q1: 77%
USA beds). increased to 90% or
higher over time (Q16
and later).
Comparisons over All patient charts for 87 Feedback 2. Perc of reassessment 2. overall 78% (range
6 years (quarterly 24h periods (n = 2478) routine chart (within 1 h) 40–100%)
n = 24 time periods) audit
Reminders

RCT – randomized controlled trial.

a
G – guideline (recommendations); PM – protocol or programme for pain management; PAT – (new) pain assessment tool; PPA – practice or policy of pain assessment.
b
According description of authors, not according to EPOC classification; mths – months; exp. – experimental group; contr. – control group; PACU – post anaesthesia care unit; Ph. – phase; vs. – versus; (a) –
approximately, percentage from figure; ND – not described; A&F – audit & feedback; CDSS – computerized decision support system; PARIHS model – promoting action on research implementation in health
services; GRIP – getting research into practice programme; RU – research utilization; Perc. – percentage; QI – quality improvement; PDSA – plan do study act cycles; ED – emergency department; NRS – numeric
rating scale; ICU – intensive care unit; VAS – visual analogue scale; CURN – conduct and utilization of research in nursing model; Q – quarter.
 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568 561

2008). Most of the studies aimed to evaluate the quality of Reminders of pain assessment were typically in the
pain management following quality improvement pro- shape of posters, laminated pocket pain tools, manage-
jects. The quality of the studies varied considerably. The ment cards, or guidelines (including assessment) (Ang and
most common problem was the limited detail provided on Chow, 2010; Campbell et al., 2004; Dulko et al., 2010; Ellis
the baseline characteristics of the sample. Only, Morrison et al., 2007; Erdek and Pronovost, 2004; Innis et al., 2004;
et al. (2006) controlled for patient characteristics in O’Connor, 2003; Phelan, 2010) (Table 2).
the statistical analysis. Furthermore, sample sizes were Structural interventions in terms of ‘‘changes in
often small, adequate controls were lacking, and statis- physical structure, facilities or equipment’’ (e.g. modified
tical analyses were not rigorous enough. Only one study pain scales, new or revised pain management protocols,
provided a sample size calculation, relating to establish- algorithms or guidelines), belonging to the EPOC category
ing significant differences between adherence rates. organizational strategies, were used in 14 studies (see
The quality of the studies of Morrison et al. (2006) Table 2).
and Baumann et al. (2007) was considered good. Finally, in seven studies a single strategy was applied,
These provided detailed methodological and statistical either directed at the health professionals (e.g. education
information. or feedback) (Dalton et al., 2001; Duncan and Pozehl, 2001;
Michaels et al., 2007; White, 1999) or at the organization
3.3. Implementation strategies (e.g. facilities) (Baumann et al., 2007; Michaels et al., 2007;
O’Connor, 2003) (see Table 2). The remaining studies used
Most of the studies aimed to improve adherence to multifaceted intervention programmes.
pain assessment recommendations combined with pain A rationale for the choice of implementation strategy
treatment. Four studies focused on improving pain was only given in four studies. In three of them, it was
assessment, reassessment or pain documentation only based on an analysis of barriers and facilitators (Ang and
(Gelinas et al., 2011; Idell et al., 2007; O’Connor, 2003; Chow, 2010; Bookbinder et al., 1996; Erdek and
Treadwell et al., 2002). The quality of the descriptions of Pronovost, 2004). Only Jordan-Marsh et al. (2004) used
the implementation strategies varied and the description a theoretical underpinning, i.e. the social ecological
was sometimes ambiguous. However, most of the studies model as an ideal approach for overcoming the limita-
described the strategies in detail, so that they could be tions of interventions that focus simply on changing
replicated. Table 2 provides the EPOC classification of the individuals’ behaviours at individual, organizational and
strategies used in the various studies. The studies societal level (Smedley and Syme, 2000). In eight studies,
reported only strategies related to the domains of health the implementation was based on a model (Ang and
professionals (e.g. education, feedback and reminders) Chow, 2010; Bookbinder et al., 1996; Dalton et al., 2001;
and organization (e.g. ‘‘changes in physical structure, Ellis et al., 2007; Erdek and Pronovost, 2004; Idell et al.,
facilities and equipment’’). 2007; Treadwell et al., 2002). However, these models did
Education was the most commonly used strategy, in 17 not provide a rationale for the choice of implementation
of 23 studies. However, type and duration of education strategy.
largely differed between studies. We identified educa-
tional programmes on the relevance of pain assessment 3.4. Effectiveness of type of implementation strategy on
and barriers in pain assessment (Ang and Chow, 2010; adherence
Bach, 1995; Bookbinder et al., 1996; Dalton et al., 2001;
Michaels et al., 2007; Narasimhaswamy et al., 2006); Overall, the different implementation strategies used
educational workshops with discussion on attitudes and had widely ranging effects on adherence to pain assess-
values of assessment (Ellis et al., 2007); and face-to-face ment recommendations. In all studies but one, the absolute
education (Phelan, 2010). Educational sessions lasted from adherence rates had improved after implementation
30 to 60 min (Innis et al., 2004; Phelan, 2010; White, 1999) compared to the before measurement, by 9% up to 49%
to 4 h (Ellis et al., 2007). (see Table 1 and Fig. 3). Fig. 3 shows percentages reported
In three studies, education was the only intervention in the individual studies broken down for EPOC type of
(Dalton et al., 2001; Michaels et al., 2007; White, 1999); in implementation strategy, i.e. health professional strate-
the remaining 14 studies it was combined with other gies, organizational strategies and combinations of both.
implementation strategies, such as feedback, reminders, After implementation, adherence rates ranged from 24 to
outreach visits, or changes in structure (e.g. revised pain 83% and 59 to 100%, for professional strategies and the
protocol). combination of professional and organizational strategies
Feedback and reminders were used in 11 and 10 of respectively. Furthermore, nearly half of the reviewed
the studies respectively. We identified feedback on studies reached an 80% or higher pain assessment
individual performance (Duncan and Pozehl, 2001; Idell adherence rate.
et al., 2007; Johnston et al., 2007; Narasimhaswamy The effect of educational strategies alone varied from 9
et al., 2006) and feedback at unit level (Ang and Chow, to 40% improvement (Dalton et al., 2001; Michaels et al.,
2010; Campbell et al., 2004; Ellis et al., 2007; Morrison 2007; White, 1999). Dalton et al. (2001) found that the
et al., 2006; Oakes et al., 2008). Only Morrison et al. adherence to postoperative pain assessments showed a
(2006) and Johnston et al. (2007) used a controlled significant programme effect from 43 to 73%. Michaels
design to study the effect of audit and feedback (A&F) on et al. (2007) found no effect (from 43 to 52%) of education
adherence. on adherence to postoperative pain assessments, or
562 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

Table 2
Implementation strategies classified according EPOC taxonomy.

Health professional strategies Health professional and Organizational strategies OS

Narasimhaswamy et al. (2006)

Erdek and Pronovost (2004)
Duncan and Pozehl (2001)

Jordan-Marsh et al. (2004)

Bookbinder et al. (1996)

Treadwell et al. (2002)

Baumann et al. (2007)

Campbell et al. (2004)
Morrison et al. (2006)
Ang and Chow (2010)

Michaels et al. (2007)

Johnston et al. (2007)

Gelinas et al. (2011)

Dalton et al. (2001)

Oakes et al. (2008)

Dulko et al. (2010)

Innis et al. (2004)

Idell et al. (2007)
Ellis et al. (2007)

O’Connor (2003)

Phelan (2010)
White (1999)

Bach (1995)
Health professionals
1 Distribution of X
educational materials
2 Education meetings X X X X X X X X X X X X X X X X
16x
3 Local consensus X X X X
process
4 Outreach visits X X X X
5 Local opinion leaders X X X X
6 Patient-mediated
intervention
7 Audit and feedback X X X X X X X X X X
8 Reminders
a. Computerized X
decision support
b. Concurrent reports
c. Inter-visit reminders
d. Enhanced laboratory
report
e. Administrative X
support
f. Implicit reminders X X X X X X X X X X X
9 Tailored interventions X X X
10 Peer review
11 Combined strategies X
Organizational interventions
12 Structural
interventions
a. Changes in the
settings and/or site of
service delivery
b. Telemedicine
c. Changes in medical X
records systems
d. Changes in
arrangements to
maintain/retrieve
information
e. Changes in physical X X X X X X X X X X X
structure, facilities and
equipment
f. Changes in scope and/
or nature of services
g. Changes in presence X X X X X
and organization of
quality
h. Staff organization
i. Other (rounds) X
13 Staff-oriented
interventions
Revision of professional
roles
Multidisciplinary
teams
Case management
Other integration of
services
 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568 563

Table 2 (Continued )

Health professional strategies Health professional and Organizational strategies OS

Narasimhaswamy et al. (2006)

Erdek and Pronovost (2004)
Duncan and Pozehl (2001)

Jordan-Marsh et al. (2004)

Bookbinder et al. (1996)

Treadwell et al. (2002)

Baumann et al. (2007)

Campbell et al. (2004)
Morrison et al. (2006)
Ang and Chow (2010)

Michaels et al. (2007)

Johnston et al. (2007)

Gelinas et al. (2011)

Dalton et al. (2001)

Oakes et al. (2008)

Dulko et al. (2010)

Innis et al. (2004)

Idell et al. (2007)
Ellis et al. (2007)

O’Connor (2003)

Phelan (2010)
White (1999)

Bach (1995)
Skill mix interventions
Interventions to
improve provider
satisfaction with the
conditions of work or
its material and/or
psychological rewards
Other (e.g. case X
discussion)
14 Patient-oriented
interventions
Interventions to X X
facilitate individual
patient participation
Interventions to
facilitate patient group
participation
Other
Regulatory interventions
15 Changes in medical
liability
16 Management of patient
complaints
17 Accreditation
18 Licensure
19 Other
Financial interventions
20 Healthcare-oriented
interventions
21 Patient-oriented
interventions
OS – Organizational strategies.

reassessments within one hour, as compared to the control 3.6. Pain intensity, effects on treatment
wards. In contrast, White (1999) detected an effect of
education on adherence to pain assessment recommenda- Fourteen studies reported effects on pain intensity or
tions (from 40 to 80%). treatment after implementation of pain assessment or pain
Morrison et al. (2006) and Johnston et al. (2007) management programmes (Bach, 1995; Baumann et al.,
reported the effects of audit and feedback on adherence 2007; Dulko et al., 2010; Ellis et al., 2007; Erdek and
rates. Both studies found a significant increase in Pronovost, 2004; Gelinas et al., 2011; Innis et al., 2004;
adherence rates, from 64 to 74% and from 15 to 58% Johnston et al., 2007; Jordan-Marsh et al., 2004; Morrison
respectively, for pain assessment and documentation. et al., 2006; Narasimhaswamy et al., 2006; Oakes et al.,
2008; Treadwell et al., 2002; White, 1999). These effects
3.5. Evaluation interval and effects over time were diverse, however, and are hard to compare because
different outcomes are reported (e.g. median pain score,
The effects of implementation strategies were mea- amount of administered analgesics).
sured at different time points after completion of the Only White (1999), Erdek and Pronovost (2004) and
implementation, ranging from 2 weeks to 6 months. Oakes et al. (2008) reported that patients’ pain scores
Sustained effects, established through repeated measure- decreased after implementation of pain assessment and
ment after implementation, were reported in three studies management. The majority of the studies found no
(Jordan-Marsh et al., 2004; Oakes et al., 2008; White, effects on treatment or pain intensity (Baumann et al.,
1999). These authors found adherence rates of 80% or 2007; Dulko et al., 2010; Ellis et al., 2007; Innis et al.,
higher in the long-term (1–6 years). 2004; Johnston et al., 2007; Morrison et al., 2006;
564 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

Fig. 3. Adherence rates documented before and after implementation.

Narasimhaswamy et al., 2006; Treadwell et al., 2002). reviewed studies. Also, getting local opinion leaders or
Treatment effect was also reported in terms of increase champions involved could benefit pain management
or decrease in analgesics consumption. In the study of (Bookbinder et al., 1996; Campbell et al., 2004; Ellis
Gelinas et al. (2011) analgesics consumption dropped; et al., 2007). Furthermore, selecting implementation
Johnston et al. (2007) reported no change in analgesics strategies that effectively pull down barriers result in
consumption. In contrast, White (1999) reported higher more successful implementation. The studies in this
analgesics consumption, coupled with lower pain scores. review that used this approach showed also an improve-
ment in adherence to pain assessment recommendations
4. Discussion (Ang and Chow, 2010; Bookbinder et al., 1996; Erdek and
Pronovost, 2004). Tailor-made implementation strategies
This systematic review provides an overview of effects based on content analysis of barriers and facilitators seem
of implementation strategies for improving adherence to to be more effective (Baker et al., 2010; Bosch et al., 2007).
pain assessment recommendations in hospitalised Not many of the reviewed studies used tailoring, however,
patients. In summary, the body of properly conducted and further research in this field is recommended. Based on
studies was too small to allow for a clear conclusion on the the adherence rates it seems that multifaceted strategies –
effectiveness of implementation strategies. Most of the combinations of professional and organizational strategies
studies were uncontrolled before–after designs with small – proved more effective than only professional (single)
sample sizes. However, studies such as performed by strategies. This is in line with the literature (Grimshaw
Morrison et al. (2006), Dalton et al. (2001) and Johnston et al., 2004). The latter conclusion is a tentative one,
et al. (2007) had higher quality. however, as the studies had methodological limitations.
Theoretical grounding of implementation strategies
4.1. Implementation strategies brings potential benefit in terms of facilitating and
understanding of implementation (Bhattacharyya et al.,
Lack of knowledge (Berben et al., 2012; Czarnecki et al., 2006; Davies et al., 2010). Nevertheless, this approach was
2011); low priority given to pain management; time little used in these reviewed studies.
constraints; and insufficient physician medication orders Education as implementation strategy was applied in
(Czarnecki et al., 2011) may stand in the way of good most of the studies but in different ways. We believe that
adherence to clinical pain assessment or management education is indeed needed to provide healthcare profes-
guidelines. Education as an implementation strategy sionals with the latest insights on pain management, but
removed the barrier of lack of knowledge in most of the dissemination of written educational materials or didactic
 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568 565

education at large scale seems an insufficient strategy to care gains transparency with the help of performance
improve pain assessment and treatment (Grimshaw et al., indicators. For Dutch hospitals, pain assessment is a
2004; Grol and Grimshaw, 2003). On the other hand, performance indicator initiated by the Health Care
interactive educational workshops and are largely effective Inspectorate. However, high adherence rates for pain
(Forsetlund et al., 2009; O’Brien et al., 2007). These findings assessment is not a guarantee for high quality of care.
could be confirmed partially in this study. Meta-analysis Moreover, this performance indicator should be linked to
has revealed that audit and feedback is a more effective treatment, so as to improve patient outcomes. This review
strategy and this could be confirmed in this review showed that this issue was studied insufficiently.
(Jamtvedt et al., 2009). Also, it remains still problematic for physicians,
Finally, several studies had a broader scope, focusing nurses and other healthcare professionals alike to adhere
also on improvement of pain treatment. In these studies it to guidelines and protocol in healthcare (Cabana et al.,
proved hard to unravel the effects of implementation 1999; Lugtenberg et al., 2011, 2009). It would seem
strategies directed only at improving adherence to pain essential, therefore, to devise implementation strategies
assessment. Also, these studies were performed in many – that are effective to increase adherence to pain
different settings. Setting and staff attitude may have an assessment recommendations. First, we must provide
impact: in oncology departments, for example, it will be good education on the importance of pain assessment
easier to motivate nurses to assess pain, whereas in the itself. Second, tailored and barrier-driven implementa-
emergency setting one tends to move quickly to solve the tion strategies focusing on specific barriers in a certain
problem, with less attention to pain as a result. Further- setting are needed to improve adherence in practice.
more, several studies only briefly described their imple- Third, feedback on performance could be useful to
mentation strategies, which makes it difficult for others to increase the intrinsic motivation of nurses (Johnston
replicate these strategies. et al., 2007; Morrison et al., 2006). Finally, patient-centric
strategies might be useful, but only if they have an
4.2. Adherence outcome measure of improving nurses adherence to pain
assessment recommendations (Coulter and Ellins, 2007).
All studies but one showed improved adherence to Such strategies were not used in the reviewed studies,
pain assessment recommendations after implementation but would be worth exploring in future research. Recent
of single or combined strategies. Bach (1995), as the one paediatric and adult studies already found that parent or
exception, found that standard postoperative pain assess- patient education about pain management improved
ment was less adhered to after implementation of a nurses’ pain assessment and documentation (Franck
modified guideline. The modified guideline implied more et al., 2011; Haller et al., 2011).
frequent pain assessments (every 2 h) than before which
may explain the decreased adherence. In our experience it 4.3. Time effects
is crucial to tailor the frequency of pain assessments in
specific patient groups; not too often which demotivates The improved nurses adherence on pain assessment
nurses and not too infrequent with the risk of missing recommendations were assessed at different time points,
pain. ranging from 2 weeks to 6 months after implementation,
Overall, half of the reviewed studies reported a which may affect validity of the results considerably.
clinically relevant adherence rate of 80% or higher after Effects may wane over time or other factors contribute as
the implementation activities. However, we should con- well to the results. On the other hand, after a shorter period
sider that the frequency of prescribed pain assessments the results may be too rosy or implementation has not yet
varies considerably among these studies. Although the set in. This underlines the need to assess adherence
American Pain Society designated pain as the ‘‘Fifth Vital repeatedly for a longer period of time. Most of the reported
Sign’’, next to heart rate, blood pressure, respiratory rate, effects on pain treatment and intensity (scores) were
and temperature, based on this review we believe that established in uncontrolled before–after studies. So, the
adherence to pain assessment is still suboptimal. It has observed changes may be due to factors other than the
been suggested that the basic principles and individual intervention.
activities related to pain assessment have been inade- Three studies explored the sustainability over time;
quately incorporated into patterns of daily practice (Dahl however, on methodological grounds (e.g. small sample
et al., 2003). size and chart audits) the results and their generalizability
The central question is, why do nurses not or poorly must be interpreted with caution. Furthermore, the
adhere to pain assessment or management protocols, and nature of quality improvement is such that it is hard to
is it harmful for patients? Poor adherence may reflect separate the outcomes from the effects of environmental
unspoken resistance to use of methods that are overly factors. Nevertheless, Oakes et al. (2011) succeeded in
simplistic, burdensome to patients, often inaccurate and sustained improved pain management over a 10 years
perhaps even disrespectful of clinical expertise and period.
experience (Franck and Bruce, 2009). On the other hand,
untreated pain may have harmful effects and effective 4.4. Limitations of this review
treatment is therefore crucial (Ferrell et al., 1989;
Jakobsson et al., 2007; Kehlet and Holte, 2001; Morrison Several methodological limitations of this review need
et al., 2003; Perkins and Kehlet, 2000). Finally, quality of to be addressed. First, we could not perform a meta-
566 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

analysis. Second, the EPOC classification is debatable,

follow up
Patient
because the content for e.g. professional and organiza-
tional strategies is incomplete. By example, the profes-

n/a
n/a

n/a

n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
sional oriented strategy ‘‘educational meeting’’ is very
general. We would suggest to split this in different types,

follow up
Provider
e.g. large- versus small scale educational meetings and
active versus passive learning. Nevertheless, it is a

UC
UC

UC

UC
UC
UC
UC
UC
UC
UC
UC
UC
currently widely used taxonomy and may be desirable
in order to join this taxonomy in comparing study results
(Grol and Wensing, 2011).

adequate
Statistic
analysis

UC
5. Conclusions

V
V

V
X
V

X
V
V
V
X
Based on this systematic review we conclude that

outcome
measure
Reliable
implementation strategies to improve nurses’ adherence
to pain assessment recommendations vary but generally

UC
V

V
V
V
V
V
V
V
V
V
address professionals and organizational aspects. Educa-
tional and feedback strategies are often used and

contamination
effective in controlled studies. Education is necessary

Protection
to inform nurses and other professionals about the

against
necessity, values, and evidence about pain assessment,

n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
however, it is not a guarantee for excellent improve-

V
V

V
ment.
Most of the studies showed an improvement in

assessment
adherence. However, due to the heterogeneity of the

outcome
Blinded
implementation strategies it is not possible to recommend

UC
UC

UC
one preferred strategy. The more so because solid evidence

V
V
V
V
V
V
V
V
V
of effectiveness is lacking due to the general poor quality of
the studies.
Characteristics

For clinical practice we recommend interactive educa-

of control

tion sessions, involving local champions and giving feed-

back on individual or unit level to increase the intrinsic
UV
UC

UC
UC

UC

UC
UC
UC
motivation of professionals. V

V
V

V
Conflict of interest: None declared.
measure
Baseline

Funding: EI was supported by an ‘‘implementation research

UC

UC
UC

UC
V

V
V

V
V
V
V
fellowship’’ grant of The Netherlands Organization for Health
Research and Development (ZonMw), ZonMw grant:
calculation

170996007.
(Controlled) BA methodological quality assessment results and rating
Appendix B Methodological quality if included studies.

Power

Acknowledgement
X
X

X
X
V
X
X
X
V
X
X

The authors thank Ko Hagoort for carefully editing the

RCT methodological quality assessment results and rating

CCT methodological quality assessment results and rating

Provider
Provider

Provider

Provider
Provider
Provider
Provider
Provider
Provider
Provider
Provider
Provider
analysis
Unit of

manuscript.

Appendix A. Search strategy

allocation

PubMed
Unit of

Ward
Ward

Ward

Ward
Ward
Ward
Ward
Ward
Ward
Ward
Ward
Ward

(nurs*[tw] AND (complian*[tw] OR adheren*[tw] OR

adhaeren*[tw]) OR nursing audit*[tw]) AND (pain mea-
sur*[tw] OR pain assess*[tw] OR pain/diagnosis[mesh])
Erdek and Pronovost (2004)

EMbase
Duncan and Pozehl (2001)
Bookbinder et al. (1996)

((nurs* NEAR/5 (complian* OR adheren* OR adhaeren*)):-

Baumann et al. (2007)

Campbell et al. (2004)

Morrison et al. (2006)

Ang and Chow (2010)

Johnston et al. (2007)

ti,ab,de OR ‘nursing research’/exp OR ‘nurse research’:ti,ab,de

Gelinas et al. (2011)
Dalton et al. (2001)

Dulko et al. (2010)

OR ‘nurse evaluation’:ti,ab,de) AND ((pain NEAR/2 (measur*

Ellis et al. (2007)

OR assess* OR diagnos*)):ti,ab,de)
Bach (1995)
First author
 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568 567

References

adequate
follow up

Statistic
analysis
Patient

American Pain Society, 1995. Quality improvement guidelines for the

n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a

V
X
treatment of acute pain and cancer pain. American Pain Society
Quality of Care Committee. Journal of the American Medical Associa-
tion 274 (23), 1874–1880.
follow up

Ang, E., Chow, Y.L., 2010. General pain assessment among patients with
Provider

cancer in an acute care setting: a best practice implementation

project. International Journal Evidence Based Healthcare 8 (2), 90–96.
UC
UC
UC
UC
UC
UC
UC
UC

outcome
measure
Bach, D.M., 1995. Implementation of the Agency for Health Care Policy

Reliable
and Research postoperative pain management guideline. Nursing
Clinics of North America 30 (3), 515–527.

V
V
adequate

Baker, R., Camosso-Stefinovic, J., Gillies, C., Shaw, E.J., Cheater, F., Flottorp,
Statistic
analysis

S., Robertson, N., 2010. Tailored interventions to overcome identified

barriers to change: effects on professional practice and health care
UC
V
X
X
V
X
X
V

outcomes. Cochrane Database Systematic Reviews (3), CD005470,

http://dx.doi.org/10.1002/14651858.CD005470.pub2.
Baumann, B.M., Holmes, J.H., Chansky, M.E., Levey, H., Kulkarni, M.,
outcome
measure

Incomplete
Reliable

Boudreaux, E.D., 2007. Pain assessments and the provision of analge-

outcome

sia: the effects of a templated chart. Academic Emergency Medicine

UC

14 (1), 47–52.
V
V
V
V
V

V
V

data

UC
UC

Berben, S.A., Meijs, T.H., van Grunsven, P.M., Schoonhoven, L., van Achter-
berg, T., 2012. Facilitators and barriers in pain management for
contamination

trauma patients in the chain of emergency care. Injury 43 (9),

1397–1402.
Protection

Bhattacharyya, O., Reeves, S., Garfinkel, S., Zwarenstein, M., 2006. Design-
against

ing theoretically-informed implementation interventions: fine in

theory, but evidence of effectiveness in practice is needed. Imple-
n/a
n/a
n/a

n/a
n/a
n/a
n/a

assessment
V

mentation science 1, 5.
outcome
Blinded

Bookbinder, M., Coyle, N., Kiss, M., Goldstein, M.L., Holritz, K., Thaler, H.,
Gianella, A., Derby, S., Brown, M., Racolin, A., Ho, M.N., Portenoy, R.K.,
assessment

V
V

1996. Implementing national standards for cancer pain management:

outcome
Blinded

program model and evaluation. Journal of Pain and Symptom Man-

agement 12 (6), 334–347.
Bosch, M., van der Weijden, T., Wensing, M., Grol, R., 2007. Tailoring
V
V
V
V
V
V
V
V

quality improvement interventions to identified barriers: a multi-

ple case analysis. Journal of Evaluation in Clinical Practice 13 (2),
Affect data
collections
Characteristics

161–168.
Cabana, M.D., Rand, C.S., Powe, N.R., Wu, A.W., Wilson, M.H., Abboud, P.A.,
of control

Rubin, H.R., 1999. Why don’t physicians follow clinical practice guide-
lines? A framework for improvement. Journal of the American
X
X

Medical Association 282 (15), 1458–1465.

UC
UC
UC
UC

Campbell, P., Dennie, M., Dougherty, K., Iwaskiw, O., Rollo, K., 2004.
X
X
V

Implementation of an ED protocol for pain management at triage

at a busy Level I trauma center. Journal of Emergency Nursing 30 (5),
measure
Baseline

431–438.
Carlson, C.L., 2010. Prior conditions influencing nurses’ decisions to adopt
intervention
UC
UC
UC
UC

UC

evidence-based postoperative pain assessment practices. Pain Man-

V
X

Shape of

agement Nursing 11 (4), 245–258.

Coulter, A., Ellins, J., 2007. Effectiveness of strategies for informing,
educating, and involving patients. British Medical Journal 335
calculation

V
V

(7609), 24–27.
Power

Czarnecki, M.L., Simon, K., Thompson, J.J., Armus, C.L., Hanson, T.C., Berg,
K.A., Petrie, J.L., Xiang, Q., Malin, S., 2011. Barriers to pediatric pain
X
X
X
X
X
X
X
X

management: a nursing perspective. Pain Management Nursing 12

V: done or adequate; X: not done or not adequate; UC: unclear.

(3), 154–162.
Dahl, J.L., Gordon, D., Ward, S., Skemp, M., Wochos, S., Schurr, M., 2003.
ITS methodological quality assessment results and rating
Provider
Provider
Provider
Provider
Provider
Provider
Provider
Provider

independent
Intervention
analysis
Unit of

Institutionalizing pain management: the Post-Operative Pain Man-

agement Quality Improvement Project. Journal of Pain 4 (7), 361–371.
Dalton, J.A., Carlson, J., Blau, W., Lindley, C., Greer, S.M., Youngblood, R.,
2001. Documentation of pain assessment and treatment: how are we
UC
UC

doing? Pain Management Nursing 2 (2), 54–64.

allocation

Davies, P., Walker, A.E., Grimshaw, J.M., 2010. A systematic review of the
Unit of

use of theory in the design of guideline dissemination and imple-

Ward
Ward
Ward
Ward
Ward
Ward
Ward
Ward

mentation strategies and interpretation of the results of rigorous

evaluations. Implementation Science 5, 14.
Dihle, A., Bjolseth, G., Helseth, S., 2006. The gap between saying and doing
Narasimhaswamy et al. (2006)

in postoperative pain management. Journal of Clinical Nursing 15 (4),

469–479.
Jordan-Marsh et al. (2004)

Dulko, D., Hertz, E., Julien, J., Beck, S., Mooney, K., 2010. Implementation of
Treadwell et al. (2002)

cancer pain guidelines by acute care nurse practitioners using an

Michaels et al. (2007)

audit and feedback strategy. Journal of the American Academy of

Oakes et al. (2008)
Innis et al. (2004)
Idell et al. (2007)

Nurse Practitioners 22 (1), 45–55.

O’Connor (2003)

Duncan, K., Pozehl, B., 2001. Effects of individual performance feedback

Phelan (2010)

White (1999)
First author

First author

on nurses’ adherence to pain management clinical guidelines. Out-

(Continued )

comes management for nursing practice 5 (2), 57–62.

Ellis, J.A., McCleary, L., Blouin, R., Dube, K., Rowley, B., MacNeil, M., Cooke,
C., 2007. Implementing best practice pain management in a pediatric
hospital. Journal for Specialists in Pediatric Nursing 12 (4), 264–277.
568 E. Ista et al. / International Journal of Nursing Studies 50 (2013) 552–568

EPOC, 2002. Available <http://epoc.cochrane.org/sites/epoc.cochrane. Johnston, C.C., Gagnon, A., Rennick, J., Rosmus, C., Patenaude, H., Ellis, J.,
org/files/uploads/datacollectionchecklist.pdf>. Ottowa (accessed Shapiro, C., Filion, F., Ritchie, J., Byron, J., 2007. One-on-one coaching
21.02.12). to improve pain assessment and management practices of pediatric
Erdek, M.A., Pronovost, P.J., 2004. Improving assessment and treatment of nurses. Journal of Pediatric Nursing 22 (6), 467–478.
pain in the critically ill. International Journal for Quality in Health Joint Commission on Accreditation of Healthcare Organizations,
Care 16 (1), 59–64. 2001. Comprehensive Hospital Accreditation Manual. Joint Com-
Ferrell, B.R., Wisdom, C., Wenzl, C., 1989. Quality of life as an outcome mission on Accreditation of Healthcare Organizations, Oakbrook
variable in the management of cancer pain. Cancer 63 (11 Suppl.), Terrace, IL.
2321–2327. Jordan-Marsh, M., Hubbard, J., Watson, R., Deon Hall, R., Miller, P., Mohan,
Forsetlund, L., Bjorndal, A., Rashidian, A., Jamtvedt, G., O’Brien, M.A., Wolf, O., 2004. The social ecology of changing pain management: do I have
F., Davis, D., Odgaard-Jensen, J., Oxman, A.D., 2009. Continuing edu- to cry? Journal of Pediatric Nursing 19 (3), 193–203.
cation meetings and workshops: effects on professional practice and Kehlet, H., Holte, K., 2001. Effect of postoperative analgesia on surgical
health care outcomes. Cochrane Database Systematic Reviews (2), outcome. British Journal of Anaesthesia 87 (1), 62–72.
CD003030, http://dx.doi.org/10.1002/14651858.CD003030.pub2. Lugtenberg, M., Burgers, J.S., Clancy, C., Westert, G.P., Schneider, E.C.,
Franck, L.S., Bruce, E., 2009. Putting pain assessment into practice: why is 2011. Current guidelines have limited applicability to patients with
it so painful? Pain Research & Management 14 (1), 13–20. comorbid conditions: a systematic analysis of evidence-based guide-
Franck, L.S., Oulton, K., Nderitu, S., Lim, M., Fang, S., Kaiser, A., 2011. Parent lines. PloS One 6 (10), e25987.
involvement in pain management for NICU infants: a randomized Lugtenberg, M., Zegers-van Schaick, J.M., Westert, G.P., Burgers, J.S., 2009.
controlled trial. Pediatrics 128 (3), 510–518. Why don’t physicians adhere to guideline recommendations in prac-
Gelinas, C., Arbour, C., Michaud, C., Vaillant, F., Desjardins, S., 2011. tice? An analysis of barriers among Dutch general practitioners.
Implementation of the critical-care pain observation tool on pain Implementation Science 4, 54.
assessment/management nursing practices in an intensive care unit Miaskowski, C., Cleary, J., Burney, R., Coyne, P., Finley, R., Foster, R.,
with nonverbal critically ill adults: a before and after study. Inter- Grossman, S., Janjan, N., Syejala, K., Weisman, S., Zahrbock, C.,
national Journal of Nursing Studies 48 (12), 1495–1504. 2005. Guideline for the management of cancer pain in adults and
Gelinas, C., Fortier, M., Viens, C., Fillion, L., Puntillo, K., 2004. Pain assess- children (clinical practice guideline; no. 3) American Pain Society
ment and management in critically ill intubated patients: a retro- (APS), Glenview (IL).
spective study. American Journal of Critical Care 13 (2), 126–135. Michaels, T.K., Hubbartt, E., Carroll, S.A., Hudson-Barr, D., 2007. Evaluating
Gordon, D.B., Dahl, J.L., Miaskowski, C., McCarberg, B., Todd, K.H., Paice, an educational approach to improve pain assessment in hospitalized
J.A., Lipman, A.G., Bookbinder, M., Sanders, S.H., Turk, D.C., Carr, D.B., patients. Journal of Nursing Care Quality 22 (3), 260–265.
2005. American pain society recommendations for improving the Morrison, R.S., Magaziner, J., McLaughlin, M.A., Orosz, G., Silberzweig, S.B.,
quality of acute and cancer pain management: American Pain Society Koval, K.J., Siu, A.L., 2003. The impact of post-operative pain on
Quality of Care Task Force. Archives of Internal Medicine 165 (14), outcomes following hip fracture. Pain 103 (3), 303–311.
1574–1580. Morrison, R.S., Meier, D.E., Fischberg, D., Moore, C., Degenholtz, H., Litke,
Gordon, D.B., Pellino, T.A., Miaskowski, C., McNeill, J.A., Paice, J.A., Laferriere, A., Maroney-Galin, C., Siu, A.L., 2006. Improving the management of
D., Bookbinder, M., 2002. A 10-year review of quality improvement pain in hospitalized adults. Archives of Internal Medicine 166 (9),
monitoring in pain management: recommendations for standardized 1033–1039.
outcome measures. Pain Management Nursing 3 (4), 116–130. Narasimhaswamy, S., Vedi, C., Xavier, Y., Tseng, C.H., Shine, D., 2006. Effect
Grimshaw, J.M., Thomas, R.E., MacLennan, G., Fraser, C., Ramsay, C.R., Vale, of implementing pain management standards. Journal of General
L., Whitty, P., Eccles, M.P., Matowe, L., Shirran, L., Wensing, M., Internal Medicine 21 (7), 689–693.
Dijkstra, R., Donaldson, C., 2004. Effectiveness and efficiency of guide- O’Brien, M.A., Rogers, S., Jamtvedt, G., Oxman, A.D., Odgaard-Jensen, J.,
line dissemination and implementation strategies. Health Technology Kristoffersen, D.T., Forsetlund, L., Bainbridge, D., Freemantle, N., Davis,
Assessment 8 (6) iii–iv, 1–72. D.A., Haynes, R.B., Harvey, E.L., 2007. Educational outreach visits:
Grol, R., Grimshaw, J., 2003. From best evidence to best practice: effects on professional practice and health care outcomes. Cochrane
effective implementation of change in patients’ care. Lancet 362 Database Systematic Reviews (4), CD000409, http://dx.doi.org/
(9391), 1225–1230. 10.1002/14651858.CD000409.pub2.
Grol, R., Wensing, M., 2011. Implementation: Effective Improvement of O’Connor, M., 2003. Pain management: improving documentation of
Patient Care (in Dutch). (Implementatie: Effectieve verbetering van assessment and intensity. Journal for Healthcare Quality 25 (1),
de patiëntenzorg)).Elsevier Gezondheidszorg, Maarssen. 17–21.
Grol, R.P., Bosch, M.C., Hulscher, M.E., Eccles, M.P., Wensing, M., 2007. Oakes, L.L., Anghelescu, D.L., Windsor, K.B., Barnhill, P.D., 2008. An insti-
Planning and studying improvement in patient care: the use of tutional quality improvement initiative for pain management for
theoretical perspectives. Milbank Quarterly 85 (1), 93–138. pediatric cancer inpatients. Journal of Pain and Symptom Manage-
Haller, G., Agoritsas, T., Luthy, C., Piguet, V., Griesser, A.C., Perneger, T., ment 35 (6), 656–669.
2011. Collaborative quality improvement to manage pain in acute Oakes, L.L., Anghelescu, D.L., Windsor, K.B., Barnhill, P.D., Faughnan, L.G.,
care hospitals. Pain medicine 12 (1), 138–147. 2011. An update: institutional quality improvement initiative for pain
Herr, K., Titler, M.G., Schilling, M.L., Marsh, J.L., Xie, X., Ardery, G., Clarke, management for pediatric cancer inpatients. 2007–2010. Journal of
W.R., Everett, L.Q., 2004. Evidence-based assessment of acute pain in Pain and Symptom Management 41 (4), e4–e6.
older adults: current nursing practices and perceived barriers. The Perkins, F.M., Kehlet, H., 2000. Chronic pain as an outcome of surgery. A
Clinical Journal of Pain 20 (5), 331–340. review of predictive factors. Anesthesiology 93 (4), 1123–1133.
Horbury, C., Henderson, A., Bromley, B., 2005. Influences of patient beha- Phelan, C., 2010. An innovative approach to targeting pain in older people
vior on clinical nurses’ pain assessment: implications for continuing in the acute care setting. Contemporary Nurse 35 (2), 221–233.
education. Journal of Continuing Education in Nursing 36 (1), 18–24. Smedley, B.D., Syme, L., 2000. Promoting Health: Intervention Strategies
Idell, C.S., Grant, M., Kirk, C., 2007. Alignment of pain reassessment from Social and Behavioral Research. National Academy Press,
practices and National Comprehensive Cancer Network guidelines. Washington, DC.
Oncology Nursing Forum 34 (3), 661–671. Treadwell, M.J., Franck, L.S., Vichinsky, E., 2002. Using quality improve-
Innis, J., Bikaunieks, N., Petryshen, P., Zellermeyer, V., Ciccarelli, L., 2004. ment strategies to enhance pediatric pain assessment. International
Patient satisfaction and pain management: an educational approach. Journal for Quality in Health Care 14 (1), 39–47.
Journal of Nursing Care Quality 19 (4), 322–327. Van Niekerk, L.M., Martin, F., 2001. Tasmanian nurses’ knowledge of
Jakobsson, U., Hallberg, I.R., Westergren, A., 2007. Exploring determinants pain management. International Journal of Nursing Studies 38 (2),
for quality of life among older people in pain and in need of help for 141–152.
daily living. Journal of Clinical Nursing 16 (3A), 95–104. White, C.L., 1999. Changing pain management practice and impacting on
Jamtvedt, G., Young, J.M., Kristoffersen, D.T., O’Brien, M.A., Oxman, A.D., patient outcomes. Clinical Nurse Specialist 13 (4), 166–172.
2009. Audit and feedback: effects on professional practice and health Yildirim, Y.K., Cicek, F., Uyar, M., 2008. Knowledge and attitudes of Turkish
care outcomes. Cochrane Database Systematic Reviews (2), oncology nurses about cancer pain management. Pain Management
CD000259, http://dx.doi.org/10.1002/14651858.CD000259.pub2. Nursing 9 (1), 17–25.