Accepted Manuscript

Nudging to Improve Hand Hygiene

M.G. Caris, H.A. Labuschagne, M. Dekker, M.H.H. Kramer, M.A. van Agtmael,
C.M.J.E. Vandenbroucke-Grauls

PII: S0195-6701(17)30532-7
DOI: 10.1016/j.jhin.2017.09.023
Reference: YJHIN 5237

To appear in: Journal of Hospital Infection

Received Date: 3 August 2017

Accepted Date: 24 September 2017

Please cite this article as: Caris MG, Labuschagne HA, Dekker M, Kramer MHH, van Agtmael MA,
Vandenbroucke-Grauls CMJE, Nudging to Improve Hand Hygiene, Journal of Hospital Infection (2017),
doi: 10.1016/j.jhin.2017.09.023.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Nudging to Improve Hand Hygiene
M.G. Caris1,2; H.A. Labuschagne1; M. Dekker2; M.H.H. Kramer1; M.A. van Agtmael1; C.M.J.E.
Vandenbroucke-Grauls2

1
Department of Internal Medicine
2
Department of Medical Microbiology & Infection Control

VU University Medical Center

PT
De Boelelaan 1117
1081 HV Amsterdam, The Netherlands

RI
Corresponding author
M.G. Caris, MD

SC
Department of Internal Medicine
VU University Medical Center
De Boelelaan 1117

U
1081 HV Amsterdam, The Netherlands
AN
Email: [email protected]
Phone: +31 20 444 45 88
M

Running title
Nudging hand hygiene
D

Key words
TE

Nudging, cognitive bias, hand hygiene

C EP
AC

1
 ACCEPTED MANUSCRIPT
Summary
Background Hand hygiene is paramount to prevent healthcare-associated infections, but improving
compliance is challenging. When healthcare workers seldom encounter healthcare-associated
infections, they will consider the odds of causing infections through poor hand hygiene negligible.
Cognitive biases, such as these, can induce non-compliance. Nudging, ‘a friendly push to encourage
desired behaviour’, could provide an easily implemented, inexpensive measure to address cognitive
biases and thus support hand hygiene interventions.

PT
Aim To investigate whether behavioural nudges, displayed as posters, can increase the use of
alcohol-based hand rub.

RI
Methods We developed nudges based on a systematic review of previously described cognitive
biases, and tested these through a cross-sectional survey among the target audience. We then
conducted a controlled before-after trial on two hospital wards, to assess the effect of these nudges

SC
on the use of alcohol based hand rub, measured with electronic dispensers.

Findings Poisson regression analyses adjusted for workload showed that nudges displayed next to

U
dispensers increased their overall use on one ward (poster 1: RR 1.6 (95%CI 1.2-2.2), poster 2: RR 1.7
(95%CI 1.2-2.5)) and during doctor’s rounds on both wards (poster 1: ward A RR 1.7 (95%CI 1.1-2.6),
AN
ward B RR 2.2 (95%CI 1.3-3.8)). Use of dispensers without adjacent nudges did not increase.

Conclusion Nudges based on cognitive biases that play a role in hand hygiene, and displayed as
M

posters, could provide an easy, inexpensive measure to increase use of alcohol-based hand rub.
When applying nudges to change behaviour, it is important to identify the right nudge for the right
D

audience.
TE
C EP
AC

2
 ACCEPTED MANUSCRIPT

Introduction
Healthcare-associated infections pose a threat to patient safety. Although hand hygiene is the most
effective strategy to prevent these infections,1,2 adherence to hand hygiene guidelines is poor.3 To
improve adherence, many methods have been applied, but structural improvement is challenging.4
Measures to support the effect of interventions are therefore essential.

PT
A possibly effective strategy, not yet commonly applied in healthcare, is nudging: a friendly push to
encourage desired behaviour. Nudging is widely used to stimulate healthy behaviour, such as
smoking cessation5 and healthy food choices.6 Nudging of healthcare workers has only recently been

RI
addressed in a few studies, in which various kinds of nudges proved successful in influencing
physician prescribing behaviour.7-10 Although the exact mechanism has not been clarified, it has been

SC
suggested that nudges influence behaviour by addressing cognitive biases underlying non-
compliance.11 Cognitive biases are often caused by "heuristics". In social sciences, heuristics are
defined as decisional short-cuts which can be useful, as they help us to quickly make judgements or
decisions. However, when heuristics lead to cognitive biases, they can confound our judgement and

U
thus influence our behaviour.
AN
In medical decision making, several cognitive biases and underlying heuristics have been proposed to
play a role.12,13 One example is the ‘availability heuristic’, which describes how we assess the
probability of an event by the ease with which we can recall it: the easier an event is to recall, the
M

more likely we consider it to occur. Conversely, this means that if prevalence of healthcare-
associated infections is low, healthcare workers are inclined to believe that chances of transmitting
D

an infection through lack of hand hygiene are also low, thus leading to non-compliance.14 Heuristics
such as these also play a role in influenza vaccination among healthcare workers, and nudges
TE

specifically designed to target these heuristics were shown to increase vaccine uptake.11 Although
evidence is scarce, research indicates that behavioural nudges could also be useful to improve hand
hygiene, both in school children15 and healthcare workers.16
EP

Nudging could provide an easily implemented, inexpensive measure to support hand hygiene
interventions. In this study, we reviewed heuristics and cognitive biases that could play a role in hand
C

hygiene non-compliance. We then developed a set of nudges addressing these biases and assessed
their effect on the use of alcohol-based hand rub in our hospital.
AC

3
 ACCEPTED MANUSCRIPT
Methods
Design and setting
In our university medical centre, we performed a mixed-methods study, combining a cross-sectional
survey with a controlled before-after trial. We systematically reviewed literature on possible
cognitive biases underlying hand hygiene non-compliance and assessed how these could be
addressed using images and slogans. Based on a survey among the target audience, a graphic
designer devised posters with the two highest scoring slogans and images. We displayed these
posters on two hospital wards and assessed their effect on the use of alcohol-based hand rub,

PT
measured with electronic dispensers.

Nudges

RI
Previous studies have proposed 43 types of cognitive biases and heuristics that could play a role in
medical decision making.11-13 We evaluated these types for their relevance to hand hygiene. Most of
these were considered irrelevant: 27 applied to pattern recognition in diagnosing patients, and two

SC
others applied to the order in which choices are presented, which does not apply to the yes/no
decision to perform hand hygiene. In addition, we systematically searched PubMed, CINAHL and
PsycINFO for studies on cognitive biases in infection control, and in hand hygiene specifically. This

U
revealed no additional types to consider. Table I shows an overview of the 14 types that could be
relevant in decisions on hand hygiene, and could therefore be important to nudge correct hand
AN
hygiene behaviour. These were used as input for the images and slogans (in Supplementary
material).
M

We sent a survey to a convenience sample of 110 students, 60 nurses from all clinical wards and 27
doctors, mostly internal medicine. The survey was distributed through SurveyMonkey
D

(SurveyMonkey Inc., San Mateo, California, USA). Respondents were presented with 10 images and
eight slogans and were asked open-ended questions on what these evoked.
TE

Two researchers (MC and HL) independently evaluated responses to the images and slogans, to
assess their value as a nudge. In order to be effective in changing behaviour, a nudge needs to be a
reminder of the sought after activity, therefore the image or slogan needed to evoke a link with hand
EP

hygiene. In addition, nudges need to evoke a reaction. Nudges generating a positive attitude are
more effective, merely negative attitudes can be counterproductive. Responses to images were
assigned points for their link with hand hygiene, and use of positive and negative adjectives (table II),
C

with a maximum of five. Responses to slogans were only assigned points for use of adjectives, with a
AC

maximum of three.

Reasons for poor hand hygiene have been shown to be different for different healthcare workers.14,18
In addition, factors such as knowledge and risk perception have been described to influence hand
hygiene behaviour of healthcare personnel, and could therefore affect the response to the images
and slogans. We used the questionnaire by Erasmus et al17to assess these factors (translated
questionnaire in supplementary material). With linear regression, we assessed scores of the slogans
and images, adjusting for overall score on the hand hygiene questionnaire and for the respondent's
profession (i.e. nurses or otherwise), using SPSS 22.0 for Windows (IBM SPSS Inc., Chicago, IL, USA).

4
 ACCEPTED MANUSCRIPT
Alcohol-based hand rub
Two hospital wards were selected based on the location of their dispensers, as to not interfere with
their current use, and on their similar characteristics: both were clinical, non-ICU wards with adult
patients of both medical and surgical specialities, and had 24 beds. Although one ward was an acute
admissions ward, both had medium-high turnover of patients (average length of stay 6-9 days), and
similar average hand hygiene compliance rates (50-60% with overt observation, based on the WHO’s
5 moments for hand hygiene).

PT
Three electronic alcohol based hand rub dispensers (Ophardt Hygiene Monitoring System (OHMS),
Ophardt Hygiene-Technik GmbH & Co.KG, Issum, Germany) were installed on each ward: one at the
entrance, the other two in different (shared) patient rooms. The dispensers were installed at the

RI
same location as the previous, non-electronic dispensers and looked and functioned the same. These
dispensers recorded the exact time they were activated and the number of times the handle was
consecutively pressed; consecutive uses were recorded as one activation. After a two-week baseline

SC
measurement, the posters were sequentially displayed at the wards’ entrance, for two weeks each.
The sequence for each ward was assigned randomly. We used a patient-based approach to measure
workload (i.e. nurse-patient ratio), defined as number of FTE (full-time equivalent) per admission

U
day, which we obtained from our hospital’s business intelligence department.
AN
We analysed changes in the number of activations through Poisson regression analyses, adjusting for
workload and correcting for correlated observations through stratification per dispenser, with
STATA/SE 14.1 for Windows (StataCorp LP, College Station, Texas, USA).
M

Results
D

Nudges
Overall survey response rate was 26%; responders were 19 medical students, 15 physicians (1
TE

haematology, 1 infectious diseases, 1 psychiatry, 12 internal medicine), and 18 nurses (A&E,

gastroenterology, ICU, internal medicine, nephrology, neurology, OBGYN, paediatrics, psychiatry,
pulmonology, surgery). The two highest-scoring images had mean scores of 3.1 and 3.0. The two
EP

highest-scoring slogans had mean scores of 2.6 and 2.4. These scores were not influenced by
profession, or by score on the hand hygiene questionnaire. The best scoring image was specifically
mentioned in response to the best scoring slogan, as was the second best scoring image in response
C

to the second best scoring slogan. These two were therefore paired on the posters (figure 1).
AC

Alcohol-based hand rub

The six dispensers together were used 2842 times. Both posters increased the overall use of the
entrance dispenser on ward B (poster 1: relative risk (RR) 1.61 95% confidence interval (CI) 1.18-2.20,
poster 2: RR 1.72 95% CI 1.17-2.53) but not of the entrance dispenser on ward A (poster 1: RR 1.04
95% CI 0.79-1.38, poster 2: RR 1.07 95% CI 0.80-1.42). There was no effect of the entrance posters on
the use of dispensers inside patient rooms. During the intervention periods, use of the entrance
dispensers showed peaks between 9 and 11 (figure 1), coinciding with the doctors’ morning rounds.
On ward B, a second peak between 16 and 18 coincided with a ward-specific afternoon doctors’
round.

5
 ACCEPTED MANUSCRIPT
Discussion
Our study indicates that nudges, based on cognitive biases and displayed as posters, can increase the
use of alcohol-based hand rub when shown next to dispensers. One of the nudges had an effect on
both wards, increasing the use of adjacent dispensers during specific times of the day, while the
other nudge increased overall use of the entrance dispenser on one ward. At the same time, there
was no change in use of the dispensers where no posters were displayed.

Although our sample is small, these findings are in line with the few studies that have used nudges to

PT
influence healthcare workers’ behaviour. Meeker and colleagues designed a nudge that appealed to
the power of public commitment, by asking general practitioners to sign a letter of commitment to
appropriate use of antibiotics. Prominent display of this letter in their examination room led to a 20%

RI
absolute reduction in inappropriate prescribing.7 Tannenbaum and colleagues nudged appropriate
prescribing by rearranging treatment options in an electronic health record, which led to a 12%
absolute reduction in aggressive treatment.10 The same approach was shown effective in improving

SC
ventilation19 and appropriate prescribing of mouthwash9 in ICU-patients. In a study specifically
addressing hand hygiene, nudges were shown to increase hand washing among school children from
4% to 74%.15 These studies all show how by targeting certain aspects of ‘choice architecture’,

U
seemingly small interventions can influence behaviour.
AN
Other studies have shown that posters solely serving as a reminder or educational message are not
effective,20 and that message framing on hand hygiene posters is crucial to achieve an effect.21 Dubov
and Phung specifically addressed cognitive biases underlying low rates of influenza vaccination
M

among healthcare workers, and showed how nudges targeting these biases can increase vaccine
uptake.11 This strengthens the idea that the effect of nudges possibly relies on addressing heuristics
D

and cognitive biases underlying non-compliance.

TE

We displayed two different nudges, appealing to different cognitive biases, and these had different
effects. Poster 1 (“which category do you belong to”) appealed to the bandwagon effect, and only
showed an effect on ward B. This could be due to differences between the wards we did not account
for, such as ward culture regarding hand hygiene or the presence or absence of role models.18 As
EP

previously described, the acceptability of a nudge depends on “the nature of the nudge, who nudges,
and who is nudged”.22 Further refined nudges, based on cognitive biases at the ward level, could
perhaps enhance acceptability. Poster 2 (“more hand hygiene, less infections”) appealed to the loss
C

aversion and relative risk biases, and had an effect on both wards. However, use mainly increased
AC

during doctor’s rounds, which could indicate that the nudge appealed more to doctors. Reasons for
poor hand hygiene have been shown to be different for doctors and nurses.14,18 Future studies, in
which nudges are designed to specifically address doctors or nurses, or in which the user’s profession
is registered along with the use of a dispenser, could shed light on this hypothesis.

Our study has strengths and limitations. We performed a thorough literature review on cognitive
biases underlying non-compliance and assessed these for their applicability to hand hygiene; this
overview can be of use to other interventions. In addition, the nudges were designed with input from
the expected end-users (i.e. nurses, doctors, and students), which has been proposed as more
effective than using ready-made posters.23 We selected the highest-scoring images and slogans for
our nudges. Scores for the slogans were high, but scores for the images were moderate. Perhaps

6
 ACCEPTED MANUSCRIPT
other, higher-scoring images could have had more effect. Although we evaluated the nudges among
the target audience and the responding nurses worked on many different units, almost all
responding doctors worked in internal medicine. In addition, response rates were low. Surveys
among a larger sample with a more diverse background, could strengthen our results. With the
electronic dispensers, data on their use was collected remotely, therefore study participants were
not aware of being monitored, eliminating response bias. It should be noted that we only collected
data on use of the dispensers, not on the appropriateness of alcohol-based hand rub use. We
therefore cannot draw conclusions on hand hygiene compliance. We only had six electronic

PT
dispensers at our disposal, for a limited time of six weeks. More dispensers with adjacent posters and
a wash-out period between posters would have strengthened our results. Unfortunately, this was
not possible. However, with this small, explorative study, we did see an interesting effect that

RI
warrants further research. The effect of changes in the environment can be subject to the novelty
effect; studies with longer follow-up are needed to assess whether the effect of nudges is sustained.

SC
There is, of course, no easy solution to increase hand hygiene compliance, and multi-faceted
intervention strategies are preferred. Nudging should not be considered a stand-alone intervention,
but rather a supportive measure to optimize intervention strategies. Our study indicates that

U
displaying nudges next to dispensers could be a useful, inexpensive measure to enhance the use of
alcohol-based hand rub.
AN
M
D
TE
C EP
AC

7
 ACCEPTED MANUSCRIPT
References

1 Pittet D, Allegranzi B, Sax H, Dharan S, Pessoa-Silva CL, Donaldson L, et al. Evidence-based

model for hand transmission during patient care and the role of improved practices. The
Lancet Infectious Diseases 2006; 6: 641-52.
2 Kirkland KB, Homa KA, Lasky RA, Ptak JA, Taylor EA, Splaine ME. Impact of a hospital-wide
hand hygiene initiative on healthcare-associated infections: results of an interrupted time
series. BMJ Qual Saf 2012; 21: 1019-26.
3 Erasmus V, Daha TJ, Brug H, Richardus JH, Behrendt MD, Vos MC, et al. Systematic review of

PT
studies on compliance with hand hygiene guidelines in hospital care. Infect Control Hosp
Epidemiol 2010; 31: 283-94.
4 Huis A, Van Achterberg T, De Bruin M, Grol R, Schoonhoven L, Hulscher M. A systematic

RI
review of hand hygiene improvement strategies: a behavioural approach. Implement Sci
2012; 7: 92.
5 Sunstein CR. Nudging smokers. N Engl J Med 2015; 372: 2150-1.

SC
6 Bucher T, Collins C, Rollo ME, McCaffrey TA, De Vlieger N, Van der Bend D, et al. Nudging
consumers towards healthier choices: a systematic review of positional influences on food
choice. Br J Nutr 2016; 115: 2252-63.
7 Meeker D, Knight TK, Friedberg MW, Linder JA, Goldstein NJ, Fox CR, et al. Nudging

U
guideline-concordant antibiotic prescribing: a randomized clinical trial. JAMA Intern Med
2014; 174: 425-31.
AN
8 Katchanov J, Kluge S, Mackenzie CR, Kaasch AJ. "Nudging" in microbiological reports: a
strategy to improve prescribing. Infection 2016.
9 Bourdeaux CP, Davies KJ, Thomas MJ, Bewley JS, Gould TH. Using 'nudge' principles for order
M

set design: a before and after evaluation of an electronic prescribing template in critical care.
BMJ Qual Saf 2014; 23: 382-8.
10 Tannenbaum D, Doctor JN, Persell SD, Friedberg MW, Meeker D, Friesema EM, et al.
Nudging physician prescription decisions by partitioning the order set: results of a vignette-
D

based study. J Gen Intern Med 2015; 30: 298-304.

11 Dubov A, Phung C. Nudges or mandates? The ethics of mandatory flu vaccination. Vaccine
TE

2015; 33: 2530-5.

12 Blumenthal-Barby JS, Krieger H. Cognitive biases and heuristics in medical decision making: a
critical review using a systematic search strategy. Med Decis Making 2015; 35: 539-57.
EP

13 Croskerry P. Achieving quality in clinical decision making: cognitive strategies and detection
of bias. Acad Emerg Med 2002; 9: 1184-204.
14 Erasmus V, Brouwer W, Van Beeck EF, Oenema A, Daha TJ, Richardus JH, et al. A qualitative
exploration of reasons for poor hand hygiene among hospital workers: lack of positive role
C

models and of convincing evidence that hand hygiene prevents cross-infection. Infect Control
Hosp Epidemiol 2009; 30: 415-9.
AC

15 Dreibelbis R, Kroeger A, Hossain K, Venkatesh M, Ram PK. Behavior Change without Behavior
Change Communication: Nudging Handwashing among Primary School Students in
Bangladesh. Int J Environ Res Public Health 2016; 13.
16 Kwok YL, Juergens CP, Mclaws ML. Automated hand hygiene auditing with and without an
intervention. Am J Infect Control 2016; 44: 1475-80.
17 Erasmus VE. Compliance to Hand Hygiene Guidelines in Hospital Care - A stepwise
behavioural approach. Rotterdam, The Netherlands: Erasmus MC; 2012.
18 Jang JH, Wu S, Kirzner D, Moore C, Youssef G, Tong A, et al. Focus group study of hand
hygiene practice among healthcare workers in a teaching hospital in Toronto, Canada. Infect
Control Hosp Epidemiol 2010; 31: 144-50.

8
 ACCEPTED MANUSCRIPT
19 Bourdeaux CP, Thomas MJ, Gould TH Malhotra G, Jarvstad A, Jones T, et al. Increasing
compliance with low tidal volume ventilation in the ICU with two nudge-based interventions:
evaluation through intervention time-series analyses. BMJ Open 2016; 6: e010129.
20 Morse L, Mcdonald M. Failure of a poster-based educational programme to improve
compliance with peripheral venous catheter care in a tertiary hospital. A clinical audit.
Journal of Hospital Infection 2009; 72: 221-6.
21 Jenner EA, Jones F, Fletcher B, Miller L, Scott GM. Hand hygiene posters: selling the message.
Journal of Hospital Infection 2005; 59: 77-82.
22 Lucke J. Context is all important in investigating attitudes: acceptability depends on the

PT
nature of the nudge, who nudges, and who is nudged. Am J Bioeth 2013; 13: 24-5.
23 Seto WH, Yuen SW, Cheung CW, Ching PT, Cowling BJ, Pittet D. Hand hygiene promotion and
the participation of infection control link nurses: an effective innovation to overcome
campaign fatigue. Am J Infect Control 2013; 41: 1281-3.

RI
24 Tversky AK, J. Anomalies: the endowment effect, loss aversion, and status quo bias. Choices,
Values, and Frames. New York: Russell Sage Foundation, Cambridge University Press; 2000. p.
159-70.

U SC
AN
M
D
TE
C EP
AC

9
 ACCEPTED MANUSCRIPT
Source of funding
This study was funded by the board of directors of the VU University Medical Center, through the
annual Innovation Impulse Programme. Ophardt provided six electronic dispensers. Members of the
board of directors who provided the funding and Ophardt had no involvement in the conduct of the
research, preparation of the article, in study design, in the collection, analysis and interpretation of
data, or in the decision to submit the article for publication.

PT
RI
U SC
AN
M
D
TE
C EP
AC

10
 ACCEPTED MANUSCRIPT
Tables

Table I. Overview of heuristics and cognitive biases that could play a role in hand hygiene behaviour

Heuristic/cognitive bias Meaning Relevance to hand hygiene

12
Affect heuristic Behaviour is influenced by the emotional response Hand hygiene can be associated with
(the affect) a stimulus evokes, which influences the both positive affect (i.e. being clean,
way risks and benefits are perceived. If an activity or providing safe care) and negative
affect (i.e. it is a nuisance, thought of

PT
stimulus evokes positive feelings, risks are perceived
as unnecessary, or it irritates the skin).
as low while benefits are considered high, and vice
versa.

RI
12
Impact bias People tend to overestimate the long-term impact of “If we always perform hand hygiene,
events. there will be no more time left for
actual care.”

SC
11,12
Optimism bias People have a tendency to believe that particular “We never see problems with
health risks are greater for other people than healthcare-associated infections on
themselves. our ward.”

U
11
Present bias People have a tendency to value costs and benefits Short-term risk of more work and skin
AN
today higher than costs and benefits in the future. irritation can weigh more heavily than
reducing healthcare-associated
infections.
11-13
Availability heuristic People assess the probability of an event by the ease “Hand hygiene compliance is low and
M

with which they can recall such an episode. The we just had an outbreak.”
occurrence is considered more likely if an episode is
easy to recall. “I don’t always wash my hands and I
D

have never caused an infection.”

12,13
Confirmation bias People have a tendency to see evidence in favour of See availability heuristic.
TE

their beliefs and to ignore evidence for the opposite.

Default bias or status Disadvantages of leaving the status quo outweigh Increasing hand hygiene performance
EP

12 24 requires leaving the status quo.

quo bias the corresponding advantages.

12
Bandwagon effect The probability of an individual adopting an activity If everyone cleans their hands, more
or behaviour increases when more people have people will and vice versa.
C

already done so.

AC

11,12
Ambiguity aversion The tendency to prefer a known risk to an unknown Balancing risks of skin irritation (easily
risk. visible) and healthcare-associated
infection (hard to see).
12,13
Commission bias People have the “tendency toward action rather Performing hand hygiene is an action.
13
than inaction”.
Loss/gain framing bias Avoiding losses is a more powerful driver of “If I always perform hand hygiene, I
12,21
or loss aversion bias decisions than the corresponding gains. will get dry skin. Besides, infections
can be easily treated with antibiotics.”
11-13
Omission bias A potentially harmful inaction (omission) is usually Omitting hand hygiene to avoid skin
preferred to a possibly less harmful action. irritation.

11
 ACCEPTED MANUSCRIPT
Frequency/percentage Risks are perceived as higher when displayed as Rates of healthcare-associated
12
framing effect frequencies. infections can be expressed as
frequencies.
12
Relative risk bias People are more likely to choose an activity when The effect of proper hand hygiene on
presented with relative risk than when the same reducing healthcare-associated
information is described as an absolute risk. infections can be expressed as relative
risk reduction instead of absolute risk
reduction.

PT
RI
U SC
AN
M
D
TE
C EP
AC

12
 ACCEPTED MANUSCRIPT
Table II. Categorization of responses to images and slogans in the online survey

Link with hand hygiene Reaction Points

No link Negative 0
no use of any words related to hand hygiene use of adjectives such as annoying, exaggerated

Indirectly linked Neutral 1

use of words such as contagious, transfer of use of neither positive nor negative adjectives
bacteria

PT
Directly linked Positive and negative 2
use of words such as hand hygiene, hand use of both positive and negative adjectives
washing “Somewhat annoying, somewhat funny”

RI
Positive 3
use of adjectives such as good, nice

U SC
AN
M
D
TE
C EP
AC

13
 ACCEPTED MANUSCRIPT
Figures

Figure 1.

PT
RI
U SC
AN
M
D
TE
C EP
AC

14
 ACCEPTED MANUSCRIPT
Figure 2.

PT
RI
U SC
AN
M
D
TE
C EP
AC

15
 ACCEPTED MANUSCRIPT
Figure legends

Legend figure 1. Nudges, developed based on survey responses. Poster 1 (left): “Half of all healthcare
workers perform well in hand hygiene. Which category do you belong to?” Poster 2: “40% increase in
hand hygiene, 40% decrease in healthcare-associated infections”

PT
RI
U SC
AN
M
D
TE
C EP
AC

16
 ACCEPTED MANUSCRIPT
Legend figure 2. Number of activations of the entrance dispensers on ward A en B per hour of the
day, during baseline and during display of posters 1 (“Which category do you belong to?”) and 2
(“More hand hygiene, less infections”). Symbols indicate the number of activations during doctor’s
rounds, during display of the poster compared to baseline (results of Poisson regression analyses,
adjusted for workload, RR Relative Risk CI Confidence Interval): *RR 2.09 95% CI 1.31-3.33 ●RR 1.67
95% CI 1.09-2.55 ●●RR 1.49 95% CI 0.91-2.42 ●●●RR 2.20 95% CI 1.28-3.78

PT
RI
U SC
AN
M
D
TE
C EP
AC

17