Preventing the use of physical restraints on residents
newly admitted to psycho-geriatric nursing home wards:
A cluster-randomized trial
Anna R. Huizinga, Jan P.H. Hamersa,*, Math J.M. Gulpersb,
Martijn P.F. Bergerc
aDepartment of Health Care Studies, Section of Nursing Science, Maastricht University, The Netherlands
bVerpleeghuis Lu ¨ckerheide (Nursing Home), MeanderGroep Zuid-Limburg, The Netherlands
cDepartment of Methodology and Statistics, Maastricht University, The Netherlands
Received 5 April 2007; received in revised form 25 March 2008; accepted 28 March 2008
Background: Although there is an urgent need for restraint-free care, the number of randomized clinical trials on preventing or
reducing physical restraints has been limited.
Objectives: To investigate the effectiveness of an educational intervention to prevent the use of physical restraints on residents
newly admitted to psycho-geriatric nursing home wards.
Design: Cluster-randomized trial.
Setting: Fourteen Dutch psycho-geriatric nursing home wards.
Participants: 138 residents admitted to 14 psycho-geriatric nursing home wards after baseline measurement of the trial were
selected, out of which 33 residents died or informed consent had not been obtained. A total of 105 residents were included in the
Methods: The nursing homewards were randomly assigned to either educational interventionor control status. The educational
intervention consisted of an educational programme for nurses combined with a nurse specialist’s consultation. The data were
collected at 1, 4 and 8 months post-intervention. The use of physical restraints was measured by blinded, trained observers on
regarding restraint status, restraint intensity, multiple restraints and types of restraints were found. One month post-
intervention, 38% of the residents newly admitted to the experimental wards were restrained. Bilateral bedrails were the
most frequently used restraints at Post-test 1 (24%), Post-test 2 (23%) and Post-test 3 (28%), followed by the use of infrared
systems at Post-tests 2 and 3.
Conclusion: An educational programme combined with the consultation of a nurse specialist does not prevent the use of
physical restraints on residents newly admitted to psycho-geriatric nursing home wards. Although other studies have shown
Available online at www.sciencedirect.com
International Journal of Nursing Studies 46 (2009) 459–469
* Corresponding author at: Department of Health Care Studies, Section of Nursing Science, Maastricht University, P.O. Box 616, 6200 MD
Maastricht, The Netherlands. Tel.: +31 43 3881549; fax: +31 43 3884162.
E-mail address: email@example.com (J.P.H. Hamers).
0020-7489/$ – see front matter # 2008 Elsevier Ltd. All rights reserved.
promising results with the effectiveness of these types of intervention on restraint reduction, the development of additional
interventions to prevent restraint usage is recommended.
# 2008 Elsevier Ltd. All rights reserved.
Keywords: Cluster-randomized trial; Consultation; Education; Resident admission; Psycho-geriatrics; Physical restraint
What is already known about the topic?
? Physical restraints are commonly used with psycho-ger-
iatric nursing home residents and restraints are used as a
routine procedure in most cases.
? Several studies have shown a need for restraint-free care
in nursing homes.
? The number of randomized controlled trails (RCTs)
investigating the effect of restraint reduction programmes
is limited and has shown varying results.
What this paper adds
? An educational programme for nurses combined with the
consultation of a nurse specialist cannot prevent the use of
bilateral bedrails and other restraint types on residents
newly admitted to psycho-geriatric nursing home wards.
? This study confirms the general assumptions concerning a
routine usage of physical restrains on psycho-geriatric
nursing home residents.
? In order to achieve restraint-free care for psycho-geriatric
nursing home residents, more effort and interventions, in
addition to education and consultation, are needed, such
as analysing the attitudes and knowledge of the (nursing)
staff population, the residents and their representatives,
and stimulating the mobilization of residents.
Physical restraints are defined as any limitation on an
individual’s freedom of movement by using measures or
equipments such as belts tied to a chair or bed, bilateral
bedrails andchairswithtables (Hantikainen, 1998). Physical
restraints are commonly used on psycho-geriatric nursing
home residents. High prevalence values have been reported
in several studies, ranging from 15% to 66% (Capezuti et al.,
2007; Hamers et al., 2004; Huizing et al., 2006). In nursing
homes, physical restraints are in most cases used for more
than 3 months and as a routine measure (Hamers et al.,
2004). The use of physical restraints is based mainly on the
assumption that restraints prevent falls and fall-related
injuries (Evans and FitzGerald, 2002; Evans and Strumpf,
1990; Hamers et al., 2004). Furthermore, residents’ char-
acteristics, such as impairment in mobility and cognitive
status, are highly associated with restraint use with nursing
home residents (Hamers et al., 2004; Sullivan-Marx et al.,
1999). There are also indications that nurses’ attitude influ-
ences the use of physical restraints (Karlsson et al., 2001;
Suen et al., 2006; Werner and Mendelsson, 2001). Various
studies, however, conclude that the use of physical restraints
is ineffective in preventing falls and fall-related injuries and
may also lead to negative physical, psychological and social
consequences, suchaspressure ulcersanddepression (Cape-
zuti et al., 2007; Evans et al., 2003; Hamers and Huizing,
2005). The need to reduce physical restraint use in nursing
homes has, therefore, often been recommended in the lit-
erature (Hamers and Huizing, 2005; Strumpf et al., 1998).
The transition to restraint-free care involves a process of
change from viewing behaviourof residents, such as fall risk
and wandering, as a problem to be controlled with physical
restraints to viewing behaviour of residents as a commu-
nication of health-state change or unmet need (Strumpf
et al., 1998). This process of change can be complex and
slow (Strumpf et al., 1998), just like other processes of
change in healthcare (Grol et al., 2005; Rogers, 2003). So
far, a limited number of RCT investigations has examined
theeffectiveness of programmes onthe reduction of restraint
use (Evans et al., 1997; Huizing et al., 2006; Testad et al.,
2005). Comparable interventions, mostly consisting of edu-
cation and consultation, were tested in these studies.
Restraint education for (nursing) staff, usually combined
with support and advice in practice, was an important
component in these interventions. Although Huizing et al.
(2006) showed no effect on restraint reduction, two of the
RCT studies reported positive effects in reducing the use of
physical restraints (Evans et al., 1997; Testad et al., 2005).
Testad et al. (2005) have shown that an education program
nursing homes. Furthermore, Evanset al. (1997)have shown
that restraint education combined with specialist nurse
consultation is effective in safely reducing the use of phy-
sical restraints. However, it is unclear whether education and
consultation prevent the use of physical restraints on newly
admitted nursing home residents. Preventing the use of
restraints on these residents may be a first step in the
transition to restraint-free care, given the new situation
where nurses can implement individualized care and break
with routines concerning restraint use. Research on this
specific population is therefore needed to further understand
the transition process to restraint-free care.
The aim of this study was to determine whether an
educational intervention, consisting of an educational pro-
specialist, has a preventive effect on the use of physical
restraints with residents newly admitted to psycho-geriatric
nursing home wards.
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469460
2.1. Design and sample
A cluster-randomized trial was performed to examine the
preventive effect of educational intervention on the use of
physical restraints with residents newly admitted to psycho-
geriatric nursing home wards. The study was part of a larger
study focusing on the use of physical restraints on psycho-
geriatric nursing home residents (see also Huizing et al.,
2006). This larger study focussed on the effects of the
educational intervention on the reduction of physical
restraints on residents already residing in the nursing home
wards. The present study examines the preventive effects of
the educational intervention on the use of physical restraints
on residents newly admitted to 14 Dutch psycho-geriatric
nursing home wards after the baseline measurement. These
residents were enrolled at each subsequent measurement.
not included in the sample. The 14 psycho-geriatric nursing
home wards were randomly assigned to educational inter-
vention (seven experimental wards) or control status (seven
educational programme on the use of physical restraints.
Further, consultation with a registered nurse specialized in
the use of physical restraints and in their reduction (nurse
specialist) was introduced in the experimental wards. There
was no educational intervention in the control group and
residents received care as usual. Data were collected at 1, 4
test 3). In this study the term nurses is defined as registered
nurses, care workers, care helpers and care assistants
(MinVWS, 1997). The qualifications within the nursing
profession decrease in complexity from registered nurses
performing the most complex activities to care helpers and
care assistants doing mainly routine and standard procedural
work (MinVWS, 1997). In the Netherlands, most of the
nurses working in nursing homes are qualified as care
workers (Van der Windt et al., 2003).
2.2. The intervention
An educational programme, combined with the consul-
tation of a nurse specialist, made up the components of the
educational intervention (see Table 1) (see also Huizing
et al., 2006). The educational intervention was designed to
encourage nurses to adopt a philosophy of restraint-free
care and to familiarize themselves with the techniques of
individualized care (Strumpf et al., 1998). As can be seen
from Table 1, the educational programme was carried out
over a 2-month period and taught by the nurse specialist. In
five meetings, each lasting 2 h, several topics concerning
are the decision-making process for restraint use, the inef-
for analysing the risk behaviour of residents and the alter-
natives to restraint use. Furthermore, the nurses were
invited to discuss real-life cases during the educational
meetings, which provided supporting evidence about how
they combined their practical experience with the informa-
tion from the educational programme. The educational
programme consisted of small-scale meetings with an
active learning environment for the nurses, meaning that
an active participation by the nurses during the meetings
was stimulated. As described in the literature, it seems that
interactive and personal educational meetings are more
effective than passive education (Grol et al., 2005). The
inclusion of ‘key figures’ (Grol et al., 2005) and different
types of ‘adopters’ (Rogers, 2003) was the basic principle
for the selection of the nurses for the educational pro-
gramme. Seven nurses per experimental ward, including
the charge nurse, were invited to attend the meetings and
exchange of knowledge and experiences among the experi-
experimental wards and one charge nurse. After the five
educational meetings, a 90-min plenary session was orga-
nized for all the nurses from each experimental ward
(including nurses who were not invited to attend the five
educational meetings) to inform them about restraint use
and restraint-free care as well. As part of the educational
intervention, the nurse specialist was available, from the
start of the educational intervention until 8-months post-
intervention, for consultation 28 h per week, visited the
wards once a week, attended multidisciplinary meetings
about residents and stimulated nurses to use alternatives to
physical restraints and to use the least restrictive restraints,
such as sensor mats. During the visits to the experimental
wards and the multidisciplinary meetings the nurse specia-
list evaluated the use of physical restraints on residents and
discussed difficulties in achieving restraint-free care with
the nurses on the wards and with the persons present at the
2.3.1. Physical restraints
The use of physical restraints per resident was measured
using an observation tool, which had been designed for the
present. Any limitation on an individual’s freedom of move-
ment was regarded as a physical restraint. Examples of
restraint types are belts tied to a chair or bed, bilateral
bedrails, tight sheets (a sheet over belly and upper legs that
overturned chairs, chairs on a board (a chair with chair legs
fixed to a board), special sheets (a fitted sheet including a
coat enclosing a mattress), sleep suits, sensor mats (includ-
ing sensor strips in beds) and infrared systems. In the
Netherlands, the use of physical restraints in psycho-ger-
iatric nursing homes is regulated by the Psychiatric Hospi-
tals (Compulsory Admission) Act. The use of restraints is
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469461
only justified when specific conditions are met, such as the
lack of suitable alternatives to restraint use and the presence
or expectancyof danger (CBO and VWR, 2001). The formal
responsibility for the use of physical restraints in nursing
homes lies with the nursing home physicians (CBO and
VWR, 2001). However, the decision-making process in the
use of physical restraints is a multidisciplinary matter (CBO
and VWR, 2001) in which nursing staff plays an important
role (Hamers and Van Wijmen, 2003). The use of physical
restraints was measured by trained observers (n = 8) on four
separate occasions(morning, afternoon, evening,night)over
a 24-h period. The observers were blinded to the experi-
mental and control conditions. The day of visit to each ward
was unannounced to discourage any artificial removal of
restraints by staff (Evans et al., 1997). To measure the inter-
rater reliability, two of the eight observers were selected and
they scored the same residents with restraints (k = 1.0). Four
variables concerning restraint use were used in the study.
observedto be restrained at anytime inthe four observations
over a 24-h period (yes/no). Secondly, all types of restraints
used with psycho-geriatric nursing home residents over the
24-h period were recorded. Thirdly, restraint intensity indi-
cated the number of times in four observations over a 24-h
period that a resident was restrained, ranging from not
restrained in four observations (score 0) to restrained during
all four observations (score 4). The duration of each restraint
use (in minutes or hours) was not measured. Fourthly,
multiple restraints indicated the number of different types
of restraints used per resident recorded in four observations
over a 24-h period. For example, when the resident is
observed with a belt and bedrails during the four observa-
tions, the resident had a multiple restraint score of 2.
2.3.2. Residents’ characteristics
Residents’ characteristics, such as cognitive status, self-
and social engagement, were measured at each Post-test
using the Minimum Data Set (MDS) version 2.1, which is
part of the Resident Assessment Instrument (RAI) (interRAI,
2002). The MDS was completed by trained nurses working
on the wards, who filled out the questionnaires for residents
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469 462
Educational intervention for the nurses of psycho-geriatric nursing home residents
(1) Educational programme
Basic principles The manual ‘Restraint-free care. Individualized approaches for frail elders’ Strumpf et al. (1998)
The paper ‘Het gebruik van vrijheidsbeperkende interventies in de zorg. Een richtlijn voor
verpleegkundigen en verzorgenden in een multidisciplinaire omgeving’ CBO and VWR (2001)
The educational programme ‘Een balans tussen: Veiligheid, Vrijheid en Vrijheidsbeperking’
Dielis-van Houts and Schuurmans (2004)
National and international literature and information from conferences
Learning goals To become familiar with techniques of individualized care and to use these in practice
To improve the knowledge about the ineffectiveness and consequences of restraint use
To improve the knowledge about the decision-making process, legislation and ethical aspects of restraint use
To improve the knowledge about risk behaviour of residents and their acceptance
To learn skills to recognize, analyse and discuss risk behaviour of residents and to use specific interventions
To apply the learned knowledge and skills in daily practice
TopicsEffectiveness and consequences of restraint use, including experiences of residents and caregivers
concerning restraint use
Decision-making process towards restraint use, including legislation and ethical aspects
Risk behaviour of residents and strategies to analyse this behaviour
Alternatives to restraint use
Power Point presentations, literature, videotapes, photo-reportages, real-life cases and homework assignments
Nurse specialist (=registered nurse specialized in the use of physical restraints and their reduction)
Five small-scale meetings carried out of a 2-month period, each lasting for 2 h; 7 seven nurses per ward,
including the charge nurse, were invited and divided into seven groups
One plenary session per ward, each lasting for 90 min; all nursing staff per ward were invited
Tasks of nurse specialist
Supporting the nursing staff in achieving restraint-free care in daily practice
Nurse specialist (=registered nurse specialized in the use of physical restraints and their reduction)
Available for consultation 28 h per week
Visiting wards once a week
Attending multidisciplinary meetings about residents
Stimulating nursing staff to use alternatives to physical restraints and to use the least restrictive restraints
intheir care. Different scales based on MDSitems were used
to determine the residents’ characteristics. The cognitive
status of residents was measured using the Cognitive Per-
formance Scale (Morris et al., 1994). The scores of the scale
ranged from 0 (intact) to 6 (very severe impairment). Self-
performance in activities of daily living (ADL status) was
measured using the MDS ADL Self-performance Hierarchy,
in which the scores ranged from 0 (independent) to 6 (total
dependency) (Morris et al., 1999). A mobility scale was
developed from seven MDS items (movement in bed, trans-
fer in and out of bed, transfer to standing up, walking in the
room, walking in the corridor, locomotion on ward, locomo-
scores of the mobility scale ranged from 0 (independent) to
28 (total dependency). Depression in residents was mea-
sured using the Depression Rating Scale, in which the scores
ranged from 0 to 14 (Burrows et al., 2000). Residents who
scored ?3 (cut-off score) on the scale required further
evaluation to diagnose depression. The Social Engagement
Scale was used to determine social engagement in residents
(Mor et al., 1995). The scores ranged from 0 (lowest level of
social engagement) to 6 (highest level of social engage-
ment). Cronbach’s alphas for the scales have been tested in a
previous study (submitted for publication) and found to be
sufficient for all the scales (acognitive status= 0.64, aADL
status= 0.86, amobility= 0.97, adepression= 0.66, asocial engage-
ment= 0.77). In addition, the reliability and validity of the
MDS and related scales have been found to be sufficient in
other studies (Burrows et al., 2000; Holtkamp, 2003; Mor
et al., 1995; Morris et al., 1994, 1999).
2.4. Ethical considerations
Approval for the study was obtained from the Medical
Ethical Committee of the University Hospital and the
University. Representatives of the residents received writ-
ten information about the study from the nursing home and
the university. Based on this information the representa-
tives were asked to give written consent for the use of
personal data on the residents in the study. Nurses and
other employees of the nursing homes were informed
about the study through presentations and written informa-
tion. The trial identification number of the study is
2.5. Data analysis
Descriptive statistics were computed for the character-
istics of the psycho-geriatric nursing home residents. Dif-
ferences between the experimental and the control group in
restraint use and residents’ characteristics were investigated
using the x2-test for variables at a dichotomous level
(restraint status, types of restraints, gender), and the inde-
pendent samples t-test for variables at an interval and ratio
level (restraint intensity, multiple restraints, age, cognitive
status, self-performance in daily activities, mobility, depres-
sion, social engagement). Fisher’s exact test values were
computed for tables including a cell with a frequency less
than 5. The normality of the variables was checked using
frequencies and distribution tables. Based on these results
the variable mobility was categorised into five groups,
ranging from 0 ‘independent’ to 4 ‘total dependency’.
Furthermore, differences between the experimental and
thecontrolgroup inrestraint intensityandmultiple restraints
were also tested using the Mann–Whitney U-test because of
a non-normality in these variables. p ? 0.050 was consid-
ered to be statistically significant. However, to examine
differences on the use of physical restraints per restraint
type between the experimental and the control group, the
testing because of the increasing risk of Type 1 errors. The
data were analysed using SPSS, Version 13.0.
A total of 138 residents were admitted to the wards
during the study period. Of this sample, 33 residents were
not included in the analyses, mainly because these residents
died or informed consent had not been obtained. A total of
105 residents were included in the analyses. The total
response flow of the study is shown in Fig. 1.
As can been seen from Table 2, the residents admitted to
the experimental wards were statistically significantly more
dependent in ADL (p = 0.045) as compared to the residents
admitted to the control wards at Post-test 1. Furthermore, no
significant differences on the residents’ characteristics
between the experimental and the control group were found.
programme (30% of the total nurses in the experimental
wards), one of whom attended none of the meetings. The
mean age of the sample was 37 years (S.D. = 10.0) and most
of the nurses were women (80%). These persons were
qualified as charge nurses (13%), registered nurses (11%),
care workers (64%), care helpers (4%), care assistants (2%)
and student care workers (2%).
3.2. The effect of the educational intervention
At Post-test 1, a total of 13 (38.2%) residents newly
admitted to the experimental wards were restrained
(Table 3). There was no statistically significant difference
between the experimental and the control group in restraint
status at Post-test 1. Moreover, restraint intensity and multi-
ple restraints did not differ between groups. At Post-test 2, a
total of 22 (51.2%) residents in the experimental wards were
restrained. The restraint status, restraint intensity and multi-
ple restraints did not differ statistically significantly between
groups (Table 3). Also at Post-test 3, no statistically sig-
nificant differences were found between groups. A total of
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469 463
25 (47.2%) residents in the experimental wards were
During the study period, several types of restraints were
used in residents newly admitted to the nursing home wards
(Table 4). Bilateral bedrails were the most frequently used
restraints at Post-test 1 (13 (24.1%)), Post-test 2 (17
(23.0%)) and Post-test 3 (25 (27.8%)). After the use of
bilateral bedrails, infrared systems were the most commonly
used at Post-tests 2 and 3. For each type of restraint no
statistically significant differences between groups were
found at Post-tests 1, 2 and 3. Only the use of belts in
bed tended to be lower in the experimental group (0 (0%)) as
compared to the control group (3 (8.1%)) at Post-test 3
(p = 0.066). At Post-test 2, the use of bilateral bedrails
tended to be higher in the experimental group (14
(32.6%)) as compared to the control group (3 (9.7%))
(p = 0.026). Furthermore, the use of deep and overturned
chairs tended to be higher in the experimental group
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469 464
Fig. 1. Response flow of the study.
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469
Comparison of residents’ characteristics in the experimental group versus control group at Post-test 1 (n = 54), Post-test 2 (n = 74) and Post-test 3 (n = 90)
Post-test 1Post-test 2Post-test 3
(n = 20)
(n = 34)
(n = 31)
(n = 43)
(n = 37)
(n = 53)
Age in years
82.5 (7.0) 80.1 (7.0)0.231
83.1 (7.9) 80.4 (7.0)0.136
83.2 (7.7) 80.4 (7.1)0.079
aAbsolute numbers and (%).
bMDS Cognitive Performance Scale, range 0 (intact) to 6 (very severe impairment).
cMDS ADL Self-performance Hierarchy, range 0 (independent) to 6 (total dependency).
dScale developed from seven MDS items; the variable mobility was categorised into five groups, range 0 (independent) to 4 (total dependency).
eMDS Depression Rating Scale, range 0–14; scores ?3 indicate symptoms of depression.
fMDS Social Engagement Scale, range 0 (lowest level of social engagement) to 6 (highest level of social engagement).
Difference in the use of physical restraints between the experimental and the control group at Post-test 1 (n = 54), Post-test 2 (n = 74) and Post-test 3 (n = 90)
Use of physical
Post-test 1Post-test 2Post-test 3
(n = 20)
(n = 34)
(n = 31)
(n = 43)
(n = 37)
(n = 53)
0.55 (1.05) 0.91 (1.42)0.327b(0.442c)0.61 (1.15)1.05 (1.41)0.164b(0.110)c
0.78 (1.25)1.02 (1.39) 0.414b(0.440c)
0.35 (0.59)0.56 (0.82)0.325b(0.433c) 0.45 (0.77) 0.84 (0.97)0.071b(0.072c)0.59 (0.90)0.72 (0.86)0.516b(0.417c)
aAbsolute numbers and (%).
bThe independent samples t-test.
cThe Mann–Whitney U-test.
dRestraint intensity, range 0 (not restrained in four observations) to 4 (restrained during all four observations over a 24-h period).
eMultiple restraints indicates the number of different types of restraints used per resident recorded in four observations over a 24-h period (ranging from 0 to infinite).
(6 (11.3%)) than in the control group (0 (0%)) (p = 0.041) at
This study has shown that an educational intervention
does not prevent the use of physical restraints in residents
use of physical restraints on residents newly admitted to the
wards in the experimental group was not statistically sig-
nificantly different from that of the control group at each
Post-test. Furthermore, no significant differences in restraint
status, restraint intensity, multiple restraints and types of
restraints were found between the groups. To the best of our
knowledge, this was the first RCT study investigating the
use of physical restraints with newly admitted residents. The
results are, however, not in line with the positive effects of
education and consultation in reducing the use of physical
restraints found in the two previous RCT studies (Evans
et al., 1997; Testad et al., 2005), albeit that Huizing et al.
(2006) showed no effect on restraint reduction. In the RCT
studies comparable interventions were tested, mostly con-
sisting of education and consultation. Nevertheless, differ-
ences in research, such as data-collection methods used,
definitions of physical restraints and differences among the
countries involved regarding legislation concerning restraint
use, the job responsibilities and attitudes of nursing staff in
nursing homes, might be an explanation for the differences
in the study results. For example, studies about restraint-
reduction programmes collected restraint data using differ-
ent data-collection methods, such as questionnaires, resi-
dents’ records and interviews (Mason et al., 1995; Neufeld
et al., 1999; Testad et al., 2005), in which those persons
involved in collecting restraint data sometimes also parti-
cipated in the intervention. This may result in an over-
estimation of the intervention effect. In the present trial,
the use of physical restraints was determined by blinded and
trained observers. Furthermore, a broad definition of phy-
sical restraints was used in our trial, which differed from the
definition as used in the trial of Evans et al. (1997). Evans
et al. (1997) excluded the use of bedrails, the most prevalent
type of restraint in our trial. Finally, differences between
countries in attitudes of nursing staff regarding the use of
physical restraints can explain differences in the effective-
ness of restraint-reduction programmes. In this respect, a
recent international study (submitted for publication) has
shown that nursing staff from three European countries have
different attitudes and opinions towards physical restraint
use in nursing home residents.
The intervention in the present study did not change
behaviour and maintain changes in the target group,
although education seems to be a necessary component of
implementation strategies (Grol et al., 2005). By using
interactive education and a nurse specialist’s consultation
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469466
Differences between the experimental and the control group in the use of physical restraints with psycho-geriatric nursing home residents per restraint type at Post-test 1 (n = 54), Post-test 2
(n = 74) and Post-test 3 (n = 90)
Restraint type (absolute numbers and (%))
Control (n = 20)
Experimental (n = 34)
Control (n = 31)
Experimental (n = 43)
Control (n = 37)
Experimental (n = 53)
Belt in chair
Belt in bed
Chair with a table
Deep or overturned chair
Chair on a board
(Wheel)chair on the brakes at the table
*Statistically significant difference (between groups) tested with x2-test (or Fisher’s exact test was computed for tables including a cell with a frequency less than 5); p ? 0.010 (correction for
in the study, the intervention does seem to fulfil the condi-
tions for the dissemination of an innovation (see Grol et al.,
2005). Furthermore, the educational programme, as part of
the intervention, was comparable for duration and content
with restraint education in other studies showing positive
effects on reducing restraints (see review Evans et al., 2002).
However, Grol et al. (2005) stated that effective dissemina-
tion does not guarantee implementation of the intervention.
For example, implementation of the intervention on a nur-
sing home level, as in other RCT studies (Evans et al., 1997;
Testad et al., 2005), involving the entire organization and
overcoming organizational hindering may therefore be
worthwhile. In the present study, the ineffectiveness of
the intervention may, therefore, also be related to the imple-
mentation on a ward level, which may limit an organiza-
Furthermore, a continuing multidisciplinary educational
programme (with more professions involved) may also be
worthwhile in order to achieve restraint-free care.
The ineffectiveness of the intervention in preventing phy-
sical restraints may also imply a mainly routine usage of
physical restraints on psycho-geriatric nursing home resi-
dents. This is certainly not in accordance with the principles
of individualized care (Strumpf et al., 1998). As stated by
Strumpf et al. (1998) the use of physical restraints is incon-
sistent with the principles of individualized care. It was
hypothesized that preventing the use of restraints on these
residents may be a first step in the transition to restraint-free
care, given the new situation where nurses can implement
individualized care and break with routines concerning
restraint use. However, for changing the routine usage of
physical restraints,educationand consultation were not suffi-
towhat extent nurses acquired any knowledge about restraint
restraint-free care, the necessary transition process to
restraint-free care was clearly not attained. More efforts are
therefore needed to break the routine usage of physical
restraints with psycho-geriatric nursing home residents. In
this respect, research examining the attitudes and knowledge
about restraint use is recommended. This would appear to be
an important first step in order to develop effective interven-
clarify differences between countries about their use of phy-
During the study period, the types of restraints used with
residents did not differ between the experimental and the
control group. Bilateral bedrails were the most frequently
used restraints with newly admitted residents, which is in
studies (Capezuti et al., 2007; Hamers et al., 2004; Huizing
et al., 2006). However, the types of restraints subsequent to
the use of bedrails differed from other studies, which have
shown more restrictive types of restraints (Hamers et al.,
2004; Hantikainen, 1998; Huizing et al., 2006). Whereas
other studies showed that belts and chairs with a table were,
after bedrails, the most frequently used, in the present study
bedrails were followed by infrared systems at Post-tests 2
and 3. Although the most restrictive types of restraints, such
as belt in bed and chair, seemed to be replaced by less
restrictive restraint types, such as infrared systems, the small
sample size and the small differences in the use of these
different types of restraints must be taken into account.
Unfortunately, the use of the most restrictive types of
restraints were still used with residents newly admitted to
the experimental wards, and no statistically significant dif-
ferences were found between the experimental and the
control group. In an international study (submitted for
publication) about nurses’ attitudes towards restraint use,
the opinions of nursing staff regarding the restrictiveness of
restraint measures was examined in order to improve under-
standing of the process towards restraint-free care. In this
respect, further studies investigating the opinions of nursing
home residents and their family are also recommended.
There are some limitations to this study that need to be
consisted of nursing home residents from 14 nursing home
of older people in the south of the Netherlands. The small
sample size could have an effect on the power of the study.
However, given the small effects, enormous numbers of
residents are needed to measure an effect, which is also not
feasible. By selecting only wards from one coordinating
organization of nursing homes, an optimal similarity in the
care between these wards was assumed. Secondly, contam-
ination bias among wards might be a problem in the study.
However, efforts were made to reduce the contaminationbias
by limiting information for nurses concerning the aim and
design of the study from the outset. After randomization the
experimental wards were fully informed and were requested
to be careful with the information in relation to the control
place of residence and no information was available about
restraint status before admission. However, it was assumed
that it would be easier to introduce restraint-free care to
residents newly admitted to a nursing home ward than to
residents already residing in these wards, because for these
the consultation of the nurse specialist, as part of the inter-
vention, was carried out by only one nurse specialist. This
possibly might not have been enough to assist nurses suffi-
ciently in reducing restraint use. Fifthly, no information was
gathered regarding the extent to which nurses had embraced
the philosophy of restraint-free care and had gained knowl-
edge about restraint use through the educational intervention.
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469467
of the intervention on the knowledge and skills of the nurses.
According to this study, an educational programme
combined with a nurse specialist’s consultation does not
prevent the use of physical restraints on residents newly
admitted to psycho-geriatric nursing home wards. The con-
tinued high prevalence values for restraint use and the
ineffectiveness of the intervention in preventing physical
restraints seem to imply a mainly routine usage of physical
restraints on psycho-geriatric nursing home residents, which
is certainly not in accordance with the principles of indivi-
dualized care. Although other studies have shown promising
results with these types of interventions with regard to
restraint reduction, it is important to develop additional
interventions in order to prevent restraint usage with newly
admitted psycho-geriatric nursing home residents.
This study partly has been funded by the Province of
Approval for this study was obtained from the Medical
The authors would like to thank the residents and staff of
the nursing homes Lu ¨ckerheide, Firenschat, Vroenhof, De
Dormig, Dependance De Dormig Kerkrade, Heereveld and
Bocholtz for their participation and support.
Competing interest: None.
Conflict of interest
Burrows, A.B., Morris, J.N., Simon, S.E., Hirdes, J.P., Phillips, C.,
2000. Development of a Minimum Data Set-based depression
ratingscale forusein nursinghomes.AgeAgeing29 (165–172).
Capezuti, E., Wagner, L.M., Brush, B.L., Boltz, M., Renz, S.,
Talerico, K.A., 2007. Consequences of an intervention to reduce
restrictive side railuse innursinghomes.J.Am.Geriatr.Soc.55,
CBO, VWR, 2001. Het gebruik van vrijheidsbeperkende interven-
ties in de zorg. Een richtlijn voor verpleegkundigen en verzor-
genden in een multidisciplinaire omgeving (The use of restraints
in care situations. A guideline for nurses in a multidisciplinary
environment) Dutch Institute for Healthcare Improvement
Dielis-van Houts, A., Schuurmans, M., 2004. Een balans tussen:
Veiligheid, Vrijheid en Vrijheidsbeperking (The balance
between safety, freedom of movement and use of restraints).
Evans, D., FitzGerald, B., 2002. Reasons for physically restraining
patients and residents: a systematic review and content analysis.
Int. J. Nurs. Stud. 39, 735–743.
Evans, D., Wood, J., Lambert, L., 2003. Patient injury and physical
restraint devices: a systematic review. J. Adv. Nurs. 41 (3), 274–
Evans, D., Wood, J., Lambert, L., 2002. A review of physical
restraint minimization in the acute and residential care settings.
J. Adv. Nurs. 40 (6), 616–625.
Evans, L.K., Strumpf, N.E., 1990. Myths about elder restraint.
Image J. Nurs. Sch. 22 (2), 124–128.
Evans, L.K., Strumpf, N.E., Allen-Taylor, S.L., Capezuti, E., Mai-
nursing homes. J. Am. Geriatr. Soc. 45 (6), 675–681.
Hamers, J.P.H., Gulpers, M.J.M., Strik, W., 2004. Use of physical
restraints with cognitively impaired nursing home residents. J.
Adv. Nurs. 45 (3), 246–251.
Hamers, J.P.H., Huizing, A.R., 2005. Why do we use physical
restraints in the elderly? Z. Gerontol. Geriat. 38, 19–25.
zorg: een verpleegkundige beslissing (The use of restraints in
health care: a nursing decision). TvZ Tijdschrift voor Verpleeg-
kundigen (3), 30–33.
Hantikainen, V., 1998. Physical restraint: a descriptive study in
Swiss nursing homes. Nurs. Ethics 5 (4), 330–346.
Holtkamp, C.C.M., 2003. Effects of the Resident Assessment
Instrument on Quality of Care and Quality of Life in Nursing
Homes. NIVEL, Utrecht.
Huizing, A.R., Hamers, J.P.H., Gulpers, M.J.M., Berger, M.P.F.,
2006. Short-term effects of an educational intervention on phy-
interRAI, 2002. RAI for Nursing Home Care (RAI 2.1) interRAI
Karlsson, S., Bucht, G., Eriksson, S., Sandman, P.O., 2001. Factors
relating to the use of physical restraints in geriatric care settings.
J. Am. Geriatr. Soc. 49 (12), 1722–1728.
Mason, R., O’Connor, M., Kemble, S., 1995. Untying the elderly:
response to quality of life issues. Geriatr. Nurs. 16 (2), 68–72.
MinVWS, 1997.Qualified for the Future.Coherent Training System
for Nursing and Patient Care in the Netherlands: A Summary.
Ministry of Health, Welfare and Sport, Zoetermeer.
Mor, V.,Branco, K., Fleishman, J., Hawes,C.,Phillips, C., Morris,J.,
home residents. J. Gerontol.: Psychol. Sci. B 50 (1), P1–P8.
Morris, J.N., Fries, B.E., Mehr, D.R., Hawes, C., Phillips, C., Mor,
V., Lipsitz, L.A., 1994. MDS Cognitive Performance Scale. J.
Gerontol.: Med. Sci. 49 (4), M174–M182.
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469 468
Morris, J.N., Fries, B.E., Morris, S.A., 1999. Scaling ADLs
within the MDS. J. Gerontol.: Med. Sci. A 54 (11), M546–
Neufeld, R.R., Libow, L.S., Foley, W.J., Dunbar, J.M., Cohen, C.,
Breuer, B., 1999. Restraint reduction reduces serious injuries
among nursing home residents. J. Am. Geriatr. Soc. 47 (10),
Strumpf, N.E., Robinson, J.P., Wagner, J.S., Evans, L.K., 1998.
Restraint-free Care; Individualized Approaches for Frail Elders.
Springer Publishing Company, Inc., New York.
Suen, L.K.P., Lai, C.K.Y., Wong, T.K.S., Chow, S.K.Y., Kong,
S.K.F., Ho, J.Y.L., Kong, T.K., Leung, J.S.C., Wong, I.Y.C.,
2006. Use of physical restraints in rehabilitation settings: staff
knowledge, attitudes and predictors. J. Adv. Nurs. 55 (1),
Sullivan-Marx, E.M., Strumpf, N.E., Evans, L.K., Baumgarten, M.,
Maislin, G., 1999. Predictors of continued physical restraint use
Am. Geriatr. Soc. 47 (3), 342–348.
Testad, I., Aasland, A.M., Aarsland, D., 2005. The effect of staff
training on the use of restraint in dementia: a single-blind
randomised controlled trial. Int. J. Geriatr. Psychiatry 20 (6),
Van der Windt, W., Calsbeek, H., Talma, H., Hingstman, L., 2003.
Feiten over verpleegkundige enverzorgende beroepen in Neder-
land 2003 (Facts about nursing professions in the Netherlands
2003). The Netherlands Centre for Excellence in Nursing
to use physical restraints with older people: testing the theory of
reasoned action. J. Adv. Nurs. 35 (5), 784–791.
A.R. Huizing et al./International Journal of Nursing Studies 46 (2009) 459–469 469