Risk factors for coercion length at psychiatric hospitals in Israel: Relationship with staff

Abstract

Background:

Coercive interventions continue to be applied frequently in psychiatric care when patients are at imminent risk of harming themselves and/or others.

Aim:

The purpose of this study was to demonstrate the relationship between the length of coercion and a variety of factors, including the sociodemographic background of patients, their diagnoses and the characteristics of hospital staff.

Methods:

This is a one-year cross-sectional retrospective study, including records of 298 patients who underwent restraint and/or seclusion interventions in male acute, closed wards in two psychiatric hospitals in Israel.

Results:

A higher proportion of academic nurses to nonacademic nurses on duty leads to a shorter coercion time (P < 0.000). The number of male staff on duty, without any relation to their level of education, also leads to the shortening of the coercion time.

Conclusion:

The presence of registered, academic female nurses, male staff on duty and the administration of medication before coercive measures can reduce the length of restriction.

INTRODUCTION

Inpatient violence has not disappeared despite the pharmacological treatment of mental disorders, and it is a widespread problem in psychiatric wards often with serious consequences. Coercive interventions continue to be applied frequently in psychiatric care when patients are at imminent risk of harming themselves and/or others.

This topic has recently gained prominence in scientific discussions and research studies. Many questions remain without a clear answer, such as the origins of violence, the reasons for using seclusion and restraint measures and the effective alternatives for these interventions. The use of coercive measures is controversial, as it can lead to psychological and physical consequences for both the patient and the staff, and its therapeutic efficacy still requires empirical validation.[1] The recent literature provides some evidence that seclusion and restraint have deleterious physical or psychological consequences.[2] As coercive measures can be dangerous for psychiatric patients, international mental healthcare organisations promote the elimination of these interventions.[2,3]

We have recently published a study about the relationship between demographic and clinical characteristics of restrained and secluded psychiatric inpatients.[4] This study found that patients with personality disorders were physically limited for the longest time, while schizophrenia patients were restricted for the shortest time compared with other diagnoses. The study also demonstrated a negative correlation between the length of the use of coercive measures and the presence of registered nurses in the wards. Moreover, we found that female nurses generally used restraint less than male nurses did. However, both male and female academic nurses used fewer coercive measures than their nonacademic counterparts.[4]

In this study, we introduce a broader sample of patients, including data from two other closed psychiatric wards from two mental health centres located in different geographic areas in Israel. The aim of this study was to demonstrate the relationship between the length of coercion and a variety of factors, including the sociodemographic background of patients, their diagnoses and the characteristics of hospital staff. The use of coercion is always of clinical interest, and many articles on the subject have been published in recent 3 years. Finding at least 28 publications during the time passed from our previous publication[4] is proof that this topic still continues to be an issue of interest. These publications can be grouped according to seven issues: A) education and quantity of nursing staff related to coercion[3,5,6,7,8,9,10,11,12], B) review and meta-analyses[1,2,3,4,13,14,15,16], C) clinical and demographic characteristics of patients[2,15,16,17,18,19,20,21], D) human rights[22,23,24], E) de-escalation[25,26], F) treatment[13,27] and G) duration of coercion.[28]

The rationale for our study was to evaluate data from other wards in the same mental health centre and in another mental health centre, including both patients and staff, to confirm or deny our previous findings.

MATERIALS AND METHODS

This is a one-year cross-sectional retrospective study of the use of restraint and seclusion in an upholstered room in male acute, closed psychiatric wards in two government mental health centres in Israel. It covers the period from 1 January 2014 to 31 December 2014. This period was chosen since it was the last year before the reform of the Israeli Ministry of Health concerning the policy towards physical restriction of mentally ill patients. The aim of this reform was to reduce the use of coercive measures. Following this reform from 2015 on, there was a drastic decrease in the number of restraint patients. To validate our previous results, we designed this recent study by trying to replicate the results using data from other wards and another mental health centre in the same year, while the policy of coercion was the same all over the country. In our new study, we replicated the method to adopt or neglect the previous results.

This study is an extension of our previous restricted one, which is related to only one ward.[4] The current work is expanded and includes two closed male psychiatric wards in different mental centres in Israel: Be’er Sheva governmental mental health centre and Jerusalem governmental mental health centre.

Each ward contains 30 beds and admits both voluntary and involuntary patients. In these hospitals, there are only gender-separated acute, closed wards. The gender-separated wards neutralise the effect of interactions between male and female patients. The staff was gender-mixed with varying levels of professional education.

In the study, we included all hospitalised patients who underwent coercive measures during the period under study. The data were retrieved from each patient’s records, and analysis was performed on the following variables: age, marital status, education, psychiatric diagnosis according to International Classification of Diseases, Tenth Revision (ICD-10), length (in days) of hospital stay, type of event (seclusion or restraint), reasons for coercion, time (morning/noon/night) of event, number of events per patient during every admission, total length (hours) of coercion, number of previous hospitalisations, aggression in past, present treatment, treatment during the procedure and the per cent of inpatient occupancy in the ward at the time of event. Our dependent variable is length of coercion, and the independent variables are duration of hospitalisation, number of hospitalisations, number of male staff, number of academic nurses¹, number of academic male nurses and number of academic female nurses.

We evaluated correlations between the length of coercion (seclusion or restraint), duration and number of hospitalisations, the number of staff members and their gender and level of professional education.

The Israeli law for treatment of mentally ill patients (1991) sets a policy for using means of coercion as follows: A. restraint method may refer to seclusion or restriction; B. use of coercive measures for the hospitalised patients should be made only to the extent required for medical treatment of the patient or to prevent danger to himself or others; and C. the medical directive regarding the use of a coercive measure should be given in writing by a physician for a limited period, in a state of emergency, and in the absence of a physician, the nurse is permitted to provide a coercive instruction.

During the use of coercive measures, patients received either their regular treatment (but earlier than usual) or additional sedative drug therapy. There were two groups of additional sedative medications: 1) antipsychotics and 2) benzodiazepines. The choice of the medication was based on the clinician’s judgement.

The Institutional Review Board approved the study.

Statistical analysis

Categorical variables were summarised using frequencies and percentages. Quantitative variables are presented as mean and standard deviation (SD) if normally distributed or median and range from minimal to maximal (min–max) if not normally distributed.

For univariate analysis, associations between the dependent variable and the categorical independent variables were studied using the Mann–Whitney or the Kruskal–Wallis test, as appropriate. For determining associations of the dependent variable with the independent quantitative variables, we used Spearman’s rank correlation coefficient rho.

For multivariate analysis, multivariate general linear regression was created. This regression model includes covariates that were found to be statistically significant (or border significant) in the univariate analysis and/or were clinically significant. Also, we tested possible interaction between independent variables.

P value < 0.05 was considered as statistically significant; 0.1 > P > 0.05 was considered as borderline significant.

Statistical analyses were performed using IBM Statistical Package for the Social Sciences (SPSS) Statistics 21 for Windows (IBM Corp.: Armonk, NY).

RESULTS

During 2014, there were 973 hospitalisations (voluntary and involuntary) in total in the study wards. Of these, 298 patients (30.6% of all hospitalisations in these facilities) were restricted and/or secluded. Of these patients, 130 (43.6%) subjects were hospitalised involuntarily following a district psychiatrist’s decision; 11 (3.7%) were involuntarily admitted due to a court decision for forced treatment; and 157 (52.7%) were hospitalised voluntarily. As some patients underwent restraint or seclusion repeatedly, the total number of coercive interventions was 1216: 709 (58.3%) of them were restrained and 507 (41.7%) were secluded. Approximately 93% of the patients had a record of aggressive behaviour in the past. The mean duration of seclusion was 3.37 hours (SD = 9.9), ranging from 0.17 to 31.75 hours. The mean duration of mechanical restraint was 7.7 hours (SD = 3.2), ranging from 0.17 to 95 hours. The mean age (SD) of subjects was 33.6 (13.4) years, ranging from 18 to 90 years. Of the 298 patients, more than one-quarter (28.2%) had dual comorbidity (personality disorders accompanied by drug use). As this is a large fraction, we decided to merge them into one diagnostic group. Demographic and clinical characteristics of these patients are presented in Table 1.

Table 1.

Demographic and clinical characteristics of patients who underwent restraint and/or seclusion in an upholstered room (n=298)

	n	(%)
Age (years)
18–40	213	71.5
41–65	79	26.5
66+	6	2.0
Married status
Single	212	71.1
Married	47	15.8
Divorced	37	12.4
Widowed	0	0.0
No data	2	0.7
Education (years)
1–8	32	10.7
9–12	131	44.0
13+	20	6.7
No data	115	38.6
Cause for hospitalisation
Voluntary agreement	130	43.6
Forced hospitalisation by regional psychiatrist	157	52.7
Forced hospitalisation by court	11	3.7
Diagnoses
Schizophrenia	132	44.3
Personality disorders + mental and behavioural disorders due to psychoactive substance use	84	28.2
Acute psychotic state	38	12.8
Mood disorders	22	7.4
Organic mental disorders	14	4.7
Examination for medico-legal reasons	6	2.0
Mental retardation	2	0.7
Past aggression
No	22	7.4
Yes	276	92.6

Time spent in the upholstered room or being physically restrained was the longest for patients with personality disorders + behavioural disorders due to psychoactive substance use (median 4 hours, range 0.33–60.0 hours). The shortest time was among patients with mental retardation (median 3 hours, range 0.50–4.0 hours).

A negative correlation was found between the length of coercion (in hours) and the following factors: duration of present hospitalisation (rho = −0.251, P < 0.001); number of male nurses on duty (rho = −0.252, P < 0.001); number of registered, academic nurses on duty (rho = −0.122, P < 0.001); and number of registered, academic male nurses on duty (rho = −0.189, P < 0.001) [Table 2].

Table 2.

Correlation between clinical variables and staff characteristics with length of coercive measures in hours

Variables	rho*	P
Duration of hospitalisation	–0.251	<0.001
Number of hospitalisations	0.094	0.001
Number of male staff	–0.252	<0.001
Number of academic nurses	–0.122	<0.001
Number of academic male nurses	–0.189	<0.001
Number of academic female nurses	0.158	<0.001

*Spearman’s correlation coefficient

A positive correlation was found between the length of coercion (in hours) and the number of previous hospitalisations (rho = 0.094, P < 0.001).

Table 3 represents the comparison between different data that influence the duration of coercion time. We found nine factors from 12 that were significantly associated with diminishing coercion time. Among them are age more than 41 years, all psychiatric diagnoses other than personality disorder and mental behavioural disorders due to psychoactive substances, absence of previous aggressive behaviour, voluntary hospitalisation, seclusion, presence of male personal on duty, presence of academic nurses on duty and pharmacological intervention before restriction.

Table 3.

Factors affecting the duration of restriction time

Positive (prolongation of coercion)	Negative (shortening of coercion)	P
Young age (18–40 years)	Old age (41+ years)	0.039
Personality disorders + mental and behavioural disorders due to psychoactive substance use	Schizophrenia	0.003
	Organic mental disorder	0.048
	Mental retardation	0.000
	Mood disorder	1.000
	Acute psychotic state	0.007
	Examination for medico-legal reasons	1.000
Aggression history	Without aggression history	0.033
Divorced and single marital status	Married marital status	0.202
Forced hospitalisation by regional psychiatrist	Agreed hospitalisation	0.000
Short duration of present hospitalisation	Long duration of present hospitalisation	0.000
Restriction	Seclusion	0.000
High number of patients in ward	Low number of patients in ward	0.219
Female staff	Male staff	0.000
Number of nonacademic nurses	Number of academic nurses	0.000
Weekend days	Weekdays	0.059
Receiving medication treatment during restriction and/or seclusion	Receiving medication treatment before restriction and/or seclusion	0.000

DISCUSSION

The limitation of freedom is a cardinal issue in democratic, modern countries. Psychiatric patients, especially those who are admitted to closed wards, sometimes present a conflict between ensuring patients’ freedom on the hand and keeping them from self-harming behaviour and protecting the safety of other patients and the staff on the other hand. There are effective medications, which can control aggressiveness, but unfortunately sometimes it takes too long to achieve the desired effect. Furthermore, high doses of these drugs can induce serious side effects, including sudden death.[29,30,31,32,33,34,35] Modern psychiatry still uses coercive tools, such as seclusion and/or restraint, in cases where sedative pharmacotherapy and de-escalation techniques are not effective enough to reduce aggressiveness.

Our findings in this cross-sectional retrospective study demonstrate that during the period under study about one-third of all admitted patients (30.6%) required the use of coercive measures. These results correspond to our previous results and to other studies performed on a similar sample size with a similar duration.[4,18]

In retrospective studies, the outcome of interest has already occurred (or not occurred—e.g. in controls) in each individual by the time she or he is enrolled, and the data are collected either from records or by asking participants to recall exposures. There is no follow-up of participants.[36]

The present sample emphasises our previous findings that restraint was used more often than seclusion (58.3% vs 42.7%). These data do not correspond with the results of another study, where the inverse ratio was found: 46% of patients were restrained and 54% were secluded.[28] This difference can be explained by the fact that the upholstered room seclusion technique is a relatively new method for controlling violence in Israel, while restraint has long been in common use. Based on the available literature, it cannot be determined whether seclusion is superior to mechanical restraint or vice versa. Some authors recommend that further studies using strict methodology should be designed to clarify this issue.[16,20]

The factors connected with longer coercive time are age from 18 to 40 years, personality disorder and mental behavioural disorders due to psychoactive substances, history of past aggression, involuntary hospitalisation, a shorter duration of hospitalisation, additional medication treatment during the restriction and female staff on duty. Our findings concerning the influence of age on the coercion period of time are consistent with the results of other researchers, such as Dumais et al.,[37] Cole C et al.,[38] Roy C et al.[1] and Chieze et al.,[39] and with our previous publications, which came to the conclusion that younger age is one of the risk factors for coercion and its prolongation.[4]

It should be noted that our findings are not compatible with some publications that found that patients who suffered from organic mental disorder and schizophrenia more often undergo a longer time of coercion.[9,27] Our data demonstrate that substance use and personality disorders are more significant risk factors for a longer coercion time.

Mood disorder and examination for medico-legal reasons demonstrated only a trend of significance. Although patients with a principal diagnosis of a personality disorder together with psychoactive substance use were in second place after schizophrenia patients in terms of the number of coercion events (132 (44.3%) vs 84 (28.2%)), they were restricted for the longest period in comparison with patients with other diagnoses (P < 0.001). The use of physical restraint in this particular group brought temporary relief from the feelings of regret and remorse about the trouble they had caused others during their repeated cycles of violence and aggression. In our opinion, this discrepancy may be related to the fact that personality disorder patients, especially those with aggressive and violent behaviour, frequently cause a negative countertransference in the personal staff. In some cases, hopefully rare, staff might use these measures as punishment. We suggest that this unconscious reaction may lead to a longer time of restriction. Therefore, it is important to educate staff and raise their awareness about this issue. Moreover, physical restraint is often not the sole solution for patients with a personality disorder who exhibit aggressive and violent behaviour. They also need pharmacological and psychological treatment to achieve long-term, effective management of their behaviour.

A history of aggressive behaviour is a risk factor for a longer time of coercion. These findings are consistent with the results described by Luciano et al.[40] in their review concerning predictive factors of coercion use. As concluded in other reports, our data support the fact that a longer time from admission is likely to lower rates of restraint and seclusion.[41]

Concerning the issue of the connection between the number of patients in the ward and aggressiveness leading to coercion, some researchers report a positive association.[42] Our previous findings[4] did not support this. However, the present study reveals a trend of positive association, but it still does not reach statistical significance. We assume that the broadened sample size makes the present study resemble the other researchers’ results.

Nurses usually view coercive measures as undesirable, but still a necessary technique to cope with aggression. Nurses express the need for less intrusive interventions, although familiarity with the patients probably influences this perception of intrusiveness.

The results described by Gerace and Muir Cochrane note that factors for reduction in coercive measures include empathy and rapport between staff and patients combined with utilising trauma-informed care principles. In any case, when faced with dangerous situations, nurses felt moderately safe at work and were confident that they could use their clinical skills to ensure safety.[10]

In the literature, there are data concerning the correlation of characteristics of the staff and their years of experience with a positive attitude towards coercion. According to Al-Maraira and Hayajneh, working with male psychiatric patients and in acute psychiatric units increases the risk of using seclusion and/or restraint.[6] However, the literature concerning the relationship between staff characteristics and coercive measures is inconclusive.[3]

In contrast to our previous study, in this study we found that male personnel lead to shortening the time of coercion. At the same time, we found again that the number of academic nurses has a negative correlation with the duration of coercion measures. However, some researchers found that increasing the number of nurses on duty may not contribute to reduced use of seclusion and restraint.[9]

This current study demonstrates a trend of coercion time lengthening on weekends in contrast to weekdays. This can be explained according to the findings of Luciano et al., who note that availability of staff, specifically young doctors, reduces coercion. In Israel, the number of medical staff members (including physicians) is lower on weekends.[40]

We found that patients who received antipsychotic and/or sedative drugs before the coercion spent a shorter time under restriction than those who received the same drugs during restriction. We can explain this by the fact that these patients have already been under medication in contrast to patients of the second group, who became sedated later.

Our study has marked limitation. Unexpectedly, we found an asymmetry concerning the education and gender of nurses in the two public governmental psychiatric hospitals that we chose for our study. We used data from two sites with different staff members and different levels of professional education. This difference interfered with the comparison between the two sites. In any event, these two psychiatric sites are regular closed wards and not highly secured. Therefore, our conclusions should be accepted with caution concerning other kinds of psychiatric institutes. The limitation of this study involves the accessibility of some data concerning staff-related and environmental-related levels of risk factors for coercion length.

CONCLUSIONS

Our study demonstrates some evidence that patients’ diagnoses and characteristics of medical staff have an influence on the length of coercion time.

The presence of registered, academic female nurses and male staff on duty can reduce the length of restriction. Moreover, the administration of medication before coercive measures can also reduce the length of restriction. Substance use and personality disorders are significant risk factors for a longer coercion time.

Implications for nursing practice

Patients’ diagnoses, education and sex of medical staff on duty affect the duration of coercion time. Male staff leads to shortening the time of coercion. At the same time, the number of academic nurses has a negative correlation with the duration of coercion measures.

Ethical statement

The study protocol was accepted by the Institutional Review Board.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.