Abstract
Objective
To summarize the current state of knowledge on the use of seclusion and restraint with children and adolescents and to report the findings of an exploratory study to identify factors that place a child or adolescent at increased risk of seclusion during their admission.
Method
Literature searches were undertaken on MEDLINE, CINAHL and PsyclNFO databases. Articles were identified that focused specifically on seclusion and restraint use with children and adolescents or contained material significant to this population. The study reports findings from a retrospective review of patient charts, seclusion registers and staffing from an Australian acute inpatient facility.
Results
The data available in regard to seclusion use in this population is limited and flawed. Further research is needed on the use and outcomes of seclusion and restraint and on alternative measures in the containment of dangerousness. Both the literature and this study find that patients with certain factors are at increased risk of being secluded during an inpatient stay. These factors include being male, diagnoses of disruptive behaviour disorder and a previous history of physical abuse. Staffing factors did not show a relationship to the use of seclusion.
Conclusions
There are patient factors that predict increased risk of seclusion; these factors and their interrelationships require further elucidation. Further research is also needed on the outcomes, both positive and negative, of seclusion use and of alternatives to seclusion.
There has been increasing scrutiny of seclusion and restraint use in the last few decades, from both within and outside the clinical arena. Social, political and legal changes and advances in psychopharmacology and understanding of the therapeutic milieu have led to reductions in seclusion use. However, the research on the use of seclusion and restraint with children and adolescents is limited. This paper provides a literature review and reports from a retrospective, exploratory study of seclusion use in an Australian child and adolescent inpatient unit. The study aimed to identify patient and environmental factors that may place a child or adolescent at increased risk of being secluded during their inpatient stay.
Literature review
This review has focused on literature pertaining to the use of seclusion and restraint in the care of children and adolescents. Literature searches were undertaken on MEDLINE, CINAHL and PsycINFO and articles were followed up from the lists of references of articles initially identified.
Definitions and rationales
In 1998, the American Medical Association (AMA) Annual Meeting resolved to develop updated guidelines for the use of seclusion and restraint techniques with children and adolescents [1]. The AMA Council on Scientific Affairs adopted the following definitions which are used throughout this paper:
In addition, Swett distinguishes mechanical restraint from physical restraint, the use of bodily force for greater than 5 min to restrain the patient [2].
An American survey of the use of seclusion with children and adolescents gathered responses from 36 states [3]. Thirty-five permitted seclusion in an emergency and 18 allowed it as part of a therapeutic program. Only six had regulations specifically addressing the seclusion of children. Definitions of seclusion varied widely, from placement alone in a locked room to separating the child from the group. Danger to self or others was the most common indication. Other indications were the destruction of property, disruption of the therapeutic environment and chaotic behaviour. Contraindications were: use as a punishment; for staff convenience; and as a substitute for staff services or individualized treatment.
There is debate over whether seclusion constitutes a therapeutic intervention, a means of behaviour control or is even necessary. Murray and Sefchik [4] and Goren et al. [5] argue that restraint fails to teach children appropriate behaviour, sanctions the use of force as a response to conflict and may paradoxically increase aggressive behaviour. Irwin [6] and Lapierre [7] describe managing aggressive behaviour without the use of seclusion and mechanical restraint, although physical restraint was used by Irwin.
Two concerns in using seclusion and restraint in this population were raised by Garrison [8]. First, the risk of psychological damage may be increased because of the child's developmental and psychopathological status. Second, there may be heightened ethical concerns because the child has been involuntarily hospitalized by a parent or guardian. He concurred with the use of these strategies to maintain safety but raised the risk of staff ‘counter-aggression’ and questioned whether these procedures are therapeutic. Bath, an Australian reviewer, acknowledged a need for intervention to prevent harm [9]. He preferred physical restraint as he considered seclusion to have greater potential to be perceived and used as punishment, perceived as rejection or abandonment and to have greater legal ramifications. He recognized a role for seclusion with older adolescents with whom safe physical restraint is difficult.
A developmental clinical rationale for the use of seclusion and restraint in hospitalized children has been proposed by Gair [10] and Cotton [11]. Their rationale is grounded in theory of the normal socialization process and the ego deficits of this population. In families with normal children and ‘good enough’ parents, children gradually learn to control impulses and form adaptive relationships. Many hospitalized children have lacked appropriate socialization and have had inconsistent, abusive or neglectful parenting. They may also have difficult temperament, attention and learning problems, neuropsychological impairment and other mental illness. These problems hinder the development of age-appropriate coping strategies and impulse control. Limit-setting techniques are necessary for containment and protection of out-of-control behaviour and, as part of a therapeutic management plan, provide learning opportunities. Seclusion is one strategy to prevent violence and encourage the adaptive habit of leaving overwhelming situations.
Patterns of use
The use of seclusion and restraint has been described by a number of authors [8,12–19]. Also discussed is how these measures are influenced by the overall unit functioning [5,18,20], changes to policy and procedures [5,19,21] and legal changes [2,22]. Concerns about seclusion and restraint use arise from clinicians, administrative staff and the wider community and success in decreasing seclusion and restraint rates is reported [5,17,21,23,24].
Males are often (but not always) found to be more physically and verbally aggressive to others and more destructive than females. They are also more likely to be aggressive to others, whereas females are more likely to harm themselves. Where both seclusion and restraint is used, restraint was more often used for males, with no significant difference between the sexes for seclusion. No differences between rates of seclusion and restraint based on ethnic groups were found. There were conflicting results on the role of age. When older patients were found to require seclusion or restraint procedures more often, this appears to have been understood in terms of their greater physical size, although it has not been explicitly discussed. Where younger children have been found to be more aggressive, hypotheses put forward to account for this were: developmental differences in the expression of aggressive impulses; increased likelihood of admission because of aggressive behaviour; and staff having a limited range of responses compared to how they would manage an older child. Some found no difference in rates dependent on age.
The common diagnoses of in-patients were: disruptive behaviour disorders; affective disorder; pervasive developmental disorder; and mental retardation. Seclusion and restraint rates were higher for those with diagnoses of mental retardation, developmental disability and neurological impairment. They were also higher for patients with conduct disorder and other disruptive behaviour disorders and patients with borderline personality traits. Millstein found that children more likely to be secluded had weak verbal skills and a high incidence of specific learning disabilities and neurological problems [17]. They showed significant differences in their ability to tolerate frustration, manage stress, respond to external controls, control impulses, understand what is expected in a situation and apply learning to new situations. A previous history of aggression or assault or a history of suicide attempt was a risk factor for seclusion and restraint. Patients who were secluded were more likely to have a history of physical and sexual abuse. Some have found involuntary status predicts seclusion but other studies have found no difference.
Rates of seclusion and restraint use range from 19% to 60%; rates reported in adult units range from 2% to 66% [25]. Seclusion and restraint are used mainly in the management of aggression, with restraint use being commonly described, and also in the management of self-harm. Duration of seclusion use varied from 84 to 31 min and to 14.5 min where seclusion was used earlier as part of a therapeutic management process. By contrast, Garrison reported that the average duration of mechanical restraint was 103.8 min and physical restraint was 5.1 min [14]. Earle found that younger children had significantly shorter seclusion periods (average 1.3 h) than older children (average 1.6 h) [18]. When reported, in a third to half of aggressive incidents medication was given, and this was more likely with increasing age.
Garrison found that incidents were less likely to occur during the first 5 days of a patient's stay than during any 5-day period thereafter [14]. There was no decline in the rate of incidents as patients approached discharge (although this was seen with some individuals). Prior admission and longer duration of hospital stay were found to be predictive of seclusion and restraint use. Staff were more often the target of aggression than peers; and male staff more often than female staff. Aggression directed toward staff was more likely to result in seclusion and restraint. The possible confounding factor of male staff caring for the more aggressive patients was not addressed. Seclusion was most frequent on Mondays and on the ward's busiest day (Wednesday or Thursday) with clinical meetings and greatest personal interaction and program activity. It was lowest on Saturdays and Sundays. Over the day, seclusions occurred most frequently on either the day or evening shift and rarely at night. Generally, the most clinically sophisticated staff are present during the day, challenging the view that seclusion rates are a measure of quality of care. Times of increased risk were mornings and evenings, and transitions during the hospital day. These times are characterized by demands for age-appropriate functioning and heightened staff/child interaction. Precipitating incidents have been found to occur most either in settings with higher behavioural expectations (e.g. classrooms) or in areas where behavioural expectations were less clear (e.g. hallways and patient rooms) and during unstructured activity.
Low morale, staff conflict or disruption, lack of support from or changes in administration and threat of closure have all been thought to contribute to increased rates of seclusion and restraint. Goren reported the outcome of a project that arose from concerns about the frequency of seclusion and restraint use, the high numbers of injuries and the general atmosphere [5]. High rates of use were related to staff's perception that the units were unsafe, poor communication across disciplinary and team boundaries and low thresholds for use including for verbal threats and non-compliance. When these issues were addressed, there were substantial decreases in the use of mechanical restraint (98%), seclusion (50%) and medication. Kalogjera described the introduction of a therapeutic management strategy aimed at correcting deficient coping skills, internalizing controls and addressing developmental needs [21]. They achieved a 64% drop in seclusion use. Recommendations for decreasing seclusion use include attention to the philosophy and culture of a unit and the role of leadership, building staff skills, involving parents and treatment planning that decreased environmental and interpersonal stimulation, and planned for the therapeutic management of brief, destructive loss of control.
Political policies and legal mandates
Political policies and legislative changes influence the use of seclusion and restraint and changes are often driven by social rather than clinical factors. Swett et al. described changes in the rates of restraint and seclusion on a child and adolescent unit after the implementation of a state law that prohibited the use of seclusion [6]. The number of patients, number of episodes and hours of mechanical restraint increased but the total number of hours in restraint significantly reduced. Rates of use of chemical restraint were unchanged. Antoinette addressed the effect of the implementation of a regulation prohibiting the use of locked seclusion with children less than 14 years of age [22]. When locked seclusion was prohibited the use of chemical restraint increased markedly despite the development of other reward and privilege-based behavioural systems, leading them to support the judicious use of locked seclusion.
Patient and staff perceptions of seclusion use
There is debate in the literature about how patients, their families and staff perceive seclusion. Only a few studies have attempted to evaluate how adult populations perceive seclusion and even fewer with children and adolescents. In a study that included 15 children and 13 adolescents Martinez collected data from 69 secluded and non-secluded inpatients [26]; unfortunately the results for children and adolescents were not always separated out. In the non-secluded groups, adolescents had a negative perception of seclusion whereas children had a primarily positive reaction. In open discussion, members of all groups recognized a need for seclusion when patients are out of control or a risk to themselves or others but expressed a stigmatizing effect of seclusion. Adolescents noted that sometimes the ‘wrong kid’ was secluded because staff could not identify who had initiated the trouble. The patients were asked how seclusion could be improved. Children requested a chalkboard, calm music, punching bag and play-dough, while adolescents requested a teddy bear to hold, paper and pencil to write with and a blanket. The acceptability of various interventions for children with severe behavioural problems was investigated by Kazdin [27]. Psychiatric inpatient children and their parents/guardians rated three treatments, based on case vignettes. Parents rated time-out as the most acceptable alternative, medication and seclusion being significantly less acceptable. Children rated medication as the most acceptable treatment, with time-out and seclusion equally acceptable. The author concluded that children perceive medication as non-punitive and innocuous whereas time-out and seclusion are perceived as ‘punishment’. Parents perceive time-out as relatively innocuous and are more aware of the hazards of medication use. Tsemberis and Sullivan found that staff and children preferred seclusion to other management interventions [19]. A postal survey of staff members from 13 child and adolescent psychiatric hospitals [28], found staff rated seclusion and restraint as seldom or only occasionally having a positive effect and the paper comments on the continued use of these interventions, despite lack of belief in their efficacy. There was 70% agreement that seclusion and restraint were appropriate responses to physical aggression and 90% agreement that they were inappropriate responses to non-compliance, hallucinations or hyperactivity. Approximately 50% of respondents approved the use of seclusion and restraint for self-injury and 30% for threats of violence.
Position statements
In North America, in 1999, the Joint Commission on Accreditation of Health Care Facilities and Health Care Financing Authority issued guidelines for the use of seclusion and restraint. The Children's Health Act of 2000 established national standards for the use of seclusion and restraint with children in psychiatric treatment facilities. The International Society of Psychiatric-Mental Health Nurses (ISPN) published a Position Statement on the Use of Restraint and Seclusion in 2001 [29], While they find the goal of not using seclusion or restraint ‘laudable’ they caution against policies that could risk other adverse consequences (e.g. increase in the use of chemical restraint). The American Academy of Child and Adolescent Psychiatry have published Practice Parameters for the management of aggressive behaviour, with special reference to seclusion and restraint [30], They advocate for prevention strategies that help children and adolescents learn to cope with internal distress and external conflict but allow for seclusion and restraint use to prevent harm.
Limitations of the literature
Most of the literature on seclusion and restraint use arises from North America and describes common use of mechanical restraints to hold the child in the seclusion room or to prevent further injury by the child to themselves or property. This contrasts with practice in Australia and New Zealand where mechanical restraints are rarely used. In the early studies particularly, there is a lack of description of the patient populations served and procedures used. Nevertheless, it is clear that the units are varied. Authors’ opinions on the use of seclusion and restraint influence the way that they report their findings. Possible confounding factors are not always considered. These include the allocation of staff and the role of coercion in child and adolescent units in preventing unacceptable behaviour by the threat of seclusion. The studies that have reported on characteristics of secluded patients and factors related to seclusion and restraint use are often retrospective in design. The majority report on small numbers of patients, the patient populations and disorders are mixed and some studies lack control groups. The role of drug and alcohol use in the populations studied, which may influence rates of seclusion use, has had little comment. The clinical studies are now dated and do not reflect current inpatient child and adolescent psychiatric practice. In North America, managed care has led to a dramatic reduction in length of stay of inpatient care. A focus on acute management in the hospital with follow-up community care and a limited number of beds has also reduced the duration of admissions, and this is seen in other countries. There are few studies that have evaluated how children and adolescents perceive the use of seclusion. Importantly, and despite Garrison's [8] concerns about the increased risk of adverse consequences with this population, there is a lack of published research on the incidence and type of adverse outcomes, for example, injury or death or posttraumatic stress disorder symptoms.
Method
The study was undertaken at an 11-bed dedicated child and youth unit serving a large provincial city and taking patients up to 18 years of age. Ethical approval was gained from the local Ethics Committee.
State law defined seclusion as follows: [seclusion] ‘means confinement alone in a locked room or area during the hours of the day when the patient would ordinarily be allowed to associate freely with other patients’. State law mandates that records be kept of all seclusion use. These records are regularly checked by external agents and provided a reliable source of data. The definition goes on to specifically exclude the procedure known as ‘time-out’ when used as part of a behavioural program for psychological intervention in response to disruptive or unacceptable behaviour. Time-out, equivalent to unlocked seclusion in the literature, was used on the unit; but, as recording of this was not mandated, records were not considered reliable enough for research purposes. Seclusion is used only when danger to self, others or property cannot be reduced by another means. Patients requiring seclusion may be physically restrained by nursing staff in the process of being taken to the seclusion room. Physical restraint is discontinued when the patient is safely in the room and mechanical restraints are not used.
Data was collected anonymously for all admissions of all patients admitted over a 1-year period from the patient records, ward seclusion register and hospital records of patient numbers and staffing. Data was entered on a purpose-designed database and checked for errors. Descriptive statistics including frequencies, means and standard deviations were calculated. Categorical data was further analyzed using conventional non-parametric statistics including χ 2 . Dimensional data was analyzed using analysis of variance (ANOVA) and related parametric techniques. Correlations were calculated using Spearman's rho (rs). Multivariate analysis was undertaken using logistic regression techniques. All statistical tests were two-tailed. The critical value for alpha was set at 0.05 and alpha reduction techniques employed where necessary. As this was an explorative study, an alpha value of 0.05 was used throughout despite recognition of the increased risk of a Type I error. The statistical program SSPS [31] was used.
Results
The study encompassed 105 admissions of 78 patients, with seclusion being used in 34 admissions (32.3%) and 98 episodes of seclusion occurring in total. Twenty-seven patients were secluded during their first admission and seven in subsequent admissions. Seclusion was used once in 14.2% of admissions, twice in 7.6%, three times in 3.8%, and more than three times (range 4–18) in 6.7% of admissions. The average duration of admission was 25.02 (SD 26.83) days, range 1–196 days. The occasional long admissions were related to placement problems and extended admissions of patients with an eating disorder.
The average age of patients on the unit was 14.06 years (SD 2.72). Males were younger with a mean age of 12.63 (SD 3.02) years compared with females, whose mean age was 15.38 (SD 1.66) years. Younger age was significantly associated with seclusion use (p = 0.032). Forty-eight admissions were male and 57 admissions were female. Male patients were statistically more likely to experience an episode of seclusion (p = 0.049). Involuntary admission rates were significantly higher in the secluded group of patients, at 38.24%, as compared to the non-secluded group where only 8.45% were involuntary admissions.
Patients often had multiple diagnoses so the primary diagnosis was recorded for each admission. Diagnoses were condensed into eight categories, based on DSM-IV [32], to allow for meaningful statistical analysis. Patients with a diagnosis in the attention deficit and disruptive behaviour disorders group were significantly more likely to require seclusion (p = 0.006). Most of this difference was accounted for by the difference between those in the attention deficit and disruptive behaviour disorders group and the mood disorders group. Higher proportions of patients with a history of aggression, which was found in 42.9% of admissions, were secluded. There was no difference for patients with a history of deliberate self-harm or suicide attempt, present in 61.9% of admissions.
Previous history of any abuse was recorded then subdivided into three groups: (i) physical abuse; (ii) sexual abuse; and (iii) verbal, emotional and psychological abuse. Patients who were secluded were no more likely to have a history of sexual or emotional/verbal/psychological abuse. However, there was a statistical difference in the rates of physical abuse (p = 0.026). In male patients, there was a trend for those with a history of physical abuse (p = 0.095) to be secluded. In female patients a significant difference was found for abuse (all), (p = 0.01) and a trend for physical abuse but no relationship was found for sexual abuse or emotional/verbal/psychological abuse. Whereas nonabused male patients were more likely to be secluded than non-abused female patients, for those with a history of abuse there was no difference. In 28.6% of admissions a history of exposure to domestic violence was recorded with no difference between the secluded and non-secluded groups.
The percentage of the admission completed at the time of seclusion for each episode was calculated. The number of seclusions is high initially with no ‘honeymoon’ period, and then there is a dip in seclusion use. After the halfway point of the admission, seclusion use is again high but then declines toward discharge.
The reason that was stated for placing the patient in seclusion was taken from the seclusion forms and nursing notes. Reasons cited often included a number of factors; these were grouped by the type of behavioural disturbance:
67 episodes involved physical aggression towards person or property;
46 episodes involved verbal abuse/aggression, threats or harassment;
in 11 episodes it was commented that the patient ignored staff direction or was oppositional;
11 episodes involved uncontrollable, disruptive or obnoxious behaviour;
5 episodes involved self-harm or suicidality;
6 episodes involved property damage;
1 episode followed the patient absconding from the ward.
For some incidents the target of the aggression was recorded; this was staff in 51 episodes and other patients in 12 episodes.
Patients had a mean duration of seclusion per episode of 64 min (SD = 82 min; range 10 min to >9 h). The data is skewed by a few patients who spent long periods in seclusion. The majority of seclusions were short, with only nine being greater than 120 min. In 30 of the 98 episodes of seclusion, time-out was used first but the behaviour continued to escalate and the patient was placed in seclusion. Medication was used concurrently with seclusion in 19 cases.
The highest day for seclusion use was Fridays closely followed by Mondays. Seclusion use dipped mid-week and rates were low on the weekend. Seclusion occurred mainly on day and evening shifts, was rarer on night-shifts and only occurred once between 12 pm and 6 am. During weekdays, seclusion appeared to be related to times of high staff–patient interaction, for example the group times and bed time and lower during visiting time. On weekends there was a more even spread of seclusion throughout the day. No relationship was found between the occurrence of seclusion and the patient-to-staff ratio for the shift.
Discussion
Though seclusion and restraint are widely used, their use remains controversial and opinions vary from them being considered unnecessary evils to therapeutic interventions based on a developmental rationale. All authors recognize the potential for misuse and emphasize that seclusion and restraint be used for appropriate indications, with non-punitive implementation, consistency, debriefing, staff training and regular review. The few studies that look at the perceptions of children and adolescents in regard to seclusion find that these interventions are viewed negatively, although with some acceptance for the need to control dangerous behaviour and prevent harm.
In this unit, approximately one-third of patients experienced seclusion, compared to rates of 19–60% in the literature. In almost one-third of seclusions, the patient went first to time-out but their behaviour continued to escalate and staff proceeded to use seclusion. This suggests alternative strategies are needed in engaging and helping a patient to regain control of the behaviour. Patients can be involved in planning interventions for when they begin to lose control, increasing the likelihood that they will derive therapeutic benefits from the process.
Seclusion use predominantly relates to dangerousness to others. As in the literature, nursing staff were more often the target of the aggression. By the nature of their roles in maintaining a safe environment, enforcing the unit rules, and initiating limit-setting, they become involved when aggressive behaviour starts, placing themselves at risk. The enforcement of rules and so on, may also precipitate aggressive or disruptive behaviour. In this unit, patients with self-harm behaviours or suicidal impulses are not mechanically restrained within the seclusion room so are able to continue to harm themselves. Therefore, suicidal or self-harm behaviour is usually managed with an increase in nursing observations (up to continuous one-on-one) rather than seclusion. Medication was used with seclusion in 19.4% of cases compared to rates of 30–50% in the literature. This usually occurs when patients continue to damage themselves or the seclusion room, but are unable to be removed from seclusion due to danger to others. Medication is also used when dangerousness is arising from psychosis-related agitation. Where possible, however, medication use is avoided for children or adolescents with disruptive behaviour disorders where the patient often settles quickly without medication and the focus is on the development of internal controls.
The average duration of seclusion of 64 min is comparable to the literature (14.5–84 min). The mean was skewed by a few patients who spent long periods in seclusion. Identification of these patients and development of alternative management plans may help reduce the duration of seclusion use. The mean of 64 min is less than that reported in an adult inpatient unit in the same state, where the average duration of seclusion was 3.4 h [33].
The characteristics of patients who were secluded had similarities to previous research, being male and younger in age. This reflected the admissions of latency-age boys with disruptive behaviour disorders who were highly likely to be placed in seclusion. Children and adolescents with disruptive behaviour disorders often have a history of abuse or neglect in their backgrounds. Admission is often precipitated by aggressive behaviour that cannot be managed by their parents or caregivers and their aggressive and defiant behaviour may be aggravated by the perceived rejection admission can involve. Co-morbid learning disorders are common in this population, reflecting the research that shows increased rates of seclusion and restraint in patients with poor language and problem-solving skills. These patients will have fewer resources to cope with the demands and limits of the ward environment. Seclusion was rare in patients in the mood disorder group of which the predominant diagnosis was depression. These children and adolescents are characterized as having an ‘internalizing’ disorder and are less likely to react aggressively. Involuntary admission was also a risk factor for seclusion. These patients are not agreeable to admission and may lack insight into the nature of their illness and be antagonistic to the process of hospitalization. This is likely to increase the risk of confrontation with staff that are placing limits and demands upon them.
Abuse histories proved to be an interesting aspect of the research. The difference in seclusion between male and female patients disappears when those with a history of abuse are compared. Physical abuse predisposed patients to being secluded and exposure to domestic violence did not, so it appears to be the direct experience of aggression that led to aggressive behaviour in these patients. Patients with a history of physical abuse and female patients with any history of abuse should be considered at higher risk of requiring seclusion. Management planning should take this into consideration and care must be taken in the process of containing the aggressive behaviour that the abuse experience not be recreated.
The pattern of seclusion during the course of admission produced results at variance with those reported by Garrison [8]. He found a ‘honeymoon period’ followed by a stable rate of seclusion over time; but duration of admission was greater in his study. In this unit, high rates were found initially, which settled temporarily then increased again in the second half of the admission. We hypothesize that patients test the ward limits on admission then settle down. The increased rates in seclusion use during the second half of the admission may be related to patients having periods of leave from the unit and having to adjust from home life. Some of the children and adolescents were in foster-care or other community placement and some of these and family homes were highly stressed by the child's disruptive behaviour. Occasionally caregivers rejected children after they were admitted. This is likely to aggravate behavioural ‘acting out’ of the patient's anxieties and feelings of rejection or abandonment.
The literature found highest rates of seclusion use on Mondays and the ward's busiest day, which for this unit was Tuesdays. However, our highest rates of seclusion use occurred on Fridays and Mondays. On Monday patients are returning from weekend leave and having to re-adjust to the demands of the unit. For those that did not leave, the unit is still a relaxed, more unstructured place on the weekend. It is unclear why Friday had the highest rate. Possible explanations relate to reduced medical presence on the ward on Fridays, patients ‘winding up’ in anticipation of leave with family and patients who did not have leave due to severity of illness or lack of home to go to, acting out their distress. Review of the time of day when seclusion occurs indicates higher rates at times of higher demand on patients through transitions and group times. Seclusion use was lower during visiting time with a rise as visitors left, which may be related to perceptions of abandonment. This is followed by demands to prepare for bed and to retire.
There are major limitations of the currently available research into seclusion and restraint. The studies are becoming dated, have methodological problems and fail to address some important areas. Political and administrative systems, culture, policies and procedures differ between America and Australia and New Zealand and it is clear that these factors have significant effects on seclusion use. This explorative study aimed to begin addressing some of these deficits but contains a number of limitations. There is an increased risk of a Type I error as an alpha value of 0.05 was used throughout the analysis. The retrospective design meant that only previously collected data was available with possible omissions and lack of clarity but did prevent the possible confounding factor of seclusion rates decreasing because seclusion use was being studied. A higher number of admissions and episodes of seclusion would have been preferred but a time limit of 1 year was set for practical reasons.
Reductions of the use of seclusion and restraint, with elimination of inappropriate use, are important goals; as is maintaining a safe and therapeutic environment for patients and staff. More research is needed on the effectiveness of seclusion and restraint, the types and rates of adverse outcomes that occur and on the development of alternative strategies.