The Mental Health Commission (MHC) recently published its ninth report on the use of seclusion and restraint (2019a). This document highlights that there has been an increase in both the use of physical restraint and the duration of seclusion over the last 10 years. This was reported in the national media as demonstrating an increase in the use of coercive practices (Cullen Reference Cullen2019). One of the other striking findings contained in this report was the high degree of heterogeneity between the different units and Community Healthcare Organisations (CHOs). Levels of seclusion in the nine CHOs ranged from 9.4 to 70.3 per 100 000 population and levels of restraint ranged from 10.1 to 132.0 per 100 000 population. The report acknowledges that comparison is complicated by a wide range of factors, including culture, practice, staff training, staffing levels, and the severity and prevalence of mental illness.

There is a strong movement internationally to prohibit all involuntary treatments in mental health, which was greatly bolstered by the Convention on the Rights of Persons with Disabilities (CRPD) (2006) which Ireland ratified in 2018. The World Health Organization (WHO) has now embraced CRPD informed policies and, through its QualityRights Initiative, is overtly pushing for an end to all coercive practices (WHO 2019). The QualityRights documents echo the comments of the United Nations Special Rapporteur on torture (United Nations Human Rights Council 2013) who described seclusion and restraint as torture. Consequently, there is a strong onus on Irish mental health services to comprehensively understand the factors that influence this level of variation in the use of seclusion and restraint so that steps can be taken to reduce coercive practices. Four modifications to the seclusion and restraint reports would greatly enhance the quality of the data and, consequently, the conclusions that can be drawn.

First, the report contrasts rates of seclusion and restraint from CHOs that have varied access to seclusion rooms. Levels and duration of restraint may be longer in centres without such facilities, and this may explain some of the observed variation. A CHO with an approved centre without a seclusion room may have lower rates of seclusion compared to a CHO in which all centres have seclusion rooms. However, in this case, the variation would reflect the services available rather than the use of coercive practices. It would be more informative to only analyse centres with seclusion rooms or to include a discussion on the more comprehensive data in the appendix of the report. Stratification by ‘access to a seclusion room’ would add clarity to the report.

Second, meaningful comparison between the units and the CHOs is inhibited by the fact that there is no indication of the severity of illness experienced by the individuals attending each service. The wide variation in rates of involuntary admission suggests that different CHOs may have different burdens of severity of mental illness (Mental Health Commission 2019b). This would be hard to collect data on directly; however, HIPE (Hospital In Patient Enquiry) codes would provide some limited information. There are also multiple proxy markers that could be used to give an indication of the acuity of the presentations. For example, days of involuntary admission on the unit, average length of stay (Nielsen et al. Reference Nielsen, Lolk, Valentin and Andersen2016), percentage of all bed days that are occupied on an involuntary basis, reasons for admission or diagnosis on discharge could provide information that would enhance the generalisability of the data.

Third, in addition to calculating seclusion per population levels in each CHO, levels per bed number or per involuntary days may provide a more informative metric. Levels per bed number can be calculated from the table in appendix 3, but in its current format, direct comparison is not convenient. These would partly give consideration to the level of acuity seen in the different CHOs, and it would also allow a comparison with international levels measured using different denominators.

Fourth, the report makes no mention of restraint brought about through the use of sedating medication in the absence of a clinical indication. The WHO’s QualityRights Initiative refers to this as ‘chemical restraint’ and sees it as a central component of restrictive practice (WHO 2019). The use of pharmacological agents on individuals represents a major confounder in evaluating the use of seclusion and restraint. The collection of these data would require the largest modification, at least for voluntary patients. For involuntary patients, however, some data should exist as the number of service users administered involuntary medication and the number of administrations of involuntary medication used each month are two of the items that the MHC has instructed all approved centres to record (MHC 2014). If these data are being collected, it would be highly informative to see it included in any reporting on seclusion and restraint. For voluntary patients, any data recording the levels of medication used on the ward would give some indication as to the level of chemical restraint being employed. Even a crude measure would be highly informative, for example, intramuscular medication used on an involuntary basis, or frequency of ‘as needed’ antipsychotic use, or total dose of antipsychotic used per unit beds. These measures would have severe limitations, but significant trends or highly heterogeneous patterns of use may stimulate further research. The omission of information on chemical restraint makes the rest of the data contained in the report impossible to interpret.

National and international pressure is mounting to reduce, and even prohibit, the use of all coercive measures in the treatment of psychosocial disabilities. Collecting data on the use of seclusion and restraint is a vital first step but developing a deeper understanding of this data is required to modify current practices. The data we have on seclusion are highly informative and recently has demonstrated worrying trends. The high level of variation is a fascinating observation, but due to the nature of the data collected, it is of limited use. Modifications to the data collection, both straightforward and more complex, would greatly enhance the utility of these data and may lead to identifying steps that could help bring the practice of mental health care in Ireland more in line with the CRPD.