Mental health services in Australia

Seclusion is defined as the confinement of a patient at any time of the day or night alone in a room or area from which free exit is prevented. The purpose, duration, structure of the area and awareness of the patient are not relevant in determining what constitutes seclusion.

Seclusion also applies if the patient agrees to or requests confinement and cannot leave of their own accord. However, if voluntary isolation or ‘quiet time’ alone is requested and the patient is free to leave at any time then this social isolation or 'time out' is not considered seclusion.

While seclusion can be used to provide safety and containment at times when this is considered necessary to protect patients, staff and others, it can also be a source of distress for the patient and support persons, representatives, other patients, staff and visitors. Wherever possible, alternative less restrictive ways of managing a patient’s behaviour should be used, thus minimising the use of seclusion.

Seclusion and restraint may be used across the range of mental health services; however, the focus of the national data collections to date has been limited to the acute specialised mental health hospital service setting, since this service setting has been the focus of many of the associated quality improvement initiatives.

Overview

In 2018–19, there were 11,944 seclusion events nationally in public sector acute mental health hospital services, which represents 7.3 seclusion events per 1,000 bed days. This is an increase from 11,316 seclusion events, or 6.9 seclusion events per 1,000 bed days in 2017–18. Overall, a downward trend has been observed since 2009–10 (the first year of full national data collection – a rate of 13.9 seclusion events). Over the period from 2014–15 to 2018–19 there has been an average annual reduction in the rate of national seclusion events of 2.2%.

States and territories

In 2018–19, the Northern Territory had the highest rate of seclusion in public sector acute mental health hospital services with 13.6 seclusion events per 1,000 bed days, compared with New South Wales, which had the lowest (6.0). Seclusion rates have fallen for three of the states and territories, and risen for four jurisdictions between 2017–18 and 2018–19 (Figure RP.2). However, data for smaller jurisdictions should be interpreted with caution as small changes in the number of seclusion events can have a marked impact on their overall seclusion rate. Further data quality information can be found in the data source section.

Description of figure RP.2 - Source data: Restrictive practices tables (660KB XLS).

Notes: High numbers of seclusion events for a few individuals can have a disproportional effect on the rate of seclusion reported. The increases in the state-wide Tasmanian seclusion rate for 2012–13 and 2013–14 data, and for the ACT in 2017–18 and 2018–19 are due to a small number of clients having an above average number of seclusion events. Victoria's service delivery model produces a higher threshold for acute admission and the seclusion and restraint metrics may be inflated compared to other jurisdictions. Due to the low ratio of beds per person in the NT compared with other jurisdictions, the apparent rate of seclusion is inflated when reporting seclusion per bed day compared with reporting on a population basis.

Frequency and duration

Frequency and duration of seclusion events were collected for the first time in 2013–14. The Australian Capital Territory was unable to provide the number of admitted patient care episodes prior to 2018–19; as such, the national results for the frequency of seclusion during episodes of care up to 2017–18 exclude the Australian Capital Territory and include the Territory for 2018–19. Duration data for South Australia are excluded from the national average duration from 2013–14 to 2017–18 due to issues with the data recording methodology used in South Australia. South Australia is included in the national average duration for 2018–19.

About one in 27 (3.7%) episodes of care provided by in scope Australian public sector specialised acute hospital services involved a seclusion event in 2018–19, a decrease from 2013–14 (5.4%). The Northern Territory had the highest proportion of episodes with a seclusion event (9.4%), while Queensland had the lowest (2.8%). Nationally, there were on average 2.3 seclusion events per episode of care with seclusion, which has remained relatively stable since 2013–14 (2.1).

The average duration of a seclusion event excluding Forensic services, was 4.2 hours in 2018–19, down from 6.0 hours in 2013–14. Forensic services provide services primarily for people whose health condition has led them to commit, or be suspected of, a criminal offence or make it likely that they will re-offend without adequate treatment or containment. Forensic service data has been excluded as forensic seclusion events are typically of longer duration, and substantially skew the overall duration average. Victoria reported the longest average seclusion duration of 5.9 hours per seclusion event in 2018–19, compared with South Australia (1.8 hours) which had the shortest (Figure RP.3).

Description of figure RP.3 - Source data: Restrictive practices tables (660KB XLS).

Notes:

• South Australia reported seclusion duration in 4 hour blocks prior to 2018–19; therefore, average duration could not be calculated for South Australia. South Australia is excluded from the national average seclusion duration from 2013–14 to 2017–18 and included in national average seclusion duration for 2018–19. Comparisons of the national average seclusion duration across time should be made with care.

• Queensland and the Northern Territory do not report any acute Forensic services, however, forensic patients can and do access acute care through General units.

• The Australian Capital Territory now provides Forensic mental health acute inpatient services as of 2016–17 with the establishment of an acute inpatient service (i.e. Dhulwa Mental Health Unit).

• Victoria's service delivery model produces a higher threshold for acute admission and the seclusion and restraint metrics may be inflated compared to other jurisdictions. Higher acuity on admission may be reflected in an inflated average duration for seclusion events compared to other jurisdictions.

Target population

Seclusion data can also be presented by the target population of the acute specialised mental health hospital service where the seclusion event occurred. Around three-quarters (78.0%) of in-scope care (total number of bed days) was provided in General services (unpublished data). Older person services accounted for 14.4% followed by Forensic (4.1%) and Child and adolescent (3.6%) services.

The highest rate of seclusion was for Forensic services with 21.2 seclusion events per 1,000 bed days, followed by Child and adolescent services (14.6), General services (7.5) and Older person services (0.3). General and Older person services show reductions in the rate of seclusion events from 2009–10 to 2018–19, whilst Forensic and Child and adolescent show increases. Variability is apparent from year to year for all target populations (Figure RP.4).

Note: Data presented is the target population of the service unit; that is, the age group that the service is intended to serve, not the age of individual patients. In 2013–14, improvements were made to the reporting of target population categories. The Mixed category was removed as an option for reporting. Data for the Mixed category was most commonly a mix of General, Child and adolescent and/or Older person services. Time series data by target population should therefore be approached with caution. Seclusion and restraint metrics for a small number of Youth hospital beds reported by Victoria, Queensland, Western Australia, and the Northern Territory are also included in the General category.

Description of figure RP.4 - Source data: Restrictive practices tables (660KB XLS).

Frequency and duration

Forensic services had the highest proportion of episodes of care involving seclusion events, with 32.4% of all mental health-related episodes involving at least one seclusion event in 2018–19. This was followed by General (3.7%), Child and adolescent (3.0%), and Older person (0.5%) services. Note that some data for Older person services have been suppressed due to low numbers (Figure RP.6).

Forensic services also had the highest frequency of seclusion, with an average of 5.6 seclusion events per episode when seclusion was used at least once during an episode of care. Seclusion events that occurred in Forensic services also had the longest average duration: 48.1 hours per seclusion event, which is much greater than all other target population categories. This may be due to difficulties in applying the seclusion definition to the forensic context. General services reported an average time of 4.5 hours per seclusion event, followed by Child and adolescent (0.9 hours) services. The average time of a seclusion event decreased for all service types between 2014–15 and 2018–19.

Remoteness

Due to the small number of hospitals located in Outer Regional and Remote areas, for the purpose of remoteness analysis these categories have been combined. There were no hospitals in the seclusion dataset located in Very Remote areas.

In 2018–19, hospitals located in Major Cities had a seclusion rate of 7.3 events per 1,000 bed days. This rate was higher than that for Inner Regional facilities (6.8), and the same as that for Outer Regional and Remote area facilities combined (7.3). The proportion of mental health-related admitted care episodes with a seclusion event was lower in facilities located in Major cities (3.6%), than those in Inner regional areas (3.8%) and Outer regional and Remote areas (3.9%). However, seclusion events in facilities in Major Cities were on average longer in duration (4.3 hours) than those in Inner Regional areas (3.8 hours) and Outer Regional and Remote areas (3.9 hours).

Hospital level data

Figure RP.5 shows the variation in the seclusion rate across Australia in 2018–19. Note that data includes public sector acute mental health hospital services only and excludes forensic units. The variability in seclusion rates between hospitals may be due to a range of factors, such as the hospital’s service delivery model, the number of acute mental health service units in the hospital, the patient casemix, and the target population of the service units. High numbers of seclusion events for a few individuals can also have a disproportional effect on the rate of seclusion reported. Data for individual hospitals should be also interpreted with caution as small changes in the number of seclusion events can have a marked impact on their overall seclusion rate.

Description of figure RP.5 - Source data: Restrictive practices tables (660KB XLS).