Identification and management of data-sets

The data-sets of seven English local authorities that had collected information continuously between 1991 and 1997 were included in the analyses of trends. The local authority information had been collected by the approved social workers who coordinated the Mental Health Act assessments. The way in which the data-sets were identified and how they were made compatible, including the way in which data about ethnicity were recoded, has been described elsewhere (Reference Audini and LelliottAudini & Lelliott, 2002). Diagnosis had been coded using a variety of classification systems. This variable was recoded into five very broad categories — psychotic disorder, neurotic and mood disorders, personality disorder, dementia and learning disability.

Census data were used to standardise the rates of formal admission under the Mental Health Act for specific subgroups of the population defined by age, gender and ethnicity. Data were analysed using SPSS version 8 (SPSS, 1997) and STATA version 5 (StataCorp, 1997). For most analyses, incidents of admission under Sections 2, 3 and 4 were combined to give rates of use of admission sections under Part II of the Act.

Characteristics of the local authorities

The combined catchment population of the seven local authority areas was 2.9 million persons. The mean Jarman Underprivileged Area Score (Reference JarmanJarman, 1983) of the seven areas was 111, and their mean score on the Mental Illness Needs Index (Reference Glover, Robin and EmamiGlover et al, 1998) was 102.2. This suggests that the seven sites were skewed towards the more socially deprived end of the range for all local authorities in England.

Characteristics of those detained

The combined data-set contained 10 926 incidents of detention over 7 years. It is not known how many people this comprises because the data-sets obtained from local authorities do not include unique personal identifiers. It is inevitable that some people would have been detained on more than one occasion. Fifty-three per cent of detentions were of women; 87% were of White people, 7% of Black, 4.5% of Asian and 1.5% of people from other ethnic minority groups. The ‘other’ category for ethnic minority group contained only 158 cases and was considered too small for analyses of trends over the 7-year period. The mean age of those detained was 44.9 years (95% CI 44.6-45.3) and the median age was 40 years. Only three of the broad diagnostic categories contained more than 2% of the cases (psychosis 71%; neurotic and mood disorders 13%; dementia 6.6%). Detentions under Section 2 accounted for 55% of the cases (n=6024), Section 4 for 4% (n=431) and Section 3 for 41% (n=4471).

Overall change in rate of use of Part II

Between 1991 and 1997 the annual rate of use of Part II in the seven local authority areas increased by 32%, from 57 to 75 per 100 000 total population. All of this increase was accounted for by a steep rise in the rate of use of the Act after 1993 (Fig. 1). The rate of detentions under the treatment section of the Act (Section 3) increased by 44% during this period. This was about twice the rate of increase of detentions for assessment (24% for Section 2 and 21% for Section 4).

Fig. 1 Change in number of detentions under Part II sections per 100 000 population for each year between 1991 and 1997.

Differences between subgroups

Why are formal admission rates increasing?

Assuming that it reflects a real trend, the increase in formal admissions during the 1990s has occurred despite a continuous process of reduction in the numbers of psychiatric beds (Reference LelliottLelliott, 1996; Sainsbury Centre for Mental Health, 1998). This apparent paradox is perhaps a result of the shortage of psychiatric beds (Reference Audini, Duffett and LelliottAudini et al, 1999). This creates two types of pressure: first, pressure not to admit patients, perhaps leading to deterioration to the point where admission is required under the Mental Health Act; and second, pressure to discharge patients early, perhaps resulting in a greater likelihood of readmission, and perhaps of readmission under a section of the Mental Health Act. Furthermore, the conditions on acute psychiatric wards have deteriorated (Sainsbury Centre for Mental Health, 1998) perhaps to the point where coercion is more likely to be required to ‘persuade’ patients to reside there. It is likely that this deterioration is in part due to the increased threshold for admission. This creates a concentrating effect, resulting in a case-mix of patients with more severe disability and disturbance on acute psychiatric wards (Reference Lelliott, Audini and DarrochLelliott et al, 1995).

A further paradox is that as bed numbers have reduced, so mental health practitioners have become more safety conscious. The level of awareness of risk management has increased over the past decade. This has been driven by a spate of inquiries into homicides committed by people with a mental illness (Reference SheppardSheppard, 1996) and the preoccupation of successive governments with protection of members of the public from attacks by people with a mental illness. This preoccupation is demonstrated by the issuing of guidance such as the Care Programme Approach and by proposed new mental health legislation, designed to increase the intensity of care and surveillance of people in the community who are deemed to pose a risk. Thus, although the overall threshold for admission might have increased, it might also have been skewed towards those who are perceived as being more at risk of acting violently and of being non-compliant. It is likely that admissions of such people are more likely to require the use of the Mental Health Act.

The data provide some circumstantial support for this hypothesis. First, the rate of increase in Part II admissions in these seven local authority areas accelerated from 1994 onwards. This was the year when the most highly publicised inquiry into a homicide by a person with a mental illness published its report. The report into the case of Christopher Clunis (Reference Ritchie, Dick and LinghamRitchie et al, 1994), perhaps more than any other factor, led to a greater awareness of risk management by mental health services. Second, there has been a greater increase in the admission under section of men than women and, in particular, of young men. Most of the subjects of homicide inquiries have been men in the age range 20-45 years and it is this group of people with a mental illness who are perhaps most at risk of acting violently, or are perceived to be. It should be noted, however, that data collected routinely by the Department of Health does not show a greater increase in the rate of formal admissions after 1993 (Reference Hotopf, Wall and BuchananHotopf et al, 2000).

Why do so many more formal admissions involve young men?

The total number of people occupying a psychiatric bed fell steadily between 1982 and 1992. This was consistent with a great reduction in bed numbers. However, after slowly falling between 1982 and 1986, the number of people aged 15-44 years who were in a psychiatric hospital actually increased steadily over the subsequent 6 years. Most of the increase in numbers of young people in hospital was accounted for by young men (Reference LelliottLelliott, 1996).

Changes in society have had a different impact on men from that on women in some important respects that might have a bearing on how mental illness presents. For example, the Office for National Statistics of England and Wales (Office for National Statistics, 2002) has reported that the recession of the 1980s ‘had a much greater effect on unemployment among men than among women’ and also that ‘the largest increase in one-person households [over the past 30 years] has been among men under the age of 65’. The great increase in suicide rates among young men over the past 30 years, which has not occurred in the female population, might be another indicator of these changes (Office for National Statistics, 2002).

Are most people with dementia being admitted?

The nearly three-fold increase in the rate of admission under Part II of people diagnosed as having dementia is interesting but should be interpreted cautiously. The absolute number of cases was small and the assignment of diagnosis was made by approved social workers and not by psychiatrists. If the finding is replicated, it might herald the early stages of an effect on the prevalence of dementia as the proportion of older people in the population increases.

The value for service-planning of data about trends

The trends identified in these seven local authority areas, if they reflect changes across the country, support the need for services tailored for and targeted at the population of young men with mental illness, who seem increasingly at risk of compulsory admission and treatment. This small-scale, opportunistic study illustrates the value of trends data for this type of long-term planning and service development, both locally and nationally. The enactment of the proposed new mental health legislation in different parts of the UK will perhaps be an opportunity to develop national systems for the collection of better-quality and more-detailed information about the use of the Mental Health Act.

CLINICAL IMPLICATIONS

1. ▪ The rate of use of Part II of the Mental Health Act 1983 has increased despite a reduction in psychiatric bed numbers.

2. ▪ There is a need for services that are tailored for and targeted at the population of young men with mental illness, who seem increasingly at risk of compulsory admission and treatment.

3. ▪ The enactment of new mental health legislation in different parts of the UK should be accompanied by the introduction of better national systems for the collection of information about formal admissions.

LIMITATIONS

1. ▪ The sample in this study is of incidents of detention under the Act; it is not known how many people this comprises.

2. ▪ It is not known whether the trends identified in these seven local authorities reflect the pattern across the rest of the country.

3. ▪ The finding of a large increase in the rates of detention for people with a diagnosis of dementia is based on a small sample and needs to be replicated.