Suicide now and then … an elusive comparison

Summary

Comparing suicide rates between Victorian and modern times, and the impact of the asylum, should enable a useful historical perspective on how effective our treatment approaches really are. Difficulties include clarifying the ‘social geography’, the underlying diagnoses, the reasons for admission and the reliability of casebook data and follow-up arrangements.

Comparing historical with current data might, on the face of it, seem presumptuous, especially in the field of mental health. The diagnoses, the methods of data collection and the different social and moral attitudes could all be deemed potential confounding factors. However, among the forgotten benefits of Victorian asylum care was the associated institutional insistence on documentation. In England and Wales this was established by statute in the 1845 Asylums Act, which decreed not only that every local authority (e.g. county or borough) should build a lunatic asylum for the pauper insane, if they did not already have one, but also that proper case books should be kept. The details of every patient who was brought into the asylum, including a full admission and discharge register, had to be recorded, and there was a mandatory (and inspected) frequency with which notes had to be made (i.e. weekly, monthly, yearly), depending on the patient's length of stay.

While being most obviously reminiscent of the modern National Health Service insistence on documentation, such as the care programme approach (CPA), the resulting case books have provided us with an extraordinary resource. Although they vary in content, many of them survive in continuing series (e.g. in local archives), and it is possible to obtain a detailed picture of patient presentations, the language used to describe mental states, and the procedures (e.g. of observation and treatment) that were employed. It has also been possible to establish (for a number of the patients) the course of illnesses, and reasonable and comparative diagnoses, based on modern diagnostic criteria. The risks of retrospective diagnosis are well known, and cases need to be carefully assessed, but even at the most parsimonious level of comparison the similarities are robust enough to enable agreement (Turner, 1992).

RESPONSES TO SUICIDE

Apart from studying diagnoses (and by and large the patients in Victorian asylums suffered from very similar core symptoms to those treated today), the reasons for admission, in social and family terms, and the outcomes of treatment can also sometimes be compared with modern times. The study by Healy et al (2006, this issue) has followed in this tradition of comparative historical review by looking at suicide rates. In particular, they have looked at contrasting cohorts of patients deemed to have schizophrenia from a database of admissions between 1875 and 1924, and from the notes of patients admitted to hospital between 1994 and 1998. Since both of these cohorts come from the same area, and were admitted to the same hospital, and there does not seem to have been any significant change in the catchment size of the area population, it is not unreasonable to consider that historically informative differences might be found. In particular, the authors’ basic conclusion, that suicide rates seem to be higher at the present time than they were 100 years ago (or thereabouts), is bound to strike a chord among modern psychiatrists.

The lowering of suicide rates has become a key mental health target (Department of Health, 1999), and regular reports from the ongoing Confidential Inquiry into Suicide and Homicide by People with Mental Illness (Department of Health, 2001) aim to clarify risk factors in this regard, in order (presumably) to improve practice. Although there certainly are known risk periods for suicide – for example, in the immediate aftermath of discharge from in-patient care (Crawford, 2004) – it is nevertheless true that most suicides today are not known to mental health services. This also seems to strongly reflect the Victorian view of suicide in relation to insanity: Tuke (1892) quotes from a colleague who considered that in only about 20% of cases was there ‘any proof that the deceased had shown symptoms of mental disease, so far at least as his friends were aware of the facts’. A modern review, entitled Suicide in Victorian and Edwardian England (Anderson, 1987), also showed that rural rates were higher than urban ones (confounding the Durkheimian thesis of ‘anomie’), and emphasised that any history of suicide cannot be abstracted from its roots in social circumstances.

SOCIAL ATTITUDES

The most obvious problem therefore in Healy et al's comparison is deciding whether the two populations that were served by the North Wales Asylum (that is, the late Victorian and the late 20th-century populations) were truly similar in social terms. The population's size and age range may not reflect significant differences in transiency, employment, social class and family stability, all of which are known risk factors for attempted and completed suicide. Clarification of these would be vital in any case, but given that Healy et al's results indicate that suicide rates in patients with schizophrenia are now 20-fold higher – a disturbing level of difference – one must be sure that the baseline population is truly similar. If their figures are carefully boiled down, they show that in the course of 5 years the historical cohort had 1 suicide in 594 individuals, whereas the present-day cohort had 7 suicides in 133 individuals. Is the asylum genuinely providing a protective anti-suicidal effect? Are these ‘pharmacological life events’? Or are we not comparing like with like?

Certain other difficulties must be considered. For a start, is it possible to compare the verdicts of suicide then and now? Although the legal rejection of the suicide (banning burial in holy ground and even forfeiting all of his or her goods) seems to have ‘gone into desuetude’ by the early part of the 19th century (Tuke, 1892), it is not unreasonable to assume that attitudes to self-killing, particularly in non-urban districts of Victorian Britain, remained traditional. Protecting the family from the stigma of such a pronouncement may have been a significant factor, and therefore the true ‘suicide rate’ may not bear comparison with modern, secular pronouncements. Significant differences in suicide rates depending on the religious background of countries in the 20th century have long been a problem in suicide research, and ‘changing attitudes to suicide… were as complex and variable as the very complicated social and cultural systems in which they existed’ (MacDonald, 1995).

WHAT WERE ASYLUMS FOR?

With regard to retrospective diagnosis, although it is reasonable to accept that those patients who were deemed to have a form of schizophrenia probably did have an illness very similar to what we now call schizophrenia, we do not know what proportion of such patients were admitted. Even today not all patients with schizophrenia are admitted to hospital, particularly in non-industrial countries, and admission depends on the attitudes of and burden to the family, the nature of the symptoms and the degree of social disruption. We know that there was a rising tide of admissions to the asylums between the mid-19th and mid-20th centuries, and there has been considerable debate about the cause of this. Edward Hare (1983) suggested the possibility that the rising numbers of patients with schizophrenia might have been related to urbanisation and some form of low-grade infection, but most other commentators (e.g. Scull, 1979) have suggested that this trend was more socially generated. Once an asylum was available, people became gradually more used to putting away their difficult relatives, and if asylums were cheaper than the workhouse (as they became in the 1870s), the parish guardians would have actively promoted their use. Asylum doctors had to admit whoever turned up with the appropriate certificates, and discharge was a slow legal process.

In addition, the role of the asylum was highly structured, with up to 20% of admissions considered to involve some form of suicide risk. In fact, if one looks at the certificates that were used (i.e. the section forms), doctors often regarded a suicidal tendency as something worth writing down in order to ensure admission. Highly structured arrangements within the wards (e.g. suicide cards) made the staff and hospitals extremely sensitive to a suicide taking place, and woe betide the attendant, superintendent or responsible guardians if suicide occurred within the asylum. Given this bureaucratic stigma (and the likelihood of losing one's job), and given the regular rate of deaths in asylum (e.g. from a range of physical illnesses), is it unreasonable to suggest that it was not in the asylum worker's interests to announce that any of their patients had killed themselves? We also do not know whether suicides that occurred after discharge were actually recorded in the asylum statistics, even though some patients were followed up. Again the stigma attached to documentation of suicide may have affected this procedure.

CONCLUSIONS

Overall, therefore, and especially in the vexed area of suicidal behaviour (MacDonald, 1995), considerable caution is required when reviewing historical comparisons. We may have comparative illnesses (and the fascinating symptom similarity between then and now is essential evidence for regarding schizophrenia as a consistent and valid diagnosis), but the behaviours secondary to those illnesses will be socially generated rather than ‘biological’. It is not as if today we are living in a rising tide of suicide (rates have gone steadily down since the 1970s), even though our attitudes are probably more secular and non-avoidant of the term. Nevertheless, clinicians who are faced with the modern, revolving-door, usually male, often drug-dependent and variably insightful patient with schizophrenia may wonder what lessons are to be learned from this paper. Can we re-introduce the good bits of asylum practice into modern community care? Were some aspects of institutions in fact genuinely therapeutic? How effective are modern medications in suicide prevention (as opposed to symptom alleviation)? And should we be making more use of the data – and lessons – of history when organising care for people with severe mental illness?