868
Mental Health SUICIDE AND SAMARITAN CLIENTS B. M. BARRACLOUGH MARIAN SHEA Medical Research Council, Clinical Psychiatry Unit, Graylingwell Hospital, Chichester, Sussex
Samaritan clients are most at risk of suicide in the year following self-referral when they have a high suicide-rate of 357 per 100,000. Of the 45 suicides studied 71% (32) killed themselves within a year of self-referral, and 44% (14) of these 32 committed suicide within their first month as clients. About 4% of all suicides are former clients. Evidence suggests that Samaritan branches may differ in their ability to postpone suicide. Summary
INTRODUCTION
The Samaritans are a voluntary organisation whose declared aim is the provision of help, in a rather special way, for the suicidal and despairing. Conceived by the Rev. Chad Varah, the Samaritans began their work in the crypt of the Lord Mayor’s Church, London, in 1953. Today 105 branches in the United Kingdom provide most major cities, and many smaller towns and rural areas, with Samaritan services. The 50,000 people who last year requested the Samaritans’ help a need. The essential feature of the service is to establish a special kind of friendship between the client " and the Samaritan worker or " counsellor ". Through this " befriending " process it is hoped to resolve some of the clients’ problems by giving support with friendliness and by fostering the kind of personal changes which occur in psychotherapy, or through practical help in re-establishing social links. Besides the skills required for these two approaches the counsellors are taught to recognise and refer mental illness for psychiatric treatment. Other problems not mainly psychological-for example, clients in financial straits -are also passed on to appropriate agencies. Whether the Samaritans attract the suicidal in sufficient numbers to make suicide prevention a worthwhile aim for them is uncertain. The distribution of age, sex, and marital status among their clients is quite unlike that of suicides-younger age-groups, women, and the married predominating. Their statistical returns show a history of mental illness and suicide in near-relatives of Samaritan clients to be less common than in the general population, whereas it is higher among suicides. So there is evidence that Samaritan clients as a group do not resemble suicides. However, there is a study which suggests that they prevent suicide. Bagleycompared the suicide-rate in county boroughs before and after the introduction of a Samaritan service and, controlling with a group of ecologically similar county boroughs, demonstrated a relative reduction of nearly 20% in the suicide-rate of boroughs where the Samaritans were operating. That is, an association was established between a town having a Samaritan service and a reduction in the
impressively signify
suicide-rate; a causal link is clearly a question of judgment, since social changes could be responsible for both the appearance of a Samaritan branch and the change in the rate. There is therefore no certain evidence that the Samaritans either attract the potential suicide or, having done so, prevent his death. We decided to approach the problem more directly. We aimed to calculate a suicide-rate for Samaritan clients, secondly to discover what proportion of all suicides the Samaritans had been seeing, thirdly to examine the period elapsing between self-referral and death, and fourthly to compare Samaritan branches to see if they differed in their ability to prevent or postpone suicide. METHOD
The Samaritan suicides were identified by comparing the coroner’s list of suicides with the Samaritan client register in each of six towns in Southern England. 39 suicides were identified in this way; 6 more were found because, although they had died outside the coroner’s areas we examined, the Samaritan branches knew of their deaths. Newspaper cuttings of inquests provided proof of their suicide. The towns examined were six county boroughs in Southern England. We chose these places because they were close at hand and their coroners and Samaritan branches both agreed to help. Other nearby towns initially selected for inclusion had to be left out because either the Samaritan branch or the coroner felt unable to take part in the study, owing to ethical considerations. For each suicide the coroner’s inquest notes provided the date of death and the Samaritan records the dates of self-referral and subsequent contact. The branch records also gave us the total number of new clients seen each year from the branch’s inception, and these numbers comprised the population used to calculate the suicide-rate of clients. RESULTS
Suicide-rates The average annual rate for Samaritan clients (table i) is 119 per lDO,OOO. The numerator for this TABLE 1.—SUICIDE-RATES
PER
100,000
FOR SAMARITAN CLIENTS
* 3 suicides occurred in 1969 and are excluded as referrals in that year are unknown. t 3 suicides which took place within their first year as clients occurred in 1968. They are excluded because the 1968 clients had not a
complete year’s experience as clients. Calculated from (a) the Registrar General’s Decennial Supplement, England and Wales (1961), area mortality tables; (b) Census (1961) England and Wales. § Calculated from (a) the age and sex distributions of a random sample of Samaritan clients in the towns studied; (b) the Registrar General’s Decennial Supplement, England and Wales (1961), area mortality tables: suicide-rates for Southern Region of England. -
869
is the 42 suicides between 1961 and 1968; the denominator, the sum of the number of patients at risk in each of those years. The rate which gives a better estimate of the suicide risk of clients is that calculated from the suicides in It is the twelve months from their self-referral. 285 per 100,000. The numerator for this rate is the 26 suicides which occurred within a year of referral; the denominator the sum of the new referrals for each year-that is, the population who had experienced one complete year as clients. Table I shows this rate and those for subsequent client years, indicating a diminishing risk for suicide the longer the time from self-referral. These rates are an underestimate of the true rate, for two reasons: firstly, some Samaritan suicides will not have been located because they killed themselves outside the coroner districts we examined (6 of these were found but no systematic search could be made for the others); secondly, a proportion of Samaritan clients are anonymous and so their deaths cannot be ascertained, but they are included with the branch returns of the number of clients seen on which the suicide-rates are based. This has the effect of artificially lowering the rate. The underestimate of the suicide-rate caused by error arising from the first source is unknown. But the proportion of anonymous clients for two branches is known to be a third and a twentieth. For the remaining four branches it can be estimated as the mean of these two proportions-that is, a fifth. We corrected the client population by subtracting the appropriate proportion of anonymous clients for each branch and then calculated a second set of rates (table i). We feel that the corrected figures are closer to the true rate than the uncorrected-although still an underestimate. What conclusions can be drawn from these rates? Firstly they justify the belief that the Samaritans are attracting the potential suicide, for the national rate for people over 15 years of age is 15 per 100,000 compared with the Samaritans’ 150. Secondly they demonstrate a risk of death from suicide for Samaritan clients that is thirteen times that for the general population of comparable age and sex; a risk that in the year immediately following self-referral is thirtytwo times higher than for ordinary people. Thirdly the rates show a suicide risk that decreases in the years following self-referral, although even in the third year the suicide-rate is still about seven times that for the general population. Fourthly, comparisons between the suicide-rate of Samaritan clients and that of other defined groups provide a measure of the relative risk. Attempted suicides have an annual suicide-rate which has been estimated as 1300 per 100,000,2depressive illness patients 300 per 100,000,3 and former mental-hospital inpatients 48 per 100,000.4 Samaritan clients appear to have a risk between that of former psychiatric inpatients and that of patients diagnosed as having depressive disorders. rate
Proportion of all Suicides Consulting Samaritans 929 suicides occurred in the six coroner districts during the period under review, and 39 (4-1%) of
Interval between self-referral and suicide in 45 former Samaritan clients.
these were Samaritan clients at some time previous to their deaths. If this proportion is constant throughout the country, Samaritan branches must be seeing about 200 people each year who subsequently commit suicide. This is a measure of the potential reduction in suicide that could be effected by the Samaritans, in the existing circumstances. Time between Self-referral and Death We looked at the time elapsing between first contact with the Samaritans and death-first in years and then in months. 40 (nearly 90%) occurred within the first two years, and 32 (71%) in the initial twelve months. This concentration is most distinct for the month following referral where one-third (31%) are accounted for (see accompanying figure). In these cases dying within one month the suicidal crises must have been particularly intense and either have passed unrecognised by the Samaritans or not been amenable or accessible to their treatment. When the last medical contact of 100 consecutive suicides residing in West Sussex and Portsmouth was examined, a similar but more intense concentration was found, 71% having died within one month of having seen a doctor.õ In these terms Samaritan workers appear to do rather better than doctors, but the circumstances of the contacts are not comparable. A cluster of 8 suicide deaths occurred with a mean time of 17-8 months from referral (see figure). This does suggest that the initial experience with the Samaritans for these people may have been preventive and that they succumbed to a subsequent episode either of an illness or some suicidal crisis of a personal or social character.
Branch Differences Another observation suggests the Samaritans are postponing or preventing suicide. Table II compares branch A with the other five branches for the proportion of all its client suicides dying within a year of self-referral and in subsequent years after the first. TABLE IL-TI112E BETWEEN SELF-REFERRAL AND SUICIDE: BRANCH DIFFERENCES IN THE NUMBER OF DEATHS WITHIN THE FIRST YEAR
870
Significantly fewer died in the first year, and of 5 that did die subsequently, all did so in the second year and comprised the majority of the 8 second-year suicides mentioned in the previous paragraph. This branch alone succeeded in this way with a minimum of first-year suicides-suggesting that their methods are more effective than those of the other five branches. This is a valid conclusion only if their clients are similar. They are, when compared with a random sample of clients on the measurements available to us -namely, age, sex, marital status, social class, living alone, and history of previous psychiatric treatment. DISCUSSION
What are the implications of these findings for the Samaritan organisation? There seem to us four major points worth consideration. First, it is clear that the Samaritans are attracting the suicidal: our study shows that 3-4 per 1000 of their clients kill themselves during the twelve months following self referral-a high suicide-rate compared with the general population. However, it is a very low proportion of all clients seen. The detection of the suicide-prone must be difficult for the Samaritan worker if 99-5% of clients are not going to take their lives. A rating scale predicting the likely suicide would enable the Samaritan worker to concentrate effort on those most at risk. A future paper will describe our attempt at such a scale. The second point concerns branch differences. The average time between self-referral and death differs between branches; one branch does significantly better than the other five. The cause of this variation must be either differences in the procedures used by branches to treat clients, or differences in the urgency of the suicide drive of the different groups of suicideprone clients. On the available evidence we are not able to distinguish absolutely between the alternatives ; nevertheless from the practical point of view it seems worth while assuming that procedural differences are responsible. Then a comparison of Samaritan branch procedures with those of the branch that did best might produce some valuable knowledge about preventing or delaying suicide. The third point concerns the despondency and disappointment which the present results may cause Samaritans. If this is their response it seems to us the wrong one, for the results cannot refute their belief that they are preventing suicide and they do prove that the suicide-prone are attracted to the Samaritans-a fact previously not accepted by everyone. The high suicide-rate of clients must now ensure attention from serious people to their aims. Finally, we suggest that this exercise-the location of suicides, formerly clients-should be undertaken annually by each Samaritan branch, to see whether more can be learned about the control of suicidal clients through reviewing the events of each case. This inquiry would have been impossible without the cooperation of H.M. coroners and the directors of the Samaritan branches in the six county boroughs studied who generously made their records available to us; we are very grateful to them. The Samaritan research advisory committee provided valuable advice and support. References at foot of next column
Points of View CONCEPTIONS OF MEDICAL CARE CONFLICT BETWEEN CLAIMS OF INDIVIDUAL AND OF SOCIETY
PHILIP RHODES
Department of Gynœcology, St. Thomas’s Hospital Medical School, London S.E.1 CULTURAL phenomena underlie all our actions. We escape the fact that we have been conditioned by the environment in which we have been brought up. But the mores, the customs, the culture are largely inexplicit and not defined. Certain things are done or not done in varying situations but the definitions of what should and should not be done are vague. Although there is a generally accepted code of conduct in any community, it is subject to varying interpretations by individual members of that community. We are all members of the nation but within that nation, accepting an ill-defined morality, there are groupscannot
religious, political, social, family, neighbourhood,
and
factional-which differ in their attitudes to a variety of problems. There is a general core of culture with specialised interpretations at the periphery. IMPORTANCE OF THE INDIVIDUAL
Sick people have to rely on others for their needs. In primitive societies those who are ill turn to their relatives and friends. But in virtually all societies there are those to whom the relatives and friends can go for expert advice, those with special knowledge of illness, gleaned through special observation and reinforced by magic, superstition, and propitiation of supernatural forces. The medicine-man and the priest had a special place in the early history of medicine. The sick summon up the resources which society has to offer, which have to be paid for in one way or another. The rich and the powerful have a better chance of commanding those resources than others, while the poor and the weak may go to the wall for lack of such command. Greek philosophy gave a rational basis to the importance of the individual, and Judaic Christianity gave the concept emotional force.The value of his own life is fundamental to the activity of each individual, even though the intellectual may have passing doubts about it, the old have more lingering doubts, and suicides reject their own value. However much there may have been failures through history to act fully in accordance with the concept of the importance of each individual life, as in the Inquisition, in wars, and in the myriad examples of man’s inhumanity to man, there has
1. 2. 3. 4. 5.
DR. BARRACLOUGH, MRS. SHEA: REFERENCES Bagley, C. Soc. Sci. Med. 1968, 2, 1. Greer, S., Lee, H. A. Acta psychiat. scand. 1967, 43, 361. Thomson, I. G. Proceedings of the Fifth International Conference for Suicide Prevention, London, 1969, p. 140. Temoche, A., Pugh, T. F., McMahon, B. J. nerv. Ment. Dis. 1964, 138, 124. Barraclough, B. M., Nelson, B., Bunch, J., Sainsbury, P. Proceedings of the Fifth International Conference for Suicide Prevention.
London, 1969, p.
129.