The treatment demand for bu IÂ¡mia:a catchment area report of referral rates and demography

district regional on

In the light of proposals to change the system of providing NHS health care, it is important for future district or regional procurers to have reliable data on which to base their treatment 'contracts', bearing in mind not only current but also the potential demand of a mature market.
To my knowledge, there have been no attempts to measure treatment demand for bulimia. There are a number of reasons for this: the condition has only recently been defined; research centres with an interest in bulimia draw patients from a wide area with no consistent pattern of referral; and, many patients 'shop around' for treatment making it diffi cult to be sure that all patients from a particular area are included.
To give a true indication of potential NHS service demand, the treatment service should be 'mature' and have had a consistent referral policy and diag nostic criteria throughout its existence. Secondly, the patients should come from a circumscribed geographical area whose demographics are represen tative of a common type of British catchment area, and certainly this is so if results are to be generalised. Thirdly, it is helpful to know, when dealing with a catchment area population, who initiates the referral as well as the referring agency for if a treatment clinic is set up where previously there is none, information should be targeted so that services are taken up quickly and efficiently.

Catchment area
The circumscribed catchment area studied included the London Boroughs of Wandsworth and Merton. These urban boroughs, which have a population of 175,000 women, form a triangular wedge of South-West London, with Battersea at the apex, spreading out through Clapham, Balham, Tooting, Mitcham and Wimbledon. There are some areas of deprivation and others of great opulence but, in the main, they are typical city suburbs being terrace, villa and semi detached family dwellings or 'flat' accommodation with a stable population. The Bulimia Clinic is at St George's Hospital which is centrally situated with good transport links.

Case material
All patients were interviewed by me and were under my consultant care. To be considered an index case, all patients were diagnosed by me as having bulimia nervosa according to the criteria of DSM-III (APA, 1980) and DSM-III-R but excluding patients who have anorexia nervosa, by the criteria set by Crisp (Crisp, 1980)and also excluding the massively obese, defined as being over 50% mean-matched population weight (MMPW).

Procedure
Only patients referred by a doctor were offered an appointment. It was the practice to see patients whose doctors reported that their patients binged with food and were within a normal range of weight. All patients who binged and were reported to be overweight or underweight (see above) were still seen but in separate clinics. If they were subsequently diagnosed as having normal-weight bulimia they were transferred to the Bulimia Clinic. All patients completed extensive demographic and clinical his tory questionnaires and standardised eating disorder questionnaires. All patients were asked to attend for a two-hour assessment interview and be accompanied by an informant. Weight and height were measured in the clinic.

Findings
The clinic was established in February 1980. Data are presented up to and including January 1990, the figures being for years running February to January inclusive. Changes in referral patterns during the decade are shown by comparing the referrals during the 40 months February 1980 to May 1983 and September 1986 to January 1990 inclusive. As the condition overwhelmingly affects women, female referrals only are detailed. It could be argued that the annual referral rate should be based, not on the total female population of Wandsworth and Merton, but on the population of women from which the patients stem: women aged 15 to 40 years. In the catchment area, 42.4% of women, or 74,200, are aged within this range. Table I gives the annual rate of referral over the decade to January 1990.The Table gives the number of patients from the catchment area seen each year, expressed both in total and as an annual rate per 10,000 women. The number of patients assessed and diagnosed increased each year with the exceptions of 1984 and 1988. The implication is, that over the four years from 1986 and with referral uninfluenced by private payment, advertisement or referral restric tions (except a medical referral letter) just less than one new case in approximately 1200 women, aged 15 to 40 years, sought help from the clinic and were diagnosed as having bulimia. In the last year, there was one new case in 1000 women.

Source of referral
The majority of patients were referred either by their general practitioner (71%) or a psychiatrist (21%). In the 40 months at the beginning of the decade, twothirds of patients (66.1%) were referred by their GP and just less than a quarter (24.1%) referred by a psychiatrist. Eight percent of patients were referred by a physician. During the last 40 months of the decade, most patients were referred by their GPs (76%) with only (19%) referred by a psychiatrist. No patients were referred by physicians, gynaecologists, or surgeons.

Initiation of referral
Over half the patients initiated their own referral by asking their doctor for more specialist help; the frequency, however, varied sharply with the referring agency. Thus 77% who came directly from their general practitioner requested referral themselves, just less than half of whom (45%) had previously contacted the clinic, 18.5% referred by a psychiatrist requested onward referral but none claimed previous knowledge of the clinic. None of the patients sent by a physician requested onward referral to the clinic. Relatives and friends took a comparatively minor role initiating referral patients. Of the total, 5% had been influenced and encouraged by their parents (particularly their mothers), 3% by an elder sister and 1% by a female friend. No husband initiated referral, but, on the other hand, none hindered it either.

Demographics
The mean age of the patients was 24.8 years (s.d. = 5.0) with a range from 16 years to 40 years. The modal five-year cohort was 20 to 24 years. The mean weight was 60.9kg (s.d. = 8.7) and was not normally distri buted. The average MMPW was 105.5% (s.d. = 14.3). No patient was below 85% MMPW and 60% of the sample were over their MMPWs. Over 95% of the patients were either in employment, in full-time education or caring for their children. Three-quarters of the series were single. 21.5% were married but 3.6% were separated and a similar number were divorced. Less than 3% were non-white; 3.6% of the clinic patients were from social class 1(Wandsworth and Merton = 2.1%); 28.6% from S.C.II (22.9%); 44.6% from S.C.III (51.6%); 6.3% from S.C.IV (14.8%); and 17% S.C.V and unclassified (8.8%).

Comment
This study mirrors normal referral practices in the United Kingdom; that is, a patient is referred with out prompting by a local GP on the basis of the GP's judgement alone. Advertisement, even if only a letter to a GP requesting referrals, would have two effects I wished to avoid. First, referral rates, in any one year at least, would be artificial. Second, as bulimia fluctuates particularly in its symptoms, elicited referrals may present differently from those prompted by medical judgement in the natural course of the illness.
This study demonstrates that bulimia represents a substantial reservoir of psychiatric morbidity. The condition is common: one new case each year in 1200 women aged 15-40 years. If these figures can be generalised and if the average urban consultant's catchment area has 40,000 persons (Dr F. Caldicott, personal communication, October 1990) then each would have, very approximately 8500 women, aged 15to 40 years, giving rise to 7 new referrals each year. It is not possible to determine accurately an incidence rate from these dataâ€"clearly there are bulimics who do not wish to see a psychiatrist. However, these figures are more relevant to planners and clinicians for they indicate the number who eventually seek help and for whom provision should be made.
The clinic population comes from a broad range of social classes, generally similar to the pattern of social class in the catchment area itself. In this respect the condition is quite dissimilar to anorexia nervosa. It is therefore beholden on the NHS to provide treat ment centres for the vast majority of patients have no private alternative. The percentage of students (9%) suggest that the various surveys have been done fortuitously on a population in which bulimia is particularly prevalent.
The women attending the clinic were young and less likely to be married (21.5%) than their sisters (42.8%) or similarly-aged women in the District (51%); OPCS, 1982. The MMPW of the patients in this survey was 105.3%. Statistical comparisons for weight and height with the general population are inappropriate. A proportional survey based on OPCS (1982) data by age, social class and region suggests that the clinic population is slightly taller ( 1.65 m; 1.62 m) and heavier (60.9 kg; 59.5 kg).
Most referrals come from general practitioners and formed an increasing percentage of total referrals during the decade. Many patients initially contact the clinic direct but, on being told that assessment and treatment can only be done by way of medical referral, seek help from their GP. This approach does not deter prospective patients. The data seem to suggest that if its is desired to encourage referrals by advertisement to a clinic, then the "target groups" should either be the patient herself or general practitioners, and not psychiatrists or general physicians.
Referrals from physicians have, over the decade, fallen away as the number of direct referrals from general practitioners have increased. The increasing 205 number of referrals directly from GPs has the impli cation that medical and psychiatric colleagues are seeing less and less bulimia as the symptoms are picked up in general practice and referred directly to the clinic. Colleagues may, therefore, have the erroneous impression that the disorder is becoming more rare, while the opposite is the case. It may there fore get increasingly difficult to fund such services out of hard-pressed hospital budgets.
New legislation should mean that purchasing DHAs keep away from the internecine squabbles of the different provider units. However they must prioritise the services they purchase. Bulimia has substantial morbidity if left untreated and very effective treatments exist (e.g. Lacey, 1992). This paper makes clear that there is strong, if latent, demand across all social classes. At the end of the day I suspect that strong political pressure from the patients themselves will unlock the door to proper treatment services.