Research Article

Predicting psychiatric admission rates.

BMJ 1992; 304 doi: (Published 02 May 1992) Cite this as: BMJ 1992;304:1146
  1. B. Jarman,
  2. S. Hirsch,
  3. P. White,
  4. R. Driscoll
  1. Department of General Practice, St Mary's Hospital Medical School, London.


    OBJECTIVE--To determine the numbers of actual and expected psychiatric admissions for the residents of the district health authorities of England and to develop a model to indicate which social, health status, and service provision factors best explain the variation of the actual from the expected psychiatric admissions; to use this model to predict psychiatric admission for district health authorities as an aid to resource allocation. DESIGN--The actual psychiatric admission for district health authority residents were extracted from data of the 1986 Mental Health Enquiry. Expected admissions were calculated using the age, sex, and marital status structure of each district health authority and the national psychiatric admission rates related to age, sex, and marital status. Standardised psychiatric admission ratios were calculated as the ratios of the numbers of actual to expected psychiatric admissions. A wide range of social, health status, and service provision data were used as the explanatory variables in regression analyses to determine which combination of factors best explained the variation between districts of standardised psychiatric admission ratios. SETTING--The 168,652 psychiatric admissions recorded for the 1986 Mental Health Enquiry, after exclusion of mental handicap and psychogeriatric admissions. RESULTS--The actual number of psychiatric admissions varied from 79% above to 54% below the expected number of admissions from age, sex, and marital status for the districts of England. The most powerful variables to explain this variation were the rate of notification of drug misusers, standardised mortality ratios, and levels of illegitimacy in each district. A complex model was developed which could be used to predict district psychiatric admissions as an aid to resource allocation. A simpler model was also developed (which was less powerful than the more complex model) based on the underprivileged area score. One advantage of this model was that it could be used at the level of electoral wards as well as district health authorities.