Geographical analysis of the risk of psychiatric hospitalization in Hamburg from 1988-1994

The analysis of the geographical distribution of hospital cases is obviousl y important for the purpose of planning hospital services, but it is of eve n greater significance in the planning of psychiatric services. This concer n motivated our seven-year-long study, which examined hospitalization risks among various categories of psychiatric disorders in the major German city of Hamburg. Our database encompassed 77% (n = 64,000) of all psychiatric a dmissions in a total of 41 hospitals, most of which are general hospitals. In order to carry out the geographical analysis we employed a new statistic al method based on a mixture distribution model. According to our findings, the strongest indications of an increased frequency were among male cases of schizophrenia drug abuse and organic psychoses, and female cases of neur otic disorders, personality disorders, drug abuse and schizophrenia. We fou nd that some areas are exposed to a risk of hospitalization for these diagn ostic categories which is more than 50% above the reference. Contrary to ot her authors we did not identify an increased frequency of admission concent rated in the inner-city area far any of the diagnostic groups. The risk of hospitalization for schizophrenics was almost entirely associated with the close proximity of psychiatric units, while the risks for neuroses and pers onality disorders, as well as alcohol and drug abuse, appeared to be concen trated in areas of low social status. However, a statistically relevant cor relation between an increased risk of hospitalization and low social status could be determined only for drug abuse and alcoholism. In the end, we did identify two areas in which there was an increased risk of hospitalization for several diagnostic groups, and this information will undoubtedly facil itate the planning of hospital and psychiatric services. The fact that our findings deviate to some extent from other authors - especially with respec t to neuroses and personality disorders, but also to addiction - can be att ributed to the inclusion of psychiatric cases from general hospitals in our geographic analysis. (C) 1999 Editions scientifiques et medicales Elsevier SAS.