Some considerations of multi-variate hospital statistics

At the University Hospital in Kiel data on the medical care and administration of individual patients is collected in a hospital information system. The description leads to lists of all cases treated, arranged according to basic data descriptions, and to tabulations of aggregated data. The aim of these statistics is to provide an overview of the collected data. At present the statistics can do little more than offer some general impressions about the frequency of serious diseases requiring hospitalization and health-care needs of a particular region; this is because they are selected and therefore biased and misleading. An accurate epidemiological view of a region to evaluate the quality, efficiency and cost of its health-care system can only be obtained if all the hospitals in that area are producing summarized information. So correlation of population statistics, broken down by patients' residential area and the hospital means for evaluating the health-care demands and consumption of a particular population, and for regional planning when figures for catchment area and attraction rate are produced.