Hospital statistics describe the structural situation of hospitals and prevention or rehabilitation facilities and the utilisation of these facilities by their patients. They are subdivided into three survey parts, namely basic data, cost data and diagnosis data. As part of the basic data, equipment and staffing level are reflected, e.g. beds, large-scale medical equipment, physicians and other staff of hospitals, prevention or rehabilitation facilities and their organisational units (specialised departments). Furthermore, the hospitals show their cost of materials and personnel. Patient-related variables (diagnosis data) include, among others, the main diagnosis, age, sex, place of treatment, place of residence and the specialised department where the patient has stayed longest. The diagnosis data, which provide information on the development of morbidity, are obtained from all hospitals and since 2003 also from prevention or rehabilitation facilities with more than 100 beds. The results of hospital statistics form the statistical basis of many decisions taken in the field of health policy by the Federation and the Länder and serve as a planning base for the institutions involved in hospital financing. Moreover, they are an important part of national and international health monitoring.
How are the data for hospital statistics calculated?
Being designed as decentralised statistics, hospital statistics are compiled jointly by the statistical offices of the Federation and the Länder by means of an annual exhaustive survey whose respondents are obliged to provide information. The survey covers all hospitals and prevention or rehabilitation facilities in Germany . In addition to the traditional survey questionnaire, there has been a special software application since reference year 2003 where data from hospital EDP systems can be read in to reduce the burden on respondents. Data capture is followed by checking procedures carried out at the statistical offices of the Länder. Here errors and non-plausible data are eliminated from the data material. The Federal Statistical Office, which has been entrusted also with the conceptual preparation of the survey, aggregates the results of the Länder to obtain a federal result. While basic and cost data have been shown since 1991, diagnosis data of hospital patients have been collected only since 1993. The first amendment of the legal basis, the Hospital Statistics Ordinance, in 2001 entailed numerous modifications of the survey programme. Among other things, the range of respondents was extended for diagnosis data (see above). In addition, several variables such as the legal form of public hospitals were included into the survey for the first time.
Specific data material can be provided to users by special processing.
How accurate are the results of hospital statistics?
The quality of hospital diagnoses statistics basically depends on the coding behaviour of physicians in institutions. It is difficult for statistical offices to find out if the main diagnoses have been coded correctly in keeping with the International Classification of Diseases (ICD), and by plausibility checks the correctness can be checked only superficially. Moreover, there have been revisions of the ICD. In 2000, the 10th revision of the ICD (ICD-10) replaced the 9th revision in the hospital sector, which had applied until the end of 1999. Such a change affects the comparability of results in time series, which can then be compared only with certain reservations, e.g. on the basis of the European short list. This list permits – for selected major diagnoses – the conversion of ICD-9 to ICD-10 codes.
