Unwinding Ariadne's thread: some methodological and interpretative considerations on international hospital admission...

Hospital utilisation rates comprise the main available source of information on morbidity patterns and the need for health care in a number of developing countries. On the basis of the findings of a Greek-Swedish comparison of somatic hospital admission (presented per the 18 principal head groups of the International Classification of Diseases) between two regional University hospitals, a number of important findings and methodological questions are considered. Factors that can explain the large variations of hospital admission found are examined; in particular those related to disease definition, classification and coding, hospital services supply, professional behavior and practices, illness behavior of patients, and demographic and socio-economic characteristics but especially a true difference in morbidity. An additional statistical elaboration of the hospitalization data in the form of a logistic regression analysis is suggested, in particular at the district or regional level. However, after rigorous scrutiny, our data point to persistent remarkable differences at the respective sites in the major "welfare-related" population diseases, notably, cancer and cardiovascular diseases.