[The value of epidemiologic studies of ski injuries].

The Garmisch-Partenkirchen experience--15,513 skiing injuries between 1972 and 1990--was compared to the results of an inquiry into the skiing injuries among 565 persons of the medical profession, that were active skiers over an average of 18 years. The parameter injuries per 1000 skiing days was calculated from the outpatient clinic data as well as from all skiing injuries that had happened in the interview group and those that required surgical treatment. The mean value of 2.7 injuries/1000 skiing days--corrected for the number of patients--from the literature and the clinical data with 2.1 is comparable to the 1.8 injuries/1000 skiing days (injuries that required surgical treatment) but stands sharply against the 4.2 arising from the total number of injuries in the inquiry group.--Thus, every epidemiological study clearly underestimates the actual rate of skiing injuries. This fact may be defined as the "bypass-effect". The comparison between clinical setting and inquiry allowed for a quantification: Depending on the type of injury and the study design, a true prevalence of skiing injuries that is between 10% and 200% higher than the published data has to be expected.--Descriptive, not controlled epidemiologic studies on skiing injuries are justified and give valid trends on which therapeutic and prophylactic measures can be based--but exclusively under the following prerequisite conditions: 1. a standardized study protocol; 2. an observation.