Classifying and coding morbidity in general practice: validity and reliability in an international trial.

An international field trial of the draft defined version of the International Classification of Health Problems in Primary Care (ICHPPC) was undertaken in nine countries. Fifty-two physicians coded morbidity in 76 standard written clinical vignettes. The validity and reliability of morbidity coding using the defined classification and the extent to which these might be improved by using definitions have been estimated. There was an average of 1.7 problems per encounter, for a total of 807 different problems recorded. Sixty-two percent of the problems were coded correctly. The majority of incorrectly recorded problems were due to varied minority views of the contents of the encounters and to mistakes in recording code numbers. The results of morbidity surveys are useful only if the methods of gathering, recording, coding, collating, and analyzing the data are explicit and well documented. This study illustrates the difficulties of recording morbidity in general practice encounters. Better use of definitions and the addition of an index could improve coding accuracy.