Medical audit and patient satisfaction surveys.

INTRODUCTION Surveys among patients based on self-reporting can be severely biased if response rates are low and if the validity of the data collection instruments is low or unknown. This article presents a discussion and analysis of problems of validity and representative- ness in a patient satisfaction survey in the county of Fyn in 1991. The study was performed for the county health administration for the purpose of planning the services provided by general practice. All 291 general practitioners in the county of Fyn were invited and 136 of them participated. During 3 days in May ail patients consulting the participating practitioners were handed a self-administered ques- tionnaire on patient satisfaction, patient evaluation, and other aspects of the consultation. The doctors filled in a medical audit card with information on the consultation, part of which was directly comparable with the patients' evaluation (1). Of the target population of 6068,