Does the mode of data collection change results in a subjective knee score? Self-administration versus interview

Our objective was to compare the effect of two methods of data collection on results in a functional knee score. Two Lysholm scores were obtained for 61 patients 1 year after anterior cruciate ligament surgery at the same clinic visit. First, the patients completed a self-administered questionnaire, and second, the Lysholm score form was completed by the investigator in the course of a patient interview. A comparison of the scores revealed that the mean score was significantly lower with self-administration (self, 89.3 +/- 10.6; interviewer, 92.2 +/- 7.4) (P = 0.0035, Wilcoxon rank sum test). The assignment to one of four categories (excellent, good, fair, poor) was also significantly altered by the manner of data collection. Nineteen patients (31%) were assigned to different categories based on the mode of data collection. We believe that the major reason for a better score result with an interview was the presence of interview bias. The more the investigator is involved in the treatment of the patient, the greater the influence of this bias may be. To avoid such potential bias we suggest that a standardized self-administered questionnaire be used as the method of choice for obtaining subjective data in clinical settings.