Evaluating effects of method of administration on Walking Impairment Questionnaire

Background: Intermittent claudication resulting from peripheral arterial disease (PAD) can substantially impair walking function. The Walking Impairment Questionnaire (WIQ) assesses patient self-reported difficulty in walking. Currently this questionnaire is validated for interviewer administration only. Since this can be burdensome in a large clinical trial, we examined the effects of alternative methods of administration on patient responses on the WIQ. Methods: The WIQ, which consists of four subscales (pain severity, distance, speed, stairs), was modified to be self-administered or interviewer-administered by telephone. Patients with PAD were recruited from two sites and randomized into two groups: in group 1 the WIQ was self-administered, then telephone-administered; in group 2 the WIQ was telephone-administered, then self-administered. The two administrations occurred 4 to 7 days apart. Additional measures (SF-36, EQ-5D, and PAD symptom scale) and clinical data were included to further assess the validity of the WIQ and symptoms in patients with claudication. Telephone interviews were conducted by trained interviewers using standardized scripts. Two-week test-retest reliability was assessed for both the self-administered WIQ (group 1) and the telephone-administered WIQ (group 2). Results. Sixty patients were recruited at two sites (n = 30 per group). Seventy-eight percent were men; mean patient age was 67.1 years; and 83% of patients were white. Mean duration of PAD symptoms was 6.8 years. No significant differences were observed in WIQ subscale scores between self administration and telephone administration. No interaction effects between order and method of administration were detected. Cronbach alpha for distance, speed, and stair-climbing subscales ranged from 0.82 to 0.94. Correlations among WIQ subscales and the symptom scale were good (r = -0.34 to -0.57). Correlations of WIQ subscales with physical health subscales of the SF-36 (r = 0.24-0.59) were higher than for mental health-related subscales (r = 0.08-0.26). Conclusions. The modified WIQ demonstrated good reliability and validity with both methods of administration. These results suggest that the self-administered and telephone-administered versions of the WIQ can be used reliably and efficiently in clinical trials.