Clinical validity of a disease-specific health status questionnaire: the peripheral artery questionnaire

Measuring patient-centered outcomes is becoming increasingly important in patients with peripheral arterial disease (PAD), both as a means of determining the benefits of treatment and as an aid for disease management. In order to monitor health status in a reliable and sensitive way, the disease-specific measure Peripheral Artery Questionnaire (PAQ) was developed. However, to date, its correlation with traditional clinical indices is unknown. The primary aim of this study was to better establish the clinical validity of the PAQ by examining its association with functional indices related to PAD. Furthermore, we hypothesized that the clinical validity of this disease-specific measure is better as compared with the EuroQol-5-dimensional (EQ-5D), a standardized generic instrument.

Data on 711 consecutive PAD patients undergoing surgery were collected from 11 Dutch hospitals in 2004. At 3-year follow-up, questionnaires including the PAQ, EQ-5D, and EuroQol-Visual Analogue Scale (EQ VAS) were completed in 84% of survivors. The PAQ was analyzed according to three domains, as established by a factor analyses in the Dutch population, and the summary score. Baseline clinical indices included the presence and severity of claudication intermittent (CI) and the Lee Cardiac Risk Index.

All three PAQ domains (Physical Function, Perceived Disability, and Treatment Satisfaction) were significantly associated with CI symptoms (P values < .001-.008). Patients with claudication had significant lower PAQ summary scores as compared with asymptomatic patients (58.6 +/- 27.8 vs 68.6 +/- 27.8, P = < .001). Furthermore, the PAQ summary score and the subscale scores for Physical Functioning and Perceived Disability demonstrated a clear dose-response relation for walking distance and the Lee Risk Index (P values < .001-.031). With respect to the generic EQ-5D, the summary EQ-5D index was associated with CI (0.81 +/- 0.20 vs 0.76 +/- 0.24, P = .031) but not with walking distance (P = .128) nor the Lee Risk Index (P = .154). The EQ VAS discriminated between the clinical indices (P values = .003-.008), although a clear dose-response relation was lacking.

The clinical validity of the PAQ proved to be good as the PAQ subscales discriminated well between patients with or without symptomatic PAD and its severity as defined by walking distance. Furthermore, the PAQ subscales were directly proportional to the presence and number of risk factors relevant for PAD. For studying outcomes in PAD patients, the disease-specific PAQ is likely to be a more sensitive measure of treatment benefit as compared with the generic EQ VAS, although the latter may still be of value when comparing health status across different diseases. Regarding disease management, we advocate the use of the disease-specific PAQ as its greater sensitivity and validity will assist its translation into clinical practice.