Validity of the EuroQol-5 dimensions as a measure of recovery after pulmonary resection.

ABSTRACT

BACKGROUND: Surgical innovations advocated to improve patient recovery are often costly. Economic evaluation requires preference-based measures that reflect the construct of patient recovery. We investigated the responsiveness and construct validity of the EuroQol-5 dimensions (EQ-5D) as a measure of postoperative recovery after planned pulmonary resection for suspected malignant tumors. METHODS: Patients undergoing pulmonary resection completed the EQ-5D questionnaire and visual analog scales (VAS) for pain and fatigue at baseline (preoperatively) and at 1 and 3 mo postoperatively. Responsiveness and construct validity (discriminant and convergent) were investigated by testing a priori hypotheses. RESULTS: Fifty-five patients were analyzed (45% male, 62 ± 12 y, 29% video-assisted). There was no significant difference between median EQ-5D scores obtained at baseline (0.83 [interquartile range {IQR 0.80-1}]) compared to scores at 1 mo (0.83 [0.80-1], P = 0.86) and 3 mo after surgery (1 [0.83-1]; P = 0.09). At 1 mo after surgery, EQ-5D scores were significantly lower in patients undergoing thoracotomy versus video-assisted surgery (0.82 [IQR 0.77-0.89] versus 1 [0.83-1], P = 0.003), but there were no significant differences between patients ≥ 70-y old versus younger (0.95 [IQR 0.82-1] versus 0.83 [0.77-1], P = 0.09) or between patients with versus without complications (0.82 [IQR 0.79-0.95] versus 0.83 [0.80-1], P = 0.10). There was a low but significant correlation between EQ-5D and VAS scores of pain and fatigue (Rho -0.30 to -0.47, P ≤ 0.01). CONCLUSIONS: Despite evidence of convergent validity, the EQ-5D was not sensitive to the hypothesized trajectory of postoperative recovery and showed limited discriminant validity. This study suggests that the EQ-5D may not be appropriate to value recovery after lung resection.