The difference in loss-of-QALE between operable and inoperable NSCLC patients, or, the net difference after adjustment of potential lead-time bias for age selleck chemical between the two groups, was 9.00 ± 0.18 QALY. We conducted a sensitivity analysis for patients with performance status 0–4 (Table 2). As patients with performance status more than 2 were usually confined to bed and unavailable to answer the questionnaire, the resulted mean utility values of QoL were similar to those

of patients without including them. The difference in QALE between operable and inoperable NSCLC patients would be 10.26 QALY, which was not different from that using patients with performance status 0–1 alone. However, the difference in loss-of-QALE would be learn more underestimated slightly (= 8.36 QALY), probably because of the older mean age of inoperable patients. Another sensitivity analysis was conducted by only including

the utility values of the first QoL measurements of 518 patients in the calculations, the difference in QALE would be 10.43 QALY and the difference in loss-of-QALE would be 9.20 QALY for patients with operable and inoperable NSCLC, and these results are not significantly different from those using repeated measurements. We also performed a stratified analysis among patients with stage IIIA NSCLC (Fig. 3). Compared with inoperable stage IIIA patients, operable stage IIIA patients had a longer QALE. Moreover, the loss-of-QALE for operable stage IIIA patients was greater than that of inoperable stage IIIA patients, probably because of the younger mean age at diagnosis. There were 262 patients with operable and 621 patients with inoperable NSCLC diagnosed during the first 4 years, between 2005 and 2008, of which the survival curves were extrapolated to 2011 and compared with the Kaplan–Meier estimates based on the 7-year follow-up. The relative biases

of the extrapolation ranged between −4.6% (p = 0.099) and −6.0% (p = 0.116) after 3 years of extrapolation ( Table 3). Although the QoL for NSCLC patients has why been measured previously in several studies [20] and [21], integrating the survival with utility values of QoL to estimate the lifetime utility difference between patients with operable and inoperable NSCLC has never been comprehensively evaluated. In our study, the utility values of QoL for patients with operable and inoperable NSCLC were stratified into different age bands (Table 2), which show that compared with inoperable patients, operable patients had mean utility values closer to those for the general population (0.96, 0.93, 0.86 for men ≤54 years, 55–74 years, ≥75 years and 0.96, 0.91, 0.78 for women ≤54 years, 55–74 years, ≥75 years, respectively). In addition, we quantified the difference in loss-of-QALE (9.00 ± 0.