Because we limited the subjects to cases with pathological evidence of NSCLC and monitored them for 7 years, sex- and age-matched them

to reference subjects to estimate life expectancy, and adjusted for the utility values of QoL of an actual cohort and the corresponding referents in a real-world setting, our estimations were not confound by the preceding factors. Additionally, validation of our extrapolation method showed that the relative biases are small after 3 years of extrapolation. We thus tentatively conclude that such estimations would be useful for lifetime utility analysis of cancer under different selleck treatments, and detection of NSCLC patients at the operable stage would save more than 9 QALY. Moreover, operable IIIA patients were found to have a this website greater loss-of-QALE than inoperable IIIA patients (Fig. 3), which might imply a controversy in current practice. Since the sample size in the current study is relatively small, we recommend that future works matched on propensity scores be conducted to corroborate our results for potential

reconsideration of clinical practice guidelines. We selected patients with performance status 0–1 to estimate the differences in survival, QoL, and QALE. As patients with performance status 2–4 were usually confined to bed and physically unsuitable for curative operation, including them into the study might result in selection bias. Besides, most of them were unable to answer the questionnaire, thus the mean utility values would be overestimated. A sensitivity analysis including all subjects with performance status 0–4 (Table 2) was conducted and corroborated our conjectures. The mean utility values for patients with performance status 0–4 were

almost the same to those of patients with performance status 0–1, while the difference in loss-of-QALE was slightly underestimated because the mean age of the inoperable group became older and their loss of life expectancy became smaller. Unlike previous studies that applied Florfenicol internationally chosen life tables together with the experts’ determination of disability weights to calculate the disease burden of lung cancer using disability-adjusted life year (DALY) [22] and [23], we applied the national life tables of Taiwan and a cross-sectional sample of patients for measurement of QoL to estimate the QALE and loss-of-QALE by using QALY as the unit. While the DALY method makes international comparisons more feasible, the loss-of-QALE allows direct comparisons of different diagnosis and treatment strategies, and would likely be more useful for making decisions regarding the cost-effectiveness of national health policies. In our cohort, the 5-year survival rates for different stages of NSCLC (79.9%, 44.1%, 20.2%, and 7.7%, respectively, for stages I, II, IIIA, and IIIB-IV NSCLC) appeared comparable to those demonstrated by the National Cancer Institute [24].