[Evaluation of the interobserver variability in the systematic classification of operative morbidity in lung resection].

OBJECTIVE To quantify the interobserver variability in establishing a systematic classification for the operative morbidity of lung resection. METHODS Morbidity was classified retrospectively in a series of 499 prospective registries of patients who underwent major lung resection (458 lobectomies and 51 pneumonectomies). The systematic classification proposed by Seely et al. in 2010 was used. Each one of the authors independently classified the complications and the weighted kappa statistic was calculated. RESULTS AND COMMENTS The kappa index was 0.79. Although the value is high, it introduces a systematic bias in the classification of patient morbidity that indicates the need to very carefully evaluate the data entered into the multi-institutional registers in order to be able to obtain valid conclusions.