Editorial comment

Abstract

The volume–outcome relationship has been used as a proxy measure for quality for more than 3 decades since first proposed by Luft and colleagues1Luft H.S. Bunker J.P. Enthoven A.C. Should operations be regionalized? The empirical relation between surgical volume and mortality.N Engl J Med. 1979; 301: 1364-1369Crossref PubMed Scopus (1411) Google Scholar in 1979. In this issue of the Journal of Thoracic and Cardiovascular Surgery, Falcoz and colleagues2Falcoz P.-E. Puyraveau M. Rivera C. Bernard A. Massard G. Mauny F. et al.The impact of hospital and surgeon volume on the 30-day mortality of lung cancer surgery: A nation-based reappraisal.J Thorac Cardiovasc Surg. 2014; 147: 1805-1812Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar report the results of lung cancer resection in France captured in the Epithor database. From 2005 to 2010, the database captured the data for almost 20,000 patients undergoing anatomic lung resection for lung cancer. The most striking result of their report was the decrease in 30-day mortality from 10% in 2005 to 3.8% in 2010. The authors used sophisticated hierarchical logistic regression models to evaluate the relationship between procedure volume (both surgeon and hospital) and 30-day mortality. Rather than categorizing the procedure volume into arbitrary groups such as quartiles or quintiles, the authors appropriately modeled volume as a continuous variable using a polynomial function. The results have demonstrated that surgeon volume, but not hospital volume, was associated with 30-day mortality, with a P value < .05. Although this is statistically significant, it is unclear how clinically important surgeon volume would be as a predictor of mortality. Two simple statistical tests could have helped the reader make this determination and have been performed by our group when studying the volume–outcome relationship.3Kozower B.D. Stukenborg G.J. The relationship between hospital lung cancer resection volume and patient mortality risk.Ann Surg. 2011; 254: 1032-1037Crossref PubMed Scopus (49) Google Scholar, 4LaPar D.J. Kron I.L. Jones D.R. Stukenborg G.J. Kozower B.D. Hospital procedure volume should not be used as a measure of surgical quality.Ann Surg. 2012; 256: 606-615Crossref PubMed Scopus (72) Google Scholar First, the models could have been repeated without the volume variable to determine whether a change occurred in model performance. If no change was found in the c statistic of 0.81, volume would have contributed very little to the predictive capacity of the model. Second, the statistical significance of the fixed effects for each variable in the model could be assessed using the F test statistic. The F test provides a number that enables the reader to rank the variables in order of their contribution to the predictive capacity of the model. A P value that reaches statistical significance of <.05 could have a very small F test statistic and contribute little to a predictive model. In conclusion, the National Cancer Plan in France deserves tremendous credit for a dramatic decrease in perioperative mortality during the 5 years studied. The results of the study have demonstrated that surgeon volume is statistically associated with 30-day mortality; however, the clinical significance of this finding remains in question. Importantly, the authors advocate for the careful examination of clinical outcomes data for the measurement of quality rather than using procedure volume as a proxy measure. Participation in this systematic clinical database (Epithor) was a key component of the National Cancer Plan and allows participants to benchmark themselves and critically evaluate their performance. The impact of hospital and surgeon volume on the 30-day mortality of lung cancer surgery: A nation-based reappraisalThe Journal of Thoracic and Cardiovascular SurgeryVol. 148Issue 3PreviewOur objective was to analyze the time trend variation of 30-day mortality after lung cancer surgery, and to quantify the impact of surgeon and hospital volumes over a 5-year period in France. Full-Text PDF Open Archive