Centralization of Esophagectomy: How Far Should We Go?

• Published in: Annals of surgical oncology Vol. 21; no. 13; pp. 4068 - 4074
• Main Authors: Henneman, Daniel, Dikken, Johan L., Putter, Hein, Lemmens, Valery E. P. P., Van der Geest, Lydia G. M., van Hillegersberg, Richard, Verheij, Marcel, van de Velde, Cornelis J. H., Wouters, Michel W. J. M.
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.12.2014; Springer Nature B.V
• Subjects: Adenocarcinoma - surgery; Aged; Databases, Factual; Esophageal Neoplasms - surgery; Esophagectomy - mortality; Female; Healthcare Policy and Outcomes; Hospitals, High-Volume - statistics & numerical data; Humans; Learning Curve; Male; Medicine; Medicine & Public Health; Middle Aged; Mortality; Netherlands - epidemiology; Oncology; Retrospective Studies; Risk Factors; Surgery; Surgical Oncology; Treatment Outcome; Netherlands; Volume Category; Esophageal Cancer; Cardia Cancer; Hospital Volume; Frailty Model
• ISSN: 1068-9265; 1534-4681
• DOI: 10.1245/s10434-014-3873-5

Background This study was designed to define a statistically sound and clinically meaningful cutoff point for annual hospital volume for esophagectomy. Higher hospital volumes are associated with improved outcomes after esophagectomy. However, reported optimal volumes in literature vary, and minimal volume standards in different countries show considerable variation. So far, there has been no research on the noncategorical, nonlinear, volume-outcome relationship in esophagectomy. Methods Data were derived from the Netherlands Cancer Registry. Restricted cubic splines were used to investigate the nonlinear effects of annual hospital volume on 6 month and 2 year mortality rates. Outcomes were adjusted for year of diagnosis, case-mix, and (neo)adjuvant treatment. Results Between 1989 and 2009, 10,025 patients underwent esophagectomy for cancer in the Netherlands. Annual hospital volumes varied between 1 and 83 year, increasing over time. Increasing annual hospital volume showed a continuous, nonlinear decrease in hazard ratio (HR) for mortality along the curve. Increasing hospital volume from 20 year (baseline, HR = 1.00) to 40 and 60 year was associated with decreasing 6 month mortality, with a HR of 0.73 (95 % confidence interval (0.65–0.83) and 0.67 (0.58–0.77) respectively. Beyond 60 year, no further decrease was detected. Higher hospital volume also was associated with decreasing 2 year mortality until 50 esophagectomies year with a HR of 0.86 (0.79–0.93). Conclusions Centralization of esophagectomy to a minimum of 20 resections/year has been effectively introduced in the Netherlands. Increasing annual hospital volume was associated with a nonlinear decrease in mortality up to 40–60 esophagectomies/year, after which a plateau was reached. This finding may guide quality improvement efforts worldwide.