Quality measures for patients with early and locally advanced esophageal cancer undergoing esophagectomy have been made by national organizations. The rate of adherence to these measures as well as their association with overall survival are unknown. Esophagectomy patients were abstracted from the National Cancer Database. Because neoadjuvant status was available since 2006, the analysis of locally advanced patients began at this time point. Selected measures included: R0 resection, evaluation of 15 or more lymph nodes, and induction therapy for locally advanced tumors. Multivariate models identified variables associated with achieving quality measures. A Cox proportional hazards model evaluated factors associated with mortality. From 1998 to 2012, 4,908 of 16,040 (30.6%) early-stage esophageal cancer patients (clinical T1A to T2N0 <2cm, well-differentiated) underwent esophagectomy. Of 4,672 patients 4,518 (96.7%) achieved R0resection and 1,395 of 4,686 (29.8%) had 15 or more lymph nodes sampled. High-volume center type (>20 esophagectomies/year) was independently associated with meeting both measures (odds ratio [OR] 2.2, 95% confidence interval [CI]: 1.9 to 2.5). From 2006 to 2012, 7,747 of 20,437 (37.9%) locally advanced patients (clinical Stage IIB to IIIB) received esophagectomy. Of 6,966 patients 5,977 (85.8%) received induction therapy, 6,394 (91.8%) had R0 resection, and 2,852 (40.9%) had 15 or more lymph nodes sampled. High-volume center type was, again, associated with increased likelihood of meeting all quality measures (OR 2.17, 95% CI: 1.92 to 2.46). Meeting all quality measures was associated with the largest decrease in mortality for both early-stage (hazard ratio [HR] 0.27, 95% CI: 0.18 to 0.39) and locally advanced (HR 0.54, 95% CI: 0.40 to 0.73) esophageal cancer patients. Adherence to recommended quality measures is independently associated with improved overall survival in both early and locally advanced stages of esophageal cancer. Currently, few patients are receiving care in accordance with these recommendations.
Read full abstract