Extent of lymphadenectomy and survival in stomach cancer
17008 Background: Based on population-based studies, some investigators have posited that patients undergoing more extensive lymphadenectomy as part of their resection for stomach cancer had improved late survival rates. Such findings have prompted calls for the use of total lymph node counts as a quality indicator for hospitals. However, apparent relationships between number of lymph nodes resected and survival may be confounded by patient selection bias and provider factors. It is not clear that hospitals with higher lymph node counts have better outcomes than other hospitals. The purpose of this study is to examine relationships between total lymph node counts and survival for stomach cancer. Methods: Using the national Surveillance Epidemiology and End Results (SEER)-Medicare linked database (1992–2002), we first identified all patients undergoing major resections for gastric cancer (n=3,049). Hospitals at which the procedures were performed were categorized into 3 groups (terciles) according to the median number of nodes examined. We then assessed 5-year survival rates for each hospital group, adjusting for potentially confounding patient and hospital characteristics. Results: Hospitals with the highest median lymph node counts had slightly higher 5-year survival rates than those with the lowest node counts (31% vs. 28%; unadjusted HR for mortality 0.90, 95% CI 0.81–1.00). Hospitals with higher lymph node counts tended to treat lower risk patients and had lower procedure volumes. Adjusting for these confounding patient and provider characteristics further weakened the apparent relationship between survival and lymph node counts by hospital (adjusted HR, high vs. low hospital tercile, 0.96; 95% CI 0.85–1.09). Conclusions: Apparent relationships between total lymph node counts by hospital and 5-year survival rates after resection for stomach cancer are relatively weak and explained in large part by confounding patient and provider factors. Stronger evidence should be required before moving forward with this metric as a hospital quality indicator. No significant financial relationships to disclose.