Influence of Hospital Volume on Mortality Following Major Cancer Surgery

To the Editor: In their study on the influence of hospital volume and surgical volume on 30-day mortality after major operative procedures in patients with cancer, Dr Begg and colleagues¹ attempt to address a complex and difficult task with an elegant approach. However, this study cannot address 1 important association of quality outcome.

While a high-volume center is more likely to achieve better outcome, excellent outcome is still feasible in low-volume centers under certain circumstances. For instance, Begg et al list zero mortality after pancreatectomy in 5 hospitals with only 5 patients each during the study interval of 10 years. At the City of Hope National Medical Center, a "low-volume" institution according to a definition offered by Lieberman et al,² 54 patients have undergone partial or total pancreatectomy between 1988 and 1998, including 50 for a diagnosis of cancer. Nine surgeons performed a median of 6 such procedures during this time period (range, 2-8). All surgical oncologists were fellowship- or specialty-trained, most in high-volume cancer centers. Thirty-day or in-hospital mortality was zero. The median length of stay (LOS) was 16.5 days (12 days since 1994). The median postoperative survival, depending on cancer site, ranged from 22.5 months (pancreatic) to 56 months (ampullary and bile duct).