Who Performs Proctectomy for Rectal Cancer in the United States?

BACKGROUND: There is wide variation in surgical care for rectal cancer in the United States. OBJECTIVE: This study aimed to assess the differences in individual surgeon procedural profiles that might explain variations in the rates of restorative vs nonrestorative proctectomy for rectal cancer. DESIGN: This study was a retrospective examination of a cohort derived from trackable state hospital discharge data from 11 states. PATIENTS: We identified all patients with rectal cancer that underwent restorative proctectomy (sphincter-sparing surgery) vs nonrestorative proctectomy (colostomy formation) over a 24-month study period (January 1, 2003 through December 31, 2004). INTERVENTION: We developed an inpatient procedural profile of each treating surgeon's practice across general surgery procedure codes and summed the number of restorative vs nonrestorative proctectomies for rectal cancer by surgeon. MAIN OUTCOME MEASURES: The primary outcome measures were nonrestorative proctectomy, mortality, and length of stay. RESULTS: A total of 7519 proctectomies were performed for rectal cancer by 2588 surgeons. During the 24-month study period, 1003 (38.8%) surgeons performed only nonrestorative procedures for rectal cancer. On multivariate analysis, the likelihood that a surgeon performed only nonrestorative procedures was increased if that surgeon performed more integumentary procedures and decreased if the surgeon performed at least one ileoanal pouch procedure or more anorectal procedures. Patients who underwent proctectomy by surgeons who performed only nonrestorative procedures had significantly higher mortality (2.5 +/- 0.7%) and longer length of stay (11.3 +/- 8.8 days) in comparison with those patients treated by surgeons who performed both restorative and nonrestorative procedures (1.3 +/- 0.3% mortality and 9.2 +/- 6.9 days, P < .001 for both analyses). The volume of proctectomy performed significantly affected all analyses. LIMITATIONS: The retrospective design introduces potential selection bias. CONCLUSIONS: Over a 24-month period, 38.8% of surgeons performed only nonrestorative procedures for rectal cancer. These surgeons did not regularly perform anorectal or ileoanal pouch procedures, suggesting that they may not have a focus on colorectal disease in their practice; they had significantly higher mortality and length of stay for their patients who underwent proctectomy for rectal cancer.