Colorectal cancer prevention efficacy of primary care physician-performed vs. specialist-performed colonoscopies at one endoscopy center

Background: Lifetime risk of colorectal cancer (CRC) in the US is 5-6%. Well-performed colonoscopy screening and surveillance to clear the colon of precancerous polyps in theory can greatly reduce CRC incidence. A major hurdle for screening and surveillance is the shortage of endoscopists. A potential solution, training primary care physicians (PCPs) in colonoscopy, was called into question by a Canadian study that reported poor CRC protection among PCP-served patients. Aim: To examine the CRC prevention impact of PCPs performing screening colonoscopies under a standardized, polyp yield-maximizing clinical protocol implemented across all providers at a single endoscopy center in South Carolina. Methods: Since 2001, an innovative colonoscopy protocol along with systematic data documentation was implemented. Fifty-four PCPs were trained in colonoscopy using training processes similar to those of gastroenterology fellowship programs (hands-on expert involvement transitioning into supervised, independent performance over the first 140 procedures, the credentialing benchmark for gastroenterology fellows). Following training, PCPs were provided hands-on assistance by an endoscopy technician, with an onsite expert always available for rescue assistance. A previous study documented high rates of adenoma detection and patient safety. Post-colonoscopy CRC incidence rates for PCP vs. specialist cases, merging the center's 2001-2008 colonoscopy cohort data with 1996-2010 CRC data from the South Carolina Central Cancer Registry (SCCCR), were compared. The SCCCR is a NAACR-accredited gold-rated registry that captures all CRC cases diagnosed in South Carolina or in any of 20 surrounding states via reciprocal data exchange agreements. Age-, sex- and race-adjusted standardized incidence ratios (SIR) relative to South Carolina and SEER-17 incidence rates were calculated. Results: Of 16,364 study-eligible persons (4 excluded due to missing SSN) with screening colonoscopy 12,383 (76%) were performed by PCPs and 3,981 by specialists, with total 75,318 person-years (mean 4.5, median 4.9) of observation (PYO). Nineteen persons developed CRC (14 and 5 among PCP and specialist cases, respectively). Overall SIR was 17%. Cancer prevention efficacy was 83% for PCPs and 80% for specialists. Analyses limiting the cohort to 2001-2004 (to ensure >5-year follow-up) resulted in 7,921 patients with 49,356 PYOs (mean 6.7, median 6.6) and showed an 87% cancer prevention efficacy for PCPs and 83% for specialists. Conclusions: Colonoscopy by PCPs conferred high CRC protection similar to the highest ever documented protection rate of 76-90% reported in the 1993 National Polyp Study. PCP-performed colonoscopies with technical support under a standardized, high-performance colonoscopy protocol may be a solution to the shortage of endoscopists.