This project explores a key knowledge gap regarding the quality and colorectal cancer (CRC) protection effect of primary care physician (PCP)-performed colonoscopies under a innovative program in South Carolina that trains and utilizes primary care physicians (PCP) for colonoscopy screening. The program uses a unique clinical protocol, with inbuilt quality safeguards to compensate for PCPs'lack of formal gastroenterology training. A previous NCI-funded study documented the high quality and patient safety of PCP-performed colonoscopies, providing the evidence base for the clinical effectiveness of PCP-colonoscopies performed in a structured setting. The same study also found significantly higher colonoscopy rates among patients of trained AA PCPs relative to untrained PCPs, an increase that was led by drastic increases among AAs, particularly AA men. Though promising, questions remain to be addressed before this innovative program can be scaled up to expand colonoscopy access for AAs. The questions are: a) Do PCPs'adenoma/cancer yield rates and procedure performance indicators match those of specialists performing colonoscopies in the same clinical setting and screening patients drawn from the same population? This study aim will clarify the performance quality of PCP-colonoscopies accounting for potential confounding due to unobserved population-related variables. b) Do PCPs confer similar CRC protection rates as specialists serving the same population, and do PCP-cohort protection rates compare well with the CRC protection rates found in randomized clinical trials? This study will investigate these questions using data on 13,688 screening colonoscopies by 54 PCPs and on 5031 colonoscopies by 5 specialists. To detect differences in performance quality and case yield rates we will compare PCP-performed procedures with specialist-performed procedures on the following metrics: a) polyp detection rate and mean polyps detected per patient, b) adenoma detection rate and mean adenomas per patient, c) advanced adenoma detection rate and mean per patient, d) cancer detection rate, e) location of detected adenomas - right-sided, left-sided and cecal adenomas, f) colonoscope insertion and withdrawal times when no polyp was found and when polyp(s) were found, g) cecal intubation rate, h) reasons for incomplete colonoscopies (not achieving cecal intubation), i) repeat colonoscopy due to unsatisfactory procedure, and j) all of the above indicators classified by PCPs'training status and procedure volume. To satisfy study aim 2 we will assess the number and type of incident cancers among PCP- vs. specialist-served cohorts (65,685 person years (PYO) from 13,688 PCP colonoscopies, and 34,255 PYOs from the 5031 specialist-performed colonoscopies). We will source cancer data from the South Carolina Central Cancer Registry (SCCCR) and death data from the SC Vital Statistics registry. We will classify incident cancers by probable cause as per the documented algorithm (Pabby et al) into 4 categories: missed cancer, new cancer, incomplete removal, failed biopsy detection, and incomplete colonoscopy.
This project explores a key knowledge gap regarding the performance quality and colorectal cancer (CRC) protection effect of colonoscopies performed by primary care physicians (PCP) under a innovative program (with inbuilt quality assurance mechanisms to compensate for PCPs'lack of formal gastroenterology training) in South Carolina. The study will explore: a) whether the PCP-colonoscopy cohort rates of precancerous adenomas and cancer, and other quality measures are comparable to those of the specialist cohort performed in the same clinical setting and population, and b) whether PCP-performed colonoscopies confer similar CRC protection on patients as specialists, and as per the documented rates in randomized clinical trials. It will use colonoscopy and cancer registry data pertaining to 65,685 person years (PYO) of follow-up following 13,688 PCP colonoscopies, and 34,255 PYOs following 5031 specialist-performed colonoscopies.
