Post Colonoscopy Patient Education Survey; Do Patients Know Their Results and What Happens Next?

that capture all persons being screened or surveyed.Thus, we examined the variability in ADR for endoscopists practicing in a large, community-based setting while adjusting for the influences of patient demographic mix for each provider.Methods: We evaluated Kaiser Permanente Northern California electronic procedure and pathology records to identify adenomas detected among average-risk patients ≥50 years who underwent screening or surveillance colonoscopies in 2006 to 2008.We excluded endoscopists with <300 total exams during the study period.This resulted in 19,610 screening exams and 28,428 surveillance exams performed by 104 gastroenterologists and 4 general internists.Each physician's ADR was defined as the proportion of exams in which ≥1 adenoma or colorectal cancer was detected.Unless otherwise noted, ADRs were adjusted for patient sex, age, and race/ ethnicity to correct for differences in patient mix (e.g., if a provider had predominantly female or older patients).Variables associated with ADR (i.e., patient age, patient and physician sex and race/ethnicity, years as a physician, medical school location, specialty, procedure volume, and medical facility) were also evaluated using multilevel logistic regression models.Results: Physician ADRs for screening colonoscopies averaged 0.222 and 0.151 for male and female patients, respectively.In both sexes combined, ADRs averaged 0.180 for screening exams (5th to 95th percentile range: 0.060 -0.354) and 0.348 for surveillance exams (5th to 95th percentile range: 0.181 -0.500).Physicians with higher ADRs (prior to adjustment for patient demographics) were more likely to have performed exams on male and older patients, attended a U.S. medical school, and completed their medical training more recently.Conclusions: ADRs in large, community-based settings vary widely between colonoscopists (similar to prior reports from referral centers).ADRs for surveillance exams are substantially higher than rates for screening exams.Adjustment for patient mix changed ADR markedly (>50%) for some physicians.These results can help set benchmarks for colonoscopy quality improvement programs, and emphasize the importance of patient mix and procedure indication when setting quality improvement benchmarks.