Working up rectal bleeding in adult primary care practices

• Published in: Journal of evaluation in clinical practice Vol. 23; no. 2; pp. 279 - 287
• Main Authors: Weingart, Saul N., Stoffel, Elena M., Chung, Daniel C., Sequist, Thomas D., Lederman, Ruth I., Pelletier, Stephen R., Shields, Helen M.
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 01.04.2017
• Subjects: Adult; Aged; Aged, 80 and over; Colorectal cancer; Community Health Centers - statistics & numerical data; Diagnostic Techniques and Procedures; Diagnostic tests; Female; Gastrointestinal Hemorrhage - etiology; Guideline Adherence; Hospitals, Teaching - statistics & numerical data; Humans; Male; Medical diagnosis; medical error; Medical History Taking; Middle Aged; missed diagnosis; Patient assessment; Physical Examination; practice guidelines; Practice Guidelines as Topic; practice variation; Primary care; Primary Health Care - organization & administration; Primary Health Care - statistics & numerical data; rectal bleeding; Socioeconomic Factors; Urban Health Services - statistics & numerical data; United States > US; Massachusetts; practice variation; rectal bleeding; practice guidelines; medical error; missed diagnosis
• ISSN: 1356-1294; 1365-2753; 1365-2753
• DOI: 10.1111/jep.12596

Rationale, aims and objectives Variation in the workup of rectal bleeding may result in guideline‐discordant care and delayed diagnosis of colorectal cancer. Accordingly, we undertook this study to characterize primary care clinicians' initial rectal bleeding evaluation. Methods We studied 438 patients at 10 adult primary care practices affiliated with three Boston, Massachusetts, academic medical centres and a multispecialty group practice, performing medical record reviews of subjects with visit codes for rectal bleeding, haemorrhoids or bloody stool. Nurse reviewers ed patients' sociodemographic characteristics, rectal bleeding‐related symptoms and components of the rectal bleeding workup. Bivariate and multivariable logistic regression models examined factors associated with guideline‐discordant workups. Results Clinicians documented a family history of colorectal cancer or polyps at the index visit in 27% of cases and failed to document an abdominal or rectal examination in 21% and 29%. Failure to order imaging or a diagnostic procedure occurred in 32% of cases and was the only component of the workup associated with guideline‐discordant care, which occurred in 27% of cases. Compared with patients at hospital‐based teaching sites, patients at urban clinics or community health centres had 2.9 (95% confidence interval 1.3–6.3) times the odds of having had an incomplete workup. Network affiliation was also associated with guideline concordance. Conclusion Workup of rectal bleeding was inconsistent, incomplete and discordant with guidelines in one‐quarter of cases. Research and improvements strategies are needed to understand and manage practice and provider variation.