Mo1271 Comparison of Internal Medicine Residents to GI Fellows in the Assessment of Patients With GIB at an Inner-City University Teaching Hospital: Are Patients Triaged Appropriately?

Abstract

Comparison of Internal Medicine Residents to GI Fellows in the Assessment of Patients With GIB at an Inner-City University Teaching Hospital: Are Patients Triaged Appropriately? Priti Bijpuria*, Alexander Schlachterman, Scott Naples, Asyia S. Ahmad Gastroenterology, Drexel University, Philadelphia, PA Background: It is essential for physicians to accurately identify acute gastrointestinal bleeding (GIB) as urgent endoscopy may be indicated in patient with melena or hematemesis and for those patients with significant comorbidities such as coronary artery disease (CAD). GI specialists are dependent on primary medicine teams to accurately communicate patient information in the setting of GIB. There has been no study to date identifying the accuracy of internal medicine residents’ initial evaluation of GIB. Objective: To evaluate whether the accuracy of internal medicine (IM) residents’ assessment of GIB in patients correlates with the impression of GI specialists during consultation at an innercity university teaching hospital. Methods: Prospective study of inpatient GIB consultations requested from 8/2011 to 11/2011. A questionnaire was distributed to house staff requesting GIB consultations as well as to the GI fellows performing the consult. The questionnaire assessed physical exam, treatment, medications and past medical history. In addition, both the residents and fellows were asked to assess GIB, specifically melena, based on 1) stool color using a stool color card and 2) by rectal findings. Fellow digital rectal exam (DRE) findings served as a control for stool color identification. Results: There were 142 residents and 6 GI fellows eligible for this study. Thirty-nine GI consults have been evaluated to date. Of these consults, residents completed the questionnaire in 37 patients and fellows completed all 39 questionnaires. Residents were less likely to do a DRE to support their diagnosis of GIB 22 of 36 (61%) compared to fellows 34 of 38 (89%); p 0.006. When asked to characterize the stool of a presumed GIB, residents more frequently labeled stool as melena in 13 of 33 (39%) patients as compared with fellows in 3 of 38 (7.9%) patients; p 0.002. Residents inaccurately identified melanotic stools in ten patients. Four were incorrect based on the stool color and the other 6 were inaccurately labeled melena based on DRE findings that contradicted fellow findings. Finally, residents’ identified CAD in 14 of 29 patients (48%) whereas fellows identified CAD in only 7 of 34 patients (21%); p 0.03. Conclusion: Preliminary data indicates residents are less likely to perform DRE during evaluation for GIB and are less likely to accurately identify melena based on stool color or DRE findings. Surprisingly, residents were more likely than GI fellows to ascertain a history of CAD in GIB patients even though CAD is a significant comorbidity that impacts both the approach and timing of endoscopy. There appears to be a need to educate both residents and GI fellows on the appropriate triage of patients with acute GIBs. In addition, improved communication between services is necessary especially in high-risk patients.