Triage of colonoscopies: open access endoscopy versus outpatient consultation with a gastroenterologist.

Abstract

Background and study aims: In many Dutch hospitals, open access referral for colonoscopy is authorized by a gastroenterologist after screening a standard referral letter (SRL) without face-to-face contact with the patient. We investigated the added value of a 7.5 min outpatient consultation with a gastroenterologist (OC), regarding the patient indications, priority for colonoscopy, and the frequency of correct information about patient medications and comorbidities on SRLs. Patients and methods: In a prospective, blinded, single-center study, gastroenterologists assessed SRLs for the accuracy and priority of the colonoscopy request (SRL). These data were compared to results from the OC, and primary outcomes were the number of patients who were not recommended for colonoscopy and priority scheduling of colonoscopy for suspicion of cancer. Results: Patients were analyzed using both SRL and OC and, of 255 patients, 224 of them underwent colonoscopy. Colonoscopy was not recommended for 6.3 % and 11.4 % of patients using the SRL and OC, respectively (P = 0.02). Using the SRL, gastroenterologists did not recommend colonoscopy for seven patients, but the same patients were recommended for colonoscopy when OC was available. This was explained because the indications on the SRL did not match the information obtained from OC. Compared to OC , more colonoscopies were prioritized when the SRL was used to make decisions. Cancer was detected in 7/112 (SRL ) versus 7/65 (OC ) of priority-scheduled patients. SRLs did not report the use of coumarins and insulin in 1.6 % of patients or the prevalence of serious comorbid conditions in 52 % of patients. Conclusions: A 7.5 min outpatient consultation with a gastroenterologist improved the identification of indications for colonoscopy, decreased priority scheduling of patients, and increased the number of patients diagnosed with cancer in the prioritized group. SRLs frequently omitted patients’ medications and comorbidities.