In 2010, a multicenter randomized controlled trial reported increased postoperative complications among pancreaticoduodenectomy (PD) patients undergoing preoperative biliary decompression (PBD). We sought to evaluate rates of PBD at the population level before and after publication.
A retrospective population-based observational cohort study was performed. Patients undergoing PD for malignancy, 2005-2013, were identified and linked to administrative healthcare databases covering medical services provided to a population of 13.5 million. Patients undergoing PBD within 6 weeks prior to surgery were identified using physician billing codes, and divided into those undergoing PD before and after publication (6 month washout period). PBD rates were compared using chi-square tests. Surgeon and/or gastroenterologist consultations prior to PD were identified using billing codes, and the time interval between initial consultation and PD was calculated.
Of 2053 PD patients identified, 974 underwent surgery prior to article publication and 951 after (128 in washout period). Before publication, the rate of PBD was 47.4%, versus 40.8% after (p=0.003). The annual rate of PBD decreased from 52.5% (2005) to 40.4% (2013) (p=0.025). Among patients seen preoperatively by both a gastroenterologist and a surgeon, the median interval from initial consultation to surgery was 47 days (surgeon first), and 48 days (gastroenterologist first). Patients seen preoperatively only by a surgeon had a median interval of 30 days.
Rates of PBD have significantly decreased following publication of a randomized trial. While preoperative consultation with multiple specialists increases the time interval to surgery, the order in which they are seen does not.

An assessment of the effect of a randomized controlled trial: Preoperative biliary decompression in patients with periampullary malignancy.