Background: Comanagement is a structured collaboration between hospitalists and a surgical service to improve outcomes of surgical patients perioperatively. We identified the potential to improve patient safety and throughput metrics on the Surgical Oncology and Otolaryngology (ENT) services at our institution. We partnered with these two services to create a novel Surgical Oncology Comanagement Service.
Methods: A dedicated hospitalist rounded on Surgical Oncology and ENT patients daily. Inclusion criteria were patients admitted to a specified group of surgical attendings, ASA score 3 or 4, and expected LOS >1 day. The hospitalist managed the full range of medical issues, including, but not limited to, pre-operative assessment and testing; management of comorbidities, such as DM, CKD, and ischemic heart disease; antibiotic management; and DVT prophylaxis. The average census was 12-15 patients daily. The hospitalist communicated daily with the surgical team, and interfaced with family and social workers for goals of care and transition planning. The surgical service placed orders. The primary outcomes were in-hospital mortality, length of stay, 30-day readmission rate, and a composite of AHRQ patient safety indicators.
Results: The results reflect 13 months with Surgical Oncology (August 2014-August 2015; 641 patients) and 6 months with ENT (March 2015-August 2015; 371 patients) for a total of 1,012 patients. The comparison group was comprised of 1,254 patients cared for by the same surgical attendings in the 1-year pre-intervention period. Death among patients with a serious treatable condition as defined by AHRQ data was decreased from 11.6% (5 of 43) to 0% (0 of 62) (p< .05). There was a trend for decreased overall In-hospital mortality (0.64% vs 0.40%, p=NS). There were also trends for decreased length of stay (O:E ratio 1.01 vs 0.89), and for a composite of all AHRQ adverse outcomes (42 of 1254 patients (3.3%) vs 25 of 1012 patients (2.5%)) (p=NS). There was also a trend for decreased readmissions due to conditions unrelated to surgery (5.7% vs 4.5%), but not for conditions related to the surgical procedure (5.1% vs 6.2%). These trends were also noted when Surgical Oncology and ENT groups were analyzed separately.
Conclusions: Our Surgical Oncology Comanagement Program decreased death among patients with serious conditions as defined by AHRQ, and showed trends for improvements in overall mortality, LOS, related readmission rate, and AHRQ safety outcomes. Notably, the readmission rate for related (ie. surgical) reasons increased, while readmissions for unrelated reasons decreased. This is consistent with the premise that management of medical comorbidities and complications has been improved by comanagement. Though pre- and post-intervention comparisons can introduce confounders, the consistency of the findings across both Surgical Oncology and ENT suggest that this comanagement model can improve the quality of care for surgical patients.