Effect of a daily scrum at the gastrointestinal oncology ward on length of stay: A tertiary center experience.

Abstract

e13540 Background: The disease course of oncology patients often results in repeated hospitalization and prolonged length of stay (LOS). Next to disease and treatment complexity a causal role is attributed to psychological, social, nutritional and physical factors. Multidisciplinarity between (para)medical actors can limit LOS and reduce days with limited medical value. However this multitude of actors can also hold a reverse effect as it requires coordination between all parties. ‘Scrum’ is a methodology, originating in IT development, to approach complex issues on an operational level. It allows for quick adaptation to evolving situations requiring teamwork, quite similar to the care planned and provided for vulnerable patients with complicated pathology. The goal is to assess whether a daily scrum can impact LOS and the rate of readmissions. Methods: Starting October 2022, a daily scrum was implemented at the GI Oncology ward (capacity: 26 beds) of a tertiary center. Every day a window of 10 to 15 minutes was strictly reserved for a scrum, led by one of the two supervising medical doctors (MDs), together with the rotating residents in training and the head nurse. This daily debriefing of every single patient focused on the exact medical criteria for discharge, next to an upfront concrete plan for paramedical discharge criteria. A linear model was used to test whether the scrum succeeded at reducing LOS, while correcting for case mix (parameters: 3M All Patient Refined Diagnosis Related Groups (APR DRG), age, weight, adjustment for COVID period). All hospitalizations from October 2022 until December 2023 were included in the analysis. Results: 664 GI oncology patients were hospitalized between October 2022 and December 2023. The mean LOS was 6.07 days. After correcting for case mix, LOS was significantly reduced in the period in which a daily scrum was performed compared to four years of data prior to the scrum (-10.4%, p = 0.003). The expected LOS reduction based on the four years of data prior to the scrum was 4.0% ( p < 0.001). Case mix corrected hospital-wide unplanned readmissions within 30 days from discharge did not increase despite this lower LOS. After a change in supervising MDs in August 2023, LOS did not significantly alter (+1.0%, p = 0.837) suggesting durability of the observed effect. Conclusions: A daily scrum at the GI oncology ward significantly shortens LOS. Scrum is a useful tool to optimize the use of bed capacity by reducing days in which the patient is in hospital with no or limited value, and to prevent undesirable bedspacing (‘alternative patients’) at risk for suboptimal care and potential harm.