Background: Traditional Length of Stay (tLOS) is commonly used in Hospital Medicine to assess hospital resource utilization and to incentivize reductions in patient stays. However, tLOS is often attributed to the discharging physician, which can lead to attribution errors and fairness concerns. These limitations can reduce the metric’s utility. In response, Pierce, Harrison, and Patel (1) proposed the concept of Individualized Length of Stay (iLOS). At our institution, we have implemented iLOS, and our preliminary findings suggest that it is reliable, timely, and warrants further investigation.
Purpose: We developed a data dashboard using Epic data, which assigns patients to the physician listed as the Primary Attending at 8 AM. iLOS is calculated by dividing the number of patient-days (i.e., the number of times a physician is listed as the Primary Attending for a given discharged patient) by the total number of discharges on a given day by that physician . The data was validated over a 2-week period by cross-referencing rounding schedules and discharges, and no discrepancies were found, indicating high fidelity of the system.
Description: iLOS offers several advantages over traditional tLOS. Most importantly, it eliminates the attribution errors inherent in tLOS. For example, a physician who begins caring for a patient on hospital day 28 and discharges them would have a tLOS of 28 but an iLOS of 1. This metric encourages physicians to discharge patients promptly and discourages transferring patients to other physicians to reduce overall LOS, resource utilization, and prevent clinical errors. Unlike tLOS, which is a retrospective metric tied to discharge, iLOS is calculated daily, allowing for real-time feedback and timely interventions. This prospective nature aligns more closely with institutional goals, especially in times of high census or low throughput, where resource allocation is critical. Furthermore, it enables targeted decision-making and incentivization, which can lead to better alignment between physician performance and institutional objectives.
Conclusions: We advocate for the adoption of iLOS as a novel metric in hospital medicine and encourage further study into its clinical and financial impacts. Our institution is actively assessing how iLOS influences physician performance, with the aim of identifying outliers and optimizing throughput.