Background: With the transition to diagnosis related group (DRG)-based hospital reimbursement, length of stay has become a significant financial driver of every health system and hospital medicine group. To understand how one health system compares to national peers, length-of-stay index (LOSi) has become the standard method of reporting. LOSi is defined as the ratio of “observed LOS” divided by “expected LOS”. While LOSi is a valuable metric, it can be skewed by a small number of patients that stay an incredibly long period of time, defined as “outliers”. At our academic medical center, the hospital medicine LOSi for fiscal year 2023 was 1.13. After removing LOSi outliers, the LOSi was 1.04. To date, LOSi outliers have not been clearly described in the literature. The goal of this study is to better understand the variables that drive LOSi outliers for hospital medicine discharges and potential opportunities for both hospitalists and health system leaders to improve these metrics.
Methods: Cross-sectional analysis of FY23 LOSi outliers discharged by hospitalists at our institution. The FY23 LOSi outliers were identified through the Vizient® Clinical Data Base, which is an analytics platform that provides patient outcomes enabling health systems to benchmark against similar peers. Chart review was conducted of the 100 patients classified as LOSi outliers, with extraction of descriptive statistics including age, gender, and payer status. Hypothesized variables that may contribute to LOS were reviewed and included admit source, admit service, time spent in the ICU, days spent on other inpatient services prior to transfer to hospital medicine, discharge disposition, and reason for discharge delay.
Results: Of the 6202 patients discharged by hospitalists in FY23, 100 patients were classified as LOSi outliers. The median age was 63 years (mean 60; SD 16.6). Fifty-eight LOSi outliers had Medicare, 11 Medicaid, 19 commercial insurance, 7 self-pay, and 5 classified as other. The mean LOS was 49.2 days (SD 58.8, range 8-471) with a median of 33 days. Of the 100 LOSi outliers, 57 patients were admitted from our institution’s emergency department, 32 from outside hospitals, 10 from clinics, and 1 from a skilled nursing facility. Regarding admit service, 77 were initially admitted to hospital medicine, 15 to an ICU, 4 to surgery, and 4 to other subspecialty services. Of the 23 LOSi outliers admitted to a non-hospital medicine service, the mean hospital day in which they were transferred to hospital medicine was 25; median 18. In reviewing discharge barriers, 42 patients were classified as having a medical reason for discharge delay, 37 patients had challenges with placement, 16 patients had both a medical and placement delay, and 5 patients died prior to discharge. Upon discharge, 44 patients were discharged home, 43 discharged to a facility (SNF, SAR, AR, LTACH), 9 to hospice/death, and 4 to another hospital. Of the 34 patients that had a placement barrier, only 5 were eventually able to be discharged home.
Conclusions: The majority of LOSi outliers discharged by hospitalists at our academic medical center are admitted from the emergency department to hospital medicine services. Of the 32 LOSi outliers admitted from an outside hospital, only 4 were discharged back to the hospital and may represent an opportunity for improvement. Of the 34 patients that had a placement barrier, only 5 were eventually able to be discharged home suggesting the need for increased support from health system leaders.