EVALUATING STRATEGIES TO REDUCE PATIENT MEAN GEOMETRIC LENGTH OF STAY: A MULTI-FACILITY STUDY TO IMPROVE PATIENT OUTCOMES

Background: The increased demand for healthcare has greatly exacerbated the need for open hospital beds. While many patients may be medically able to leave the hospital, they are often still in need of assistance to manage their pain, comorbidities, complex medication regimes, and acts of daily living. These patients are often discharged to Skilled Nursing Facilities (SNFs), where they can receive the continued care they require outside of a hospital setting. Any mistakes in discharge coordination can lead patients to discontinuity of care, increasing likelihood of adverse events and readmissions, making the transitions of care between hospitals and SNFs a key area for quality improvement. This study aims to utilize a 6-step program in order to create a system of thorough integration between Hospital Medicine (HM) and Post-Acute Care (PAC) teams to improve overall patient outcomes.

Methods: The study was conducted utilizing 5 sites with HM and PAC programs: North Carolina, Florida, Nevada, Washington, and Tennessee. A 6-step program was initiated at each of these sites with the goal of improving outcomes for patients being discharged to a SNF. The primary outcome of this study was Geometric Mean Length of Stay (GMLOS), while secondary outcomes included Average Length of Stay (ALOS), Readmission Rates, and Early Discharge Order Rates. These metrics were tracked 3 months prior to implementation of the program, as well as 3 months post-implementation. The 6 steps utilized in this program are: 1) Initiate communication, 2) Actively engage, 3) Crease Team Culture, 4) Establish Collaborative Practices, 5) Define and Track Measures, and 6) Leadership Accountability.

Results: The pre-implementation GMLOS O/E to SNF was 2.70 ± 2.42, while the post-implementation GMLOS O/E to SNF was 2.59 ± 2.20. With regards to the secondary outcomes, pre-implementation ALOS to SNF was 5.14 ± 0.57, while post-implementation ALOS to SNF was 5.17 ± 0.53. The pre-implementation readmission rate also went up slightly from 11.61% ± 0.20% to a post-implementation rate of 12.48% ± 0.19%. Early discharge orders, where rates indicate the percentage of daily discharges before 11:00 am, showed a decreased rate. The pre-implementation early discharge rate was 38.97% ± 4.68%, while the post-implementation early discharge rate was 33.83% ± 5.35%.

Conclusions: Although the 4% reduction in GMLOS after implementation of the ATP program was not found to be statistically significant, it is of clinical significance given the high association between longer LOS and worse patient outcomes. The lack of significant changes to secondary measures indicates that no metric was compromised by implementation of this program. This study demonstrated an avenue by which hospital systems can streamline patient care without sacrificing quality. Future research can build on these findings by exploring how optimal strategies vary between different systems and locations.