Background: Observation care is intended for short-stay hospitalizations requiring diagnostic clarification or brief therapy; however, prolonged observation stays contribute to operational inefficiency and revenue loss. Between June 2024 and May 2025, 25.1% of Overlook Medical Center (OMC) discharges were in observation status, with an average length of stay (LOS) of 35.49 hours (including ED hold time). Extended observation stays also posed financial risk, as Medicare does not reimburse hospitals for patients exceeding two midnights without inpatient conversion. The hospital set a goal to reduce average observation LOS by 10%—to under 31.94 hours—while maintaining quality and appropriate utilization.
Methods: A multidisciplinary Lean project team was formed, co-led by the Associate Chief Medical Officer and the Director of Hospitalist Medicine, with support from Utilization Review (UR), Case Management (CM), Nursing, and Physical Therapy. The team conducted root cause analysis to identify key drivers of excessive LOS, including clinical variation, inconsistent unit placement, weekend discharge delays, and limited after-hours coverage. Interventions were implemented through a series of “tests of change”: Clinical Pathway Standardization: Developed decision trees for top observation diagnoses (chest pain, dizziness, dehydration, syncope, TIA). Geographic Cohorting: Reduced observation footprint from nine floors to core units (4W/6C, 2AB, 4AB, 7CD) to streamline workflows. Expanded UR/CM Coverage: Introduced daily huddles, extended UR hours, and early morning physician touchpoints. Education and Denial Mitigation: Rolled out “Top 5 DRGs Denial Mitigation Guide” for Hospitalists and UR staff. ED Scripting: Implemented standardized communication to improve patient understanding of observation vs. inpatient status. PT and SAR Optimization: Advanced PT consults and same-day authorization processes for discharge to subacute rehab.
Results: By October 2025, 40 hospitalists and interdisciplinary leaders participated in the initiative, reviewing 4,500 observation encounters. Early results demonstrated a reduction in average LOS from 35.5 to 33.1 hours (≈9% improvement) and a marked decrease in floor variation, with the core observation unit (4W/6C) achieving a 32.8-hour average LOS. Only 2.9% of observation patients exceeded 90 hours, down from 4.1% baseline. Weekend discharge delays decreased as UR coverage expanded, and observation rate stability (≈25%) confirmed accurate patient selection rather than status inflation. Financial modeling projected improved reimbursement capture for Medicare patients under the two-midnight rule.
Conclusions: This initiative highlights how Lean standardization and workflow redesign can reduce observation LOS without new staffing or technology. Cohorting observation of care on designated units, expanding weekend and after-hours support, and clarifying roles between UR and hospitalists led to measurable operational gains. Sustaining progress will require continued monitoring of unit-level variation and addressing residual barriers such as physical space constraints and occasional delays in diagnostic testing. OMC’s experience demonstrates that efficient observation management is achievable through cultural alignment, cross-department collaboration, and data-driven process design.