• SCVS Home
• President's Message
• Council
• Commitees
• Annual Meeting
• 37th Annual Symposium
• 36th Annual Symposium
• 35th Annual Symposium
• 34th Annual Symposium
• 33rd Annual Symposium
• Membership
• Join SCVS
• Members Only Area
• Find a Vascular Specialist
• Newsletters
• Journal
• History
• Mission Statement
• Bylaws
• Contact Us
• 37th Annual Symposium
• 36th Annual Symposium
• 35th Annual Symposium
• 34th Annual Symposium
• 33rd Annual Symposium
• Join SCVS
• Members Only Area
• Find a Vascular Specialist

Prospective Analysis of Hospital Length of Stay

Introduction: Pysicians have been held accountable for target LOS predicted by statistical models based upon administrative data. It is uncertain whether these predicted LOS correspond to appropriate/inappropriate hospitalizations for individual patients. This study was designed to prospectively analyze the appropriateness of inpatient hospitalizations on an academic vascular surgery service.
Methods: Patients were evaluated on a daily basis by a multidisciplinary team. The appropriateness of hospitalization was determined independently by an attending surgeon and a discharge coordinator for consecutive months (October/November 2006) using standard criteria for hospitalization. A variance category (e.g. Agency Related, Physician Related) and variance code (e.g. Avoidable Preoperative Days, Diagnostic Service Delays) were assigned to each inappropriate day. The actual LOS were compared to those predicted by the institutional models.
Results: 105 patients were hospitalized on the vascular service during the study dates involving 116 separate episodes of care. These accounted for 825 inpatient days. Among the total hospital days, 9% (76/825) were considered inappropriate. The actual mean LOS was significantly greater than predicted by the statistical model (10.8 + 16.4 (SD) vs 8.2 + 19.8, p = .05) and would have been greater than predicted even if all of the inappropriate days were eliminated. Breakdown of the inappropriate days by the variance codes demonstrated that surgeon/operating room availability and lack of a facility for discharge placement accounted for the greatest number of days (Figure). Patients admitted urgently/emergently accounted for 95% (39/41) of the inappropriate hospital days attributed to surgeon/operating room availability.
Conclusions: There are significant opportunities to reduce inpatient LOS although the predicted LOS determined by the statistical models may not be realistic. Potential solutions to reduce LOS include reserving time within surgeon/operating room schedules to accommodate the urgent/emergent cases and earlier identification of patients with financial/social constraints.