Are Critical Pathways Effective for Reducing Postoperative Length of Stay?

• Published in: Medical care 2003-05, Vol.41 (5), p.637-648
• Main Authors: Dy, Sydney Morss, Garg, Pushkal P., Nyberg, Dorothy, Dawson, Patricia B., Pronovost, Peter J., Morlock, Laura, Rubin, Haya R., Diener-West, Marie, Wu, Albert W.
• Format: Article
• Language: English
• Subjects: Adult; Baltimore; Cohort Studies; Comorbidity; Critical Pathways; Female; Health outcomes; Health Services Research; Hospital admissions; Hospitals, University - standards; Hospitals, University - utilization; Humans; Length of stay; Length of Stay - statistics & numerical data; Length of Stay - trends; Linear regression; Male; Middle Aged; Outcome and Process Assessment (Health Care); Postoperative care; Postoperative Care - utilization; Rates of change; Retrospective Studies; Surgery Department, Hospital - standards; Surgery Department, Hospital - utilization; Surgical procedures; Surgical Procedures, Operative - classification; Surgical Procedures, Operative - standards; Surgical specialties; United States
• ISSN: 0025-7079; 1537-1948
• DOI: 10.1097/01.MLR.0000062552.92534.BE

Background. Many hospitals use critical pathways to attempt to reduce postoperative length of stay (PLOS) for diverse conditions and procedures. Objective. To evaluate whether critical pathways were associated with reductions in postoperative PLOS after accounting for prepathway trends in PLOS. Research Design. Retrospective cohort study, from 1988 to 1998. Setting. Academic medical center department of surgery. Subjects. A total of 10,960 admissions eligible for 1 of 26 critical pathways implemented from 1990 to 1996, from 2 years before to 2 years after each pathway implementation date. Coding definitions were developed and validated to identify admissions eligible for each pathway, and data were abstracted from the hospital's discharge database. Measure. A pathway was considered effective if, after its implementation, there was a statistically significant decrease in the prepathway trend for PLOS. Results. Median number of annual eligible admissions per pathway was 59 (range, 18-706). Median PLOS for the prepathway periods was 8 days (interquartile range, 5-10 days). For 16 (62%) pathways, PLOS was already declining in the prepathway period. After adjusting for demographics, comorbidity, admission characteristics, and prepathway time trends in PLOS, 7 (27%) pathways were associated with a significant postimplementation decrease in the rate of change in PLOS (range among the 7 pathways, 5-45% decrease) and none with a significant increase from the prepathway trend for PLOS. Conclusion. Critical pathways may decrease postoperative stay for some, but not all, surgeries. Trends toward decreasing length of stay over time may reduce the impact of critical pathways on this outcome.