Can practice guidelines safely reduce hospital length of stay? Results from a multicenter interventional study.

BACKGROUND: Although practice guidelines about appropriate lengths of stay have been widely promulgated, their effects on patient outcomes are not clear. Our objective was to study the effects of length of stay practice guidelines on patient outcomes. PATIENTS AND METHODS: We performed a prospective, nonrandomized, interventional trial in six geographically distributed hospitals, among consecutively hospitalized "low-risk" patients with total hip replacement, hip fracture, or knee replacement. Case managers provided physicians with patient risk information based on guideline recommendations. We measured length of stay, compliance with recommended guideline length of stay, health status, hospital readmission rates, return to emergency department, return to work and recreation, and patient satisfaction.
RESULTS: A total of 560 patients were included in the study. For patients with knee replacement, there was a statistically significant increase in practice guideline compliance (27% baseline versus 53% intervention, P <0.0001) and reduction in length of stay (5.2 days versus 4.6 days, P <0.001) when compared with the baseline period. For hip replacement patients, there similarly was an increase in practice guideline compliance (66% baseline versus 82% intervention, P = 0.01) and reduction in length of stay (5.1 days versus 4.8 days, P = 0.03). Significant reductions in length of stay were not observed for patients recovering after hip fracture despite a significant increase in guideline compliance. There were few statistically significant changes in patient outcomes related to reductions in lengths of stay, including health status, hospital readmission rates, return to emergency department, return to work and recreation, and patient satisfaction. For patients undergoing hip replacement, very short lengths of stay (shorter than the guideline recommendation) were associated with an increased rate of discharging patients to nursing homes and rehabilitation facilities (21% versus 7%, P = 0.01), and hip fracture patients with very short lengths of stay required more visits to the doctor after discharge (56% versus 25%, P = 0.04).
CONCLUSION: Reductions in lengths of stay were most often associated with no significant change in patient outcomes. However, very short lengths of stay were associated with increased intensity of care following discharge for patients undergoing hip surgery, indicating possible cost shifting (the cost incurred by transferring patients to rehabilitation facilities may have been greater than had the patients remained in the acute care hospital for an additional 1 or 2 days and been sent directly home). These results emphasize the importance of monitoring the effects of cost containment and other systematic efforts to change patient care at the local level.