Reducing lengths of stay in the coronary care unit with a practice guideline for patients with congestive heart failure. Insights from a controlled clinical trial.

Although more than 1,000 medical practice guidelines have been developed, there have been few evaluations of their use in clinical practice or information to judge whether practice guidelines can be used to reduce health care costs. For this reason, the authors conducted a prospective controlled clinical trial with an alternating-month design at a large teaching community hospital to study the use of a practice guideline to promote early transfer of patients admitted to a hospital with congestive heart failure (CHF) from the coronary care unit (CCU) and intermediate care unit to unmonitored beds. The practice guideline was supported by locally derived risk information and recommended consideration of early "step-down" transfer of low-risk patients with CHF 24 hours after hospital admission. Physicians caring for patients identified as "low risk" received concurrent personalized written and verbal reminders concerning the guideline recommendation. Study subjects were patients admitted to a hospital CCU and intermediate care unit between November 1, 1991 and April 30, 1993 with a diagnosis of CHF or pulmonary edema. Ninety patients with CHF were identified as low risk according to the guideline during the study period. Feedback of the practice guideline recommendation was not associated with a significant increase in physician adoption of the guideline or shorter lengths of stay in the CCU or intermediate care unit. Physicians may have compensated for statistically insignificant reductions in monitored lengths of stay by increasing the length of stay in unmonitored beds (1.80 +/- 2.32 to 4.02 +/- 4.09 days, P = .002) and the total length of stay (4.73 +/- 2.43 to 6.71 +/- 5.44 days, P = .03). Quality of patient care, patient outcomes, and patient satisfaction were not affected by the guideline. Our study results suggest that implementation of a locally derived practice guideline for patients with CHF did not result in adoption of the guideline by physicians. The complexity of implementing the guideline, changes in physician practice before the study, and the failure of the guideline to address the continuum of patient care across monitored and unmonitored beds may have accounted for rejection of the guideline. Our experience demonstrates that practice guidelines, whenever possible, should be evaluated in prospective trials before they should be disseminated for widespread use.