Congestive heart failure (CHF) is a serious condition associated with high hospitalization and mortality rates. Guidelines published in 1994 by the Agency for Health Care Policy and Research (AHCPR), now the Agency for Healthcare Research and Quality (AHRQ), specify admission criteria for patients presenting with CHF. Although these guidelines were developed by a consensus of experts using recommendations of the American College of Cardiology and the American Heart Association, there has been no clinical trial of their validity. Graff and associates quantified patient risk as related to the presence or absence of AHCPR CHF hospitalization admission criteria.

Using a quality improvement initiative organized through the Connecticut Peer Review Organization, 12 acute-care hospitals reviewed medical records of patients admitted from the emergency department with the diagnosis of CHF who were discharged after January 1, 1995. Charts were abstracted for 192 variables useful in predicting patient risk of death. Patients were excluded if they had a primary diagnosis of acute myocardial infarction or some other acute medical condition contributing to CHF, such as thyrotoxicosis, cardiac tamponade or significant valvular disease. Patients had to have clear clinical evidence of heart failure and one or more physical or radiologic signs confirming the diagnosis. For a listing of the AHCPR hospital admission criteria, see the accompanying table. Two of these criteria (e.g., inadequate social support or failure of outpatient management) could not be considered because the hospital medical records did not identify whether these criteria were present. Mortality rates were considered the primary outcome measure and were noted at 30 days, six months and one year.

A total of 1,674 patients with CHF was enrolled in the study. Pulmonary edema and respiratory distress were the most common admission criteria, followed by clinical evidence of ischemia and CHF of recent onset. Eighty percent (1,340) of the patients with CHF were admitted, and 20 percent (334) were released from the emergency department. Patients in the latter group had lower mortality rates at all three end points. Patient survival was correlated with individual AHCPR criteria, with four criteria correlating with a higher mortality rate: (1) pulmonary edema, (2) hypoxia, (3) syncope and (4) significant edema. Two of the criteria, ischemic symptoms and recent-onset of CHF, did not correlate with poorer patient outcomes.

The 80 percent hospitalization rate associated with use of these criteria represents a high use of resources. A correlation was found between physician judgment and the AHCPR admission criteria; therefore, when used together, these two criteria systems can improve sensitivity and specificity while lowering the hospitalization rate.

The authors conclude that selected AHCPR CHF admission criteria correlate with increased risk. These criteria, combined with the physician's clinical judgment, may help in risk stratification of patients with CHF. Patients identified by the criteria as low risk who are now cared for in the acute setting may be candidates for outpatient management, given the advent of more intensive out-patient care programs.