Syncope, a transient loss of consciousness followed by full recovery, is common but difficult to manage. Even after a complete evaluation, in many cases there is no clear reason for syncope. However, up to 30 percent of high-risk patients die within the first year. Risk stratification of patients can help direct the clinical evaluation. Although guidelines are available to help identify patients who are at higher risk for adverse outcomes, they have not been studied prospectively.

Quinn and associates used a prospective study at a single hospital, following patients presenting to the emergency department with syncope or near syncope, to derive the San Francisco Syncope Rule. Enrolled patients were followed for seven days to identify any serious outcomes. Predictor variables—historical, related to the physical examination, and diagnostic tests—were studied. Serious outcomes were defined as death, myocardial infarction, arrhythmia, pulmonary embolism, stroke, subarrachnoid hemorrhage, significant hemorrhage, or any other event likely to bring the patient back to the emergency department.

Of the 684 patients with syncope, 79 (11.5 percent) developed serious outcomes by day 7. During the study, 55 percent of all patients presenting to the emergency department with syncope were admitted to the hospital. Of the 50 predictor variables analyzed, 26 (52 percent) were significantly associated with a serious outcome. Using multivariate analysis, a set of prediction variables was identified that could improve the prediction of serious outcomes (see accompanying table). In the study population, this rule had a 96.2 percent sensitivity for identifying patients who had a serious outcome and a specificity of 61.9 percent. It is estimated that use of this rule would place about 45 percent of patients with syncope in the high-risk category, which might decrease the number of patients requiring hospital admission.

The authors conclude that using the San Francisco Syncope Rule to predict serious outcomes of syncope will help physicians determine which patients must be admitted. Further prospective validation is recommended before physicians actively use this decision rule in practice.

In an editorial in the same journal, Gallagher supports the methodology used to derive the San Francisco Syncope Rule. The rule has 96 percent sensitivity, but the dilemma is how to manage the one syncope patient out of 25 (4 percent) who is at risk for serious outcomes but is not identified as such by this decision rule. This limitation illustrates the need to use this rule, once it is validated, as a tool that helps with decision making but not as a basis for dictating binary action because of its lack of 100 percent sensitivity.