Determining chest pain patients' suitability for transfer to a general ward following admission to a cardiac care unit

The objective of this study was to assess the performance of a consensus-derived decision algorithm in determining chest pain patients' suitability for early transfer to a lower dependency ward by predicting complications. The sample comprised 516 patients with chest pain presumed to be cardiac in origin, admitted to a cardiac care unit (CCU) in northern England from the community or from the accident and emergency department. A decision algorithm was designed following a review of the literature and amended to take into account a clinical consensus of consultant physicians. Patients were assessed on admission by CCU nurses using the algorithm, and 'triage' decisions recorded (keep on CCU or suitable for early transfer to a general ward). Admission ECGs (electrocardiographs) and baseline clinical data were recorded independently by a researcher 'blinded' to actual clinical course, and applied to the algorithm using statistical software. On discharge or death, patients' case notes were retrieved and the hospital course examined for death or severe complications. Performance of the algorithm and CCU nurses were compared for sensitivity, specificity, positive predictive value and negative predictive value. The main outcome measures were death or severe complications occurring during hospitalization, and during the first 2 days following CCU admission. While sensitivity of both the algorithm (0.98) and CCU nurses (0.95) was high, specificity was low in both groups (0.11 and 0.21, respectively), making it unlikely that the algorithm would prove useful in clinical practice. Further studies are required to develop the optimal triage tool for the assessment of patients with acute chest pain.