A computer protocol to evaluate subjects with chest pain in the emergency department: a multicenter study.

OBJECTIVE

Chest pain is a frequent cause of medical admission to the emergency department and the main differential diagnosis is between coronary and non-coronary chest pain. We elaborated a computer protocol for the management of patients with chest pain.

METHODS

The computer protocol was made of three sections according to clinical, electrocardiographic and biochemical data. Each section was coded by a letter indicating the probability of coronary chest pain for each section. The combination of the three letters formed a score string used to assign patients to four subgroups of overall probability of coronary chest pain (low, medium-low, medium-high, and high). Low-probability patients were discharged from the emergency department, whereas high-probability patients were admitted to the coronary care unit. The medium-probability patients underwent further evaluation by means of a stress test and were re-classified as having a final low probability (negative test) or high probability (positive test).

RESULTS

We evaluated 472 patients (mean age 64 years, range 18-97 years; 47% female). The incidence of coronary events in patients with low, medium-low, medium-high and high overall probability was 1.9, 12.8,13.5 and 68.0%, respectively (P < 0.05). The positive and negative predictive values of the protocol were 64.7 and 97.1%, respectively.

CONCLUSIONS

Our computer protocol represents a reliable method for the management of patients with chest pain and a non-diagnostic electrocardiogram.