Predicting a diagnosis of acute coronary syndrome during telephone evaluation by an emergency dispatcher: the SCARE predictive scale.

OBJECTIVES

Correctly identifying patients with acute coronary syndrome (ACS) on first contact is essential, yet emergency dispatchers currently lack a risk scale that can help predict an ACS diagnosis. Our main aim was to develop and validate such a risk scale.

MATERIAL AND METHODS

Prospective, observational single-center study in 2016 (January 1 to December 31). We included patients who called our emergency dispatch center to report nontraumatic chest pain. Included patients were randomly assigned to a development or a validation sample. The predictive SCARE scale was built with logistic regression analysis. Discrimination and calibration were analyzed by calculating the area under the receiver operating characteristic curve; calibration was assessed with the Hosmer-Lemeshow test.

RESULTS

The development sample included 902 patients. The regression model identified 7 variables associated with a final diagnosis of ACS: male sex, age, smoking, typical pain characteristics, first episode of chest pain, diaphoresis, and physician intuition (the teledispatcher's suspicion). When we applied the scale in the validation sample of 465 patients the area under the curve was 0.81 (95% CI, 0.76-0.87). The Hosmer-Lemeshow statistic was 5.18 (P=.74).

CONCLUSION

The SCARE scale had good discrimination and calibration properties. The scale should be further validated in an external sample from a multicenter study before it is implemented by emergency dispatch centers.