OBJECTIVE: To evaluate previously developed algorithms for the improvement of pre-hospital triage of patients with suspected acute cardiac disease. DESIGN: Prospective study. SUBJECTS: Patients with symptoms of possible cardiac origin, who were seen by a general practitioner and subsequently referred to hospital in the municipality of Rotterdam. METHODS: Prior to hospital admission, patients with suspected acute coronary disease recorded their symptoms by standardized questionnaire and a computerized ECG was made. All patients were hospitalized and a final diagnosis was established. Algorithms proposed by other investigators to distinguish patients with, from those without, acute cardiac disease were tested. MAIN OUTCOME: Identification of patients whose likelihood for acute cardiac pathology was low (stable angina, atypical chest pain, other pathology) or high (myocardial infarction, unstable angina). RESULTS: A total of 1005 patients were studied. Forty-two percent had myocardial infarction or unstable angina pectoris. Evaluation of previously developed algorithms showed that their diagnostic accuracy was poor in the pre-hospital setting. In a separate multivariate analysis, six characteristics from the clinical history and an electrocardiogram appeared to be independently and significantly associated with acute cardiac pathology. The presence of an abnormal ECG proved to be the most important predictor. CONCLUSIONS: The hospital-based algorithms were unsuitable as a predictor for pre-hospital acute cardiac pathology. A new practical hospital admission model was developed, based on six clinical predictors, including analysis of an electrocardiogram. Following appropriate validation, this out-of-hospital protocol may lead to better triage decisions by the general practitioner.
OBJECTIVE: To evaluate previously developed algorithms for the improvement of pre-hospital triage of patients with suspected acute cardiac disease. DESIGN: Prospective study. SUBJECTS:Patients with symptoms of possible cardiac origin, who were seen by a general practitioner and subsequently referred to hospital in the municipality of Rotterdam. METHODS: Prior to hospital admission, patients with suspected acute coronary disease recorded their symptoms by standardized questionnaire and a computerized ECG was made. All patients were hospitalized and a final diagnosis was established. Algorithms proposed by other investigators to distinguish patients with, from those without, acute cardiac disease were tested. MAIN OUTCOME: Identification of patients whose likelihood for acute cardiac pathology was low (stable angina, atypical chest pain, other pathology) or high (myocardial infarction, unstable angina). RESULTS: A total of 1005 patients were studied. Forty-two percent had myocardial infarction or unstable angina pectoris. Evaluation of previously developed algorithms showed that their diagnostic accuracy was poor in the pre-hospital setting. In a separate multivariate analysis, six characteristics from the clinical history and an electrocardiogram appeared to be independently and significantly associated with acute cardiac pathology. The presence of an abnormal ECG proved to be the most important predictor. CONCLUSIONS: The hospital-based algorithms were unsuitable as a predictor for pre-hospital acute cardiac pathology. A new practical hospital admission model was developed, based on six clinical predictors, including analysis of an electrocardiogram. Following appropriate validation, this out-of-hospital protocol may lead to better triage decisions by the general practitioner.
Authors: Peter Hollender; David Bradway; Patrick Wolf; Robi Goswami; Gregg Trahey Journal: IEEE Trans Ultrason Ferroelectr Freq Control Date: 2013-08 Impact factor: 2.725