Diagnostic value of a heart-type fatty acid-binding protein (H-FABP) bedside test in suspected acute coronary syndrome in primary care.

BACKGROUND: To determine the diagnostic accuracy of a rapid heart-type fatty acid-binding protein (H-FABP) test in patients suspected of acute coronary syndrome (ACS) in primary care. METHODS: General practitioners included 298 patients suspected of ACS. In all patients, whether referred to hospital or not, ECG and cardiac biomarker testing was performed. ACS was determined in accordance with international guidelines. Multivariate analysis was used to determine the value of H-FABP in addition to clinical findings. RESULTS: Mean patient age was 66 years (SD 14), 52% was female and 66 patients (22%) were diagnosed with ACS. The H-FABP bedside test was performed within 24h (median 3.1, IQR 1.5 to 7.1) after symptom onset. The positive predictive value (PPV) of H-FABP was 65% (95% confidence interval (CI) 50-78). The negative predictive value (NPV) was 85% (95% CI 80-88). Sensitivity was 39% (29-51%) and specificity 94% (90-96%). Within 6h after symptom onset, the PPV was 72% (55-84) and the NPV was 83% (77-88), sensitivity 43% (31-57%) and specificity 94% (89-97%). Adding the H-FABP test to a diagnostic model for ACS led to an increase in the area under the receiver operating curve from 0.66 (95% CI 0.58-0.73) to 0.75 (95% CI 0.68-0.82). CONCLUSION: The H-FABP rapid test provides modest additional diagnostic certainty in primary care. It cannot be used to safely exclude rule out ACS. The test can only be used safely in patients otherwise NOT referred to hospital by the GP, as an extra precaution not to miss ACS ('rule in').

Diagnosing acute coronary syndrome in primary care: comparison of the physicians' risk estimation and a clinical decision rule.

BACKGROUND: Diagnosing acute coronary syndrome (ACS) in a primary care setting poses a diagnostic dilemma for physicians. OBJECTIVE: We directly compared the diagnostic accuracy of a clinical decision rule (CDR) based on history taking and physical examination in suspected ACS with the risk estimates of the attending GP. METHODS: In a prospective multicenter study, patients suspected of ACS were included by the GP. GPs were asked to estimate the probability (0%-100%) of the presence of ACS. GPs collected patient data, but they were not aware of the CDR and did not score the patient accordingly. RESULTS: Two hundred and ninety-eight patients were included (52% female, mean age 66 years, 22% ACS). The area under the receiver operating characteristic (ROC) curve (AUC) was 0.75 [95% confidence interval (CI) 0.68-0.82] for the GP risk estimate and 0.66 (95% CI 0.58-0.73) for the CDR. There was an agreement between the risk estimation of the GP and a CDR in 51% and the prevalence of ACS in predefined low-, intermediate- and high-risk groups was similar for the GP and CDR estimates. In the low-risk group, according to the GP, four patients (8.2%) suffered an ACS. These four patients were all identified by the decision rule as high risk. CONCLUSIONS: The GP classified patients as ACS or no ACS more adequately than the CDR, judged by the AUC. However, the use of a CDR in patients that are considered at low risk for ACS by the GP could reduce the amount of missed myocardial infarctions.