Simone van den Bulk, Annelieke Hj Petrus, Robert Willemsen, Mattijs E Numans, Mark J Boogers, Maaike Pj Hermans, Joan G Meeder, Braim Rahel, Marcel Janssen, Geertjan Dinant, Christa M Cobbaert, Tobias N Bonten
Background Accurate risk stratification of patients with chest pain in primary care may improve patient safety while reducing unnecessary hospital referrals. Aim To evaluate the diagnostic performance of a clinical decision rule (CDR) combining high-sensitivity tro-ponin I (hs-cTnI) and the Marburg Heart Score (MHS) to rule out acute coronary syndrome (ACS), and specifically myocardial infarction (MI), in primary care patients with chest pain, and to compare referral rates with usual care. Design and setting Cluster-randomised diagnostic trial (2:1 intervention:control) conducted in Dutch primary care practices. Method Patients presenting with acute chest pain were assessed using a CDR applied by general practitioners, combining the MHS with a hs-cTnI point-of-care test. General practitioners in the control group provided usual care. Diagnostic performance was evaluated using sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). Secondary endpoints included evaluation of a modified HEART score and single hs-cTnI rule-out strategy. Results In total, 827 participants were included: 740 intervention and 87 controls. The CDR missed one ACS, yielding a sensitivity of 98.3% (95% CI 90.8-100%) and NPV of 99.7% (95% CI 98.0-100%). Specificity was 49.1% (95% CI 45.3-52.9%) and PPV 14.1% (95% CI 13.2-15.1%). For MI, sensitivity and NPV were both 100% (95% CI 92.6-100% and 95% CI 98.9-100%, respectively). Specificity was 48.6% (95% CI 44.8-52.4%), and PPV 11.9% (95% CI 11.1-12.7%). Referral rates did not differ significantly between groups (54.6% vs 47.1%; p=0.39). Conclusion A CDR combining the MHS with a hs-cTnI point-of-care test safely ruled out ACS and MI in(..).