Although guidelines recommend fixed cardiac troponin thresholds for the diagnosis of myocardial infarction, troponin concentrations are influenced by age, sex, comorbidities and time from symptom onset. To improve diagnosis, we developed machine learning models that integrate cardiac troponin concentrations at presentation or on serial testing with clinical features and compute the Collaboration for the Diagnosis and Evaluation of Acute Coronary Syndrome (CoDE-ACS) score (0–100) that corresponds to an individual’s probability of myocardial infarction. The models were trained on data from 10,038 patients (48% women), and their performance was externally validated using data from 10,286 patients (35% women) from seven cohorts. CoDE-ACS had excellent discrimination for myocardial infarction (area under curve, 0.953; 95% confidence interval, 0.947–0.958), performed well across subgroups and identified more patients at presentation as low probability of having myocardial infarction than fixed cardiac troponin thresholds (61 versus 27%) with a similar negative predictive value and fewer as high probability of having myocardial infarction (10 versus 16%) with a greater positive predictive value. Patients identified as having a low probability of myocardial infarction had a lower rate of cardiac death than those with intermediate or high probability 30 days (0.1 versus 0.5 and 1.8%) and 1 year (0.3 versus 2.8 and 4.2%; P < 0.001 for both) from patient presentation. CoDE-ACS used as a clinical decision support system has the potential to reduce hospital admissions and have major benefits for patients and health care providers.

BACKGROUND: Low concentrations of cardiac troponin (cTn) have been recommended for rapid rule-out of acute myocardial infarction (AMI). We examined the Beckman Coulter Access high-sensitivity cardiac troponin I (hs-cTnI) assay to identify a single test threshold that can safely rule out AMI.METHODS: This analysis used stored samples collected in 2 prospective observational studies. In all, 1871 patients presenting to a tertiary emergency department with symptoms of acute coronary syndrome had blood taken for measurement of cTnI on presentation. The endpoint was type 1 myocardial infarction (T1MI). Sensitivity and negative predictive value (NPV) were calculated for hs-cTnI values below the 99th percentile.RESULTS: Ninety-eight patients had T1MI (5.2%), and 638 (34.1%) patients had an hs-cTnI <2 ng/L (limit of detection), with sensitivity of 99.0% (95% CI, 94.4%–100%) and NPV of 99.8% (95% CI, 99.1%–100%). No hs-cTnI value above a concentration of 2 ng/L achieved sensitivity of 99%. However, an NPV of 99.5% was achieved at values <6 ng/L. A cutoff <6 ng/L enabled 1475 (78.8%) patients to be ruled out on presentation with sensitivity of 93.9% (95% CI, 87.1%–97.7%).CONCLUSIONS: A single baseline cTn <2 ng/L measured with the Access hs-cTnI assay performed well for rule-out of AMI. This cutoff concentration identified 99% of patients with AMI and could reduce the number of patients requiring lengthy assessment. A cutoff of <6 ng/L yielded a high NPV but missed more cases of AMI than would be acceptable to clinicians.

Strategies to assess patients with suspected acute myocardial infarction (AMI) using a point-of-care (POC) high-sensitivity cardiac troponin I (hs-cTnI) assay may expedite emergency care. A 2-h POC hs-cTnI strategy for emergency patients with suspected AMI was derived and validated. In two international, multi-centre, prospective, observational studies of adult emergency patients (1486 derivation cohort and 1796 validation cohort) with suspected AMI, hs-cTnI (Siemens Atellica^® VTLi) was measured at admission and 2 h later. Adjudicated final diagnoses utilized the hs-cTn assay in clinical use. A risk stratification algorithm was derived and validated. The primary diagnostic outcome was index AMI (Types 1 and 2). The primary safety outcome was 30-day major adverse cardiac events incorporating AMI and cardiac death. Overall, 81 (5.5%) and 88 (4.9%) patients in the derivation and validation cohorts, respectively, had AMI. The 2-h algorithm defined 66.1% as low risk with a sensitivity of 98.8% [95% confidence interval (CI) 89.3%–99.9%] and a negative predictive value of 99.9 (95% CI 99.2%–100%) for index AMI in the derivation cohort. In the validation cohort, 53.3% were low risk with a sensitivity of 98.9% (95% CI 92.4%–99.8%) and a negative predictive value of 99.9% (95% CI 99.3%–100%) for index AMI. The high-risk metrics identified 5.4% of patients with a specificity of 98.5% (95% CI 96.6%–99.4%) and a positive predictive value of 74.5% (95% CI 62.7%–83.6%) for index AMI. A 2-h algorithm using a POC hs-cTnI concentration enables safe and efficient risk assessment of patients with suspected AMI. The short turnaround time of POC testing may support significant efficiencies in the management of the large proportion of emergency patients with suspected AMI. Performance of the 0-/2-h algorithm for risk assessment of myocardial infarction in the Suspected Acute Myocardial Infarction in Emergency (SAMIE) and Safe Emergency Department Discharge Rate (SEIGE) cohorts. AMI, acute myocardial infarction; CI, confidence interval; hs-cTnI, high-sensitivity cardiac troponin I; NPV, negative predictive value; POC, point of care; PPV, positive predictive value.