Greenslade, Jaimi

Background: Variations in cardiac troponin concentrations by age, sex, and time between samples in patients with suspected myocardial infarction are not currently accounted for in diagnostic approaches. We aimed to combine these variables through machine learning to improve the assessment of risk for individual patients. Methods: A machine learning algorithm (myocardial-ischemic-injury-index [MI 3 ]) incorporating age, sex, and paired high-sensitivity cardiac troponin I concentrations, was trained on 3013 patients and tested on 7998 patients with suspected myocardial infarction. MI 3 uses gradient boosting to compute a value (0–100) reflecting an individual’s likelihood of a diagnosis of type 1 myocardial infarction and estimates the sensitivity, negative predictive value, specificity and positive predictive value for that individual. Assessment was by calibration and area under the receiver operating characteristic curve. Secondary analysis evaluated example MI 3 thresholds from the training set that identified patients as low risk (99% sensitivity) and high risk (75% positive predictive value), and performance at these thresholds was compared in the test set to the 99th percentile and European Society of Cardiology rule-out pathways. Results: Myocardial infarction occurred in 404 (13.4%) patients in the training set and 849 (10.6%) patients in the test set. MI 3 was well calibrated with a very high area under the receiver operating characteristic curve of 0.963 [0.956–0.971] in the test set and similar performance in early and late presenters. Example MI 3 thresholds identifying low- and high-risk patients in the training set were 1.6 and 49.7, respectively. In the test set, MI 3 values were <1.6 in 69.5% with a negative predictive value of 99.7% (99.5–99.8%) and sensitivity of 97.8% (96.7–98.7%), and were ≥49.7 in 10.6% with a positive predictive value of 71.8% (68.9–75.0%) and specificity of 96.7% (96.3–97.1%). Using these thresholds, MI 3 performed better than the European Society of Cardiology 0/3-hour pathway (sensitivity, 82.5% [74.5–88.8%]; specificity, 92.2% [90.7–93.5%]) and the 99th percentile at any time point (sensitivity, 89.6% [87.4–91.6%]); specificity, 89.3% [88.6–90.0%]). Conclusions: Using machine learning, MI 3 provides an individualized and objective assessment of the likelihood of myocardial infarction, which can be used to identify low- and high-risk patients who may benefit from earlier clinical decisions. Clinical Trial Registration: URL: https://www.anzctr.org.au . Unique identifier: ACTRN12616001441404.

AbstractBACKGROUNDIncreased cardiac troponin I or T detected by high-sensitivity assays (hs-cTnI or hs-cTnT) confers an increased risk of adverse prognosis. We determined whether patients presenting with putatively normal, detectable cTn concentrations [&gt; limit of detection and &lt; upper reference limit (URL)] have increased risk of major adverse cardiovascular events (MACE) or all-cause mortality.METHODSA prospective 5-year follow-up of patients recruited in the emergency department with possible acute coronary syndrome (ACS) and cTn concentrations measured with hs-cTnI (Abbott) and hs-cTnT (Roche) assays. Cox regression models were generated with adjustment for covariates in those without MACE on presentation. Hazard ratios (HRs) for hs-cTn were calculated relative to the HRs at the median concentration.RESULTSOf 1113 patients, 836 were without presentation MACE. Of these, 138 incurred a MACE and 169 died during a median 5.8-year follow-up. HRs for MACE at the URLs were 2.3 (95% CI, 1.7–3.2) for hs-cTnI and 1.8 (95% CI, 1.3–2.4) for hs-cTnT. Corresponding HRs for mortality were 1.7 (95% CI, 1.2–2.2) for hs-cTnI and 2.3 (95 % CI, 1.7–3.1) for hs-cTnT. The HR for MACE increased with increasing hs-cTn concentration similarly for both assays, but the HR for mortality increased at approximately twice the rate for hs-cTnT than hs-cTnI. Patients with hs-cTnI ≥10 ng/L or hs-cTnT ≥16 ng/L had the same percentage of MACE at 5-year follow-up (33%) as patients with presentation MACE.CONCLUSIONSMany patients with ACS ruled out and putatively normal but detectable hs-cTnI concentrations are at similar long-term risk as those with MACE. hs-cTnT concentrations are more strongly associated with 5-year mortality than hs-cTnI.

Abstract Objective Risk scores and accelerated diagnostic protocols can identify chest pain patients with low risk of major adverse cardiac event who could be discharged early from the ED , saving time and costs. We aimed to derive and validate a chest pain score and accelerated diagnostic protocol ( ADP ) that could safely increase the proportion of patients suitable for early discharge. Methods Logistic regression identified statistical predictors for major adverse cardiac events in a derivation cohort. Statistical coefficients were converted to whole numbers to create a score. Clinician feedback was used to improve the clinical plausibility and the usability of the final score (Emergency Department Assessment of Chest pain Score [ EDACS ]). EDACS was combined with electrocardiogram results and troponin results at 0 and 2 h to develop an ADP ( EDACS‐ADP ). The score and EDACS‐ADP were validated and tested for reproducibility in separate cohorts of patients. Results In the derivation ( n = 1974) and validation ( n = 608) cohorts, the EDACS‐ADP classified 42.2% (sensitivity 99.0%, specificity 49.9%) and 51.3% (sensitivity 100.0%, specificity 59.0%) as low risk of major adverse cardiac events, respectively. The intra‐class correlation coefficient for categorisation of patients as low risk was 0.87. Conclusion The EDACS‐ADP identified approximately half of the patients presenting to the ED with possible cardiac chest pain as having low risk of short‐term major adverse cardiac events, with high sensitivity. This is a significant improvement on similar, previously reported protocols. The EDACS‐ADP is reproducible and has the potential to make considerable cost reductions to health systems.