A machine learning algorithm to predict a culprit lesion after out of hospital cardiac arrest.

BACKGROUND: We aimed to develop a machine learning algorithm to predict the presence of a culprit lesion in patients with out-of-hospital cardiac arrest (OHCA). METHODS: We used the King's Out-of-Hospital Cardiac Arrest Registry, a retrospective cohort of 398 patients admitted to King's College Hospital between May 2012 and December 2017. The primary outcome was the presence of a culprit coronary artery lesion, for which a gradient boosting model was optimized to predict. The algorithm was then validated in two independent European cohorts comprising 568 patients. RESULTS: A culprit lesion was observed in 209/309 (67.4%) patients receiving early coronary angiography in the development, and 199/293 (67.9%) in the Ljubljana and 102/132 (61.1%) in the Bristol validation cohorts, respectively. The algorithm, which is presented as a web application, incorporates nine variables including age, a localizing feature on electrocardiogram (ECG) (≥2 mm of ST change in contiguous leads), regional wall motion abnormality, history of vascular disease and initial shockable rhythm. This model had an area under the curve (AUC) of 0.89 in the development and 0.83/0.81 in the validation cohorts with good calibration and outperforms the current gold standard-ECG alone (AUC: 0.69/0.67/0/67). CONCLUSIONS: A novel simple machine learning-derived algorithm can be applied to patients with OHCA, to predict a culprit coronary artery disease lesion with high accuracy.

Outcome of inter-hospital transfer versus direct admission for primary percutaneous coronary intervention: An observational study of 25,315 patients with ST-elevation myocardial infarction from the London Heart Attack Group.

BACKGROUND AND AIMS: In patients with ST-segment elevation myocardial infarction (STEMI), mortality is directly related to time to reperfusion with guidelines recommending patients be delivered directly to centres for primary percutaneous coronary intervention (PCI). The aim of this study was to describe the impact of inter-hospital transfer on reperfusion time and to assess whether or not treatment delays influenced clinical outcomes in comparison with direct admission to a primary PCI centre in a large regional network. METHOD AND RESULTS: We undertook an observational cohort study of patients with STEMI treated with primary PCI between 2005 and 2015 in London, UK. Patient details were recorded at the time of the procedure in databases using the British Cardiovascular Intervention Society PCI dataset. The primary end-point was all-cause mortality at a median of 4.1 years (interquartile range: 2.2-5.8 years). Secondary outcomes were in-hospital major adverse cardiac events. Of 25,315 patients, 17,560 (69.4%) were admitted directly to a primary PCI centre and 7755 (31.6%) were transferred from a non-primary PCI centre. Patients in the direct admission group were older and more likely to have left ventricular impairment compared with the inter-hospital transfer group. Median time from call for help to reperfusion in transferred patients was 52 minutes longer compared with patients admitted directly (p <0.001). However, call to first hospital admission was similar. Kaplan-Meier analysis demonstrated significantly lower mortality rates in patients who were transferred directed to a primary PCI centre compared with patients who were transferred from a non-PCI centre (17.4% direct vs. 18.7% transfer, p=0.017). Furthermore, after propensity matching, direct admission for primary PCI was still a predictor of all-cause mortality (hazard ratio: 0.89, 95% confidence interval: 0.64-0.95). CONCLUSIONS: In this large registry of over 25,000 STEMI patients treated by primary PCI survival was better in patients admitted directly to a cardiac centre versus patients transferred for primary PCI, most likely due to longer call to balloon times in patient transferred from other hospitals.