Optimising antiplatelet utilisation in the acute care setting: a novel threshold for medical intervention in suspected acute coronary syndromes.

OBJECTIVES: To construct a model to optimise and personalise recommendations for antiplatelet prescription for patients with suspected acute coronary syndrome (ACS). Acknowledging that emergency physicians work with diagnostic uncertainty, we sought to identify the point at which the probability of ACS is sufficiently high that the benefits of antiplatelet treatment outweigh the risks. Second, we evaluated the projected clinical impact of this approach by using a clinical prediction model (Troponin-only Manchester Acute Coronary Syndromes (T-MACS)) to calculate the probability of ACS.

METHODS: We conducted three systematic reviews, quantifying the effects of ticagrelor, clopidogrel or aspirin-alone treatment strategies for ACS (November 2017). We extracted data for (a) clinical outcomes and (b) weighted patient preferences (utilities) for each outcome. We then constructed utilitarian models, simulating the probability of clinical outcomes with different treatment strategies. This identified the threshold probability of ACS at which each treatment strategy became superior.We validated this approach in a prospective diagnostic study including patients with suspected ACS that was conducted at two large UK teaching hospitals (St George's Hospital London recruited October 2015 to June 2017 and Manchester Royal Infirmary: February 2015 to August 2017). We calculated the probability of ACS using T-MACS. The diagnosis of ACS was adjudicated based on serial high-sensitivity troponin testing and 30-day follow-up.

RESULTS: We constructed three models using data from six studies. Prescribing ticagrelor had greatest overall benefit when the probability of ACS exceeded 8.0%. Below that threshold, aspirin alone yielded greater benefit. The validation study included 660 patients, of which 87 (13.2%) had ACS. Prescription of combined antiplatelet strategy to patients with >8% probability of ACS had greater utility than aspirin alone.

CONCLUSION: Treatment with ticagrelor appears to yield greater net benefit for patients when the probability of ACS >8%. The clinical and cost-effectiveness of this 'precision medicine' approach warrants further study.