Dr Mark Joy
After studying pure Mathematics to post-graduate level at Warwick, Mark taught as a secondary teacher in London for 5 years. He then re-trained in Information Systems Engineering and began research in Neural Networks; specifically, studying mathematically the stability of nonlinear circuits of Hopfield-type as a system of nonlinear ordinary differential equations. He also began lecturing Mathematics and Statistics at Kingston University, London.
Mark became interested in the Statistics in HealthCare and began studying predictive models for Occupancy at the hospital-level and also classical statistical and other predictive models for HSMRs (and other risks) in secondary healthcare.
Mark has extensive Commercial experience with big data sets in HealthCare, nationally and internationally, working as a Senior Statistician at Dr Foster (one of the leading providers of healthcare variation analysis and clinical benchmarking solutions worldwide). For the last four years he has been Statistician with the Global Comparators project, a major international hospital network created in 2011 to be a global hospital benchmarking collaborative.
Nonlinear Differential EquationsNeural NetworksMulti-Level Statistical Models in MedicinePredictive Modelling of HealthCare Systems
Background: In England, most prescribing of direct-acting oral anticoagulants for atrial fibrillation (AF) is in primary care. However, there remain gaps in our understanding of dosage and disparities in use. We aimed to describe trends in direct oral anticoagulant (DOAC) prescribing, including dose reduction in people with renal impairment and other criteria, and adherence. Methods: Using English primary care sentinel network data from 2014 to 2019, we assessed appropriate DOAC dose adjustment with creatinine clearance (CrCl). Our primary care sentinel cohort was a subset of 722 general practices, with 6.46 million currently registered patients at the time of this study. Results: Of 6 464 129 people in the cohort, 2.3% were aged ≥18 years with a diagnosis of AF, and 30.8% of these were prescribed vitamin K antagonist and 69.1% DOACs. Appropriate DOAC prescribing following CrCl measures improved between 2014 and 2019; dabigatran from 21.3% (95% CI 15.1% to 28.8%) to 48.7% (95% CI 45.0% to 52.4%); rivaroxaban from 22.1% (95% CI 16.7% to 28.4%) to 49.9% (95% CI 48.5% to 53.3%); edoxaban from 10.0% (95% CI 0.3% to 44.5%) in 2016 to 57.6% (95% CI 54.5% to 60.7%) in 2019; apixaban from 30.8% (95% CI 9.1% to 61.4%) in 2015 to 60.5% (95% CI 57.8% to 63.2%) in 2019. Adherence was highest for factor Xa inhibitors, increasing from 50.1% (95% CI 47.7% to 52.4%) in 2014 to 57.8% (95% CI 57.4% to 58.2%) in 2019. Asian and black/mixed ethnicity was associated with non-adherence (OR 1.81, 95% CI 1.56 to 2.09) as was male gender (OR 1.19, 95% CI 1.15 to 1.22), higher socioeconomic status (OR 1.60, 95% CI 1.52 to 1.68), being an ex-smoker (OR 1.12, 95% CI 1.06 to 1.19) and hypertension (OR 1.07, 95% CI 1.03 to 1.17). Conclusions: The volume and quality of DOAC prescribing has increased yearly. Future interventions to augment quality of anticoagulant management should target disparities in adherence.