Domain - led time series analysis of cardiovascular disease using open data: does reduction in coronary disease increase heart failure prevalence?

Alicja Jasinska-Piadlo1, Pardis Biglarbeigi1, Raymond Bond1, David McEneaney2
1Ulster University, 2CVD Research Department, Southern Health and Social Care Trust


Abstract

Accessing healthcare data for research purposes is a lengthy process and requires data governance and ethical approval. There are however open data existing in the public domain that can be freely accessed and used for research purposes. Methods: We performed time series analysis of cardiovascular disease trends in Northern Ireland (NI) during years 2010-2020. We accessed the Raw Disease Prevalence Data for NI from the Quality Outcomes Framework (QOF) published on the government website. Statistical analysis was performed to explore trends in prevalence for patients with coronary artery disease, heart failure and atrial fibrillation. We also computed geodata and calculated the Euclidean distances between the geolocations of GP practices and primary percutaneous coronary intervention (PPCI) centers and correlated these distances with the mortality rates from heart disease from NI Statistical Research Agency website. Results & Conclusions: We observed a significant increase in the prevalence of patients with HF and AF in the NI population against a background of unchanged prevalence of coronary artery disease in each year since 2015 (Chi Squared test p-value <0.05). Our analysis shows that open data have useful research value when research questions are asked by clinical experts. We advocate for the use of publicly available data from government or other sources for research purposes and care quality improvement.