Towards the Integration of Prescription Analytics into Health Policy and General Practice Brian Cleland 1(&) , Jonathan Wallace 1 , Raymond Bond 1 , Michaela Black 1 , Maurice Mulvenna 1 , Deborah Rankin 1 , and Austin Tanney 2 1 Ulster University, Coleraine, Northern Ireland, UK {b.cleland,jg.wallace,rb.bond,mm.black, md.mulvenna}@ulster.ac.uk 2 Analytics Engines, Belfast, Northern Ireland, UK a.tanney@analyticsengines.com Abstract. The phenomenon of big data and data analytics is impacting many sectors, including healthcare. Practical examples of the application of big data to health policy and health service delivery remain scarce, however. In this paper, which summarises findings from an ongoing research project, we explore the potential for applying data analytics and anomaly detection to open data in order to support improved policy design and to enable better clinical decisions in primary care. The policy context of mental health in Northern Ireland is described, and its importance as a public health issue is explained. Based on previous work, it is proposed that depression prevalence is a mediating factor between economic deprivation and antidepressant prescribing. This hypothesis is tested by analysing a variety of open datasets. The methodology is described, including datasets used, the data processing pipeline, and analysis tools. The results are presented, identifying correlations between the three main variables, and highlighting anomalies in the data. The findings are discussed and impli- cations and opportunities for further research are described. Keywords: Health policy Á Data analytics Á Big data Á Prescribing Á Prevalence Á Deprivation 1 Introduction The increasing impact of data analytics on industry and society is often discussed under the banner of “big data”. Just as the growth of data analytics has impacted many economic sectors, so healthcare is being transformed by this phenomenon [1, 14, 29]. In 2016, the UK House of Commons noted that big data had “huge unrealised potential, both as a driver of productivity and as a way of offering better products and services to citizens” [12]. Nevertheless, with the exception of some public health surveillance [5, 31] and pharmacovigilance systems [33], exploration of the application of big data to health policy and service delivery remains limited. © Springer International Publishing AG 2017 M. Bramer and M. Petridis (Eds.): SGAI-AI 2017, LNAI 10630, pp. 193–206, 2017. https://doi.org/10.1007/978-3-319-71078-5_18