Causality driven data integration for adverse drug reaction discovery

S : scientific #bd14 | big data conference 44 Dr Chen Wang Senior Research Scientist CSIRO Computational Informatics Dr Chen Wang is a Senior Research Scientist at CSIRO Computational Informatics. He received his PhD from Nanjing University. His research interests are primarily in distributed, parallel and trustworthy systems. His current work focus on data analytics systems for drug adverse reaction discovery. His recent work include accountable distributed systems and cloud computing. He is also an Honorary Associate of the School of Information Technologies at the University of Sydney. Dr Chen Wang has industrial experience. He developed a high-throughput event delivery system and a medical image archive system, which are used by many hospitals and medical centres in USA. Causality driven data integration for adverse drug reaction discovery Chen Wang, Sarvnaz Karmi CSIRO Computational Informatics SUMMARY We describe an ongoing effort in CSIRO for partially automating causality discovery in the Adverse Drug Reaction (ADR) detection process. The proposed method integrates data from multiple sources based on rules that indicate causality. INTRODUCTION Drug adverse reactions are a major threat to public health and impose huge costs to healthcare systems. Postmarketing surveillance aims to reduce the effects of adverse drug events. There are two types of ADR discovery systems operating around the globe: passive discovery and active discovery. They differ in the data used for detecting unexpected harm caused by the normal use of a drug at the normal dosage as per label or prescription. Passive ADR discovery, which has been established for decades, uses reports that are voluntarily submitted by pharmaceutical companies, healthcare professionals and consumers. Regulatory bodies, such as TGA (Therapeutic Goods Administration) and FDA (Food and Drug Administration), maintain very large databases of such reports which they use to mine the potential safety signals. More recently, active discovery has been introduced. Active discovery monitors healthcare data from a variety of sources such as electronic health records, health insurance claims, medical literature, or even recently medical forums to identify potential signals automatically using text and data mining techniques. Active discovery intends to discover unexpected adverse events as early as possible and is therefore also called near real-time drug safety surveillance. An example of such system is recently proposed in FDA Sentinel initiative which relies on sharing deidentified patient data among a number of organisations. Both types of ADR discoveries ultimately lead to establishing the causal relationship between a drug and unexpected adverse reactions. Often ADR discovery starts with data-mining techniques for disproportionality detection of the reports about a drug and an adverse reaction in comparison to other pairs of drugs and adverse reactions. These potential ADRs are then examined in medication safety review and assessment meetings. The main task of these meetings is to establish the causality between a drug and its adverse reactions. This is largely a manual process in the current practice and often generates wide variability in assessment. Even though the shortcomings of the current process were recognised in 70s, there is not much improvement in the practice of establishing causality between a drug and an ADR so far. As ADR related data become increasingly accessible in electronic format and with the increase in processing power and techniques of dealing with big data, it is now possible to introduce carefully designed algorithms to assist the causality reasoning process and therefore automate some of the manual steps in this assessment to reduce variability. We note that the current process, endorsed by WHO, is still largely based on Naranjo’s questionnaire designed in 1981. To achieve this, there are two major requirements: first, it is essential to understand and capture the reasoning process in the existing practice. A good reasoning process tends to minimise the variability and inconsistency in assessment as shown in. Second, integrating data from various sources is essential for reaching correct conclusions in the reasoning process, e.g. additional data about background of the patients in ADR reports may help to identify causes of an ADR. This is of course only possible with collaboration of multiple health agencies to make such data accessible. Below, we propose a causality detection method to address these requirements. DESCRIPTION Previous work trying to establish causality between a drug and its unexpected adverse reactions used a well designed questionnaire to guide the assessment process. The answers of these questions were assigned different scores and the total score of each rater determines the certainty of the rater on whether a drug D causes a reaction R. A consensus among raters served as an indicator of the causality of D and R. Our proposed method contains two steps: (1) Design rules to capture the causality reasoning process using domainexpertise and the current known knowledge of ADRs per each drug or active ingredient; and (2) Process different data sources based on these rules to establish if a given drug D causes a specified adverse reaction R. A starting point for rule identification is using the existing questionnaires, and also formalising the reasoning process within the review and assessment teams inside the regulatory. For instance, consider a specific drug D and its possible adverse reaction R and a given dataset S (e.g, electronic health records and clinical notes). The following rules could be considered for causality discovery: [email protected]