Title: Pharmacovigilance through big data

Biography:

Bernard Cheung graduated from the University of Cambridge.  In 2007-2009, he held the chair in Clinical Pharmacology and Therapeutics in Birmingham.  He is now the Sun Chieh Yeh Heart Foundation Professor in Cardiovascular Therapeutics at the University of Hong Kong, and heads the Division of Clinical Pharmacology and Therapeutics.  He is an Honorary Consultant Physician of Queen Mary Hospital and the Medical Director of the Phase 1 Clinical Trials Centre.  He is an Honorary Professor at the Hong Kong University Shenzhen Hospital and a Visiting Professor of Shenzhen University. He is the Editor-in-Chief of Postgraduate Medical Journal.  He has 270 publications, 7500 citations and an h-index of 43.

Abstract

In Hong Kong 70% of inpatient care is in public hospitals and most acute emergencies are treated in public hospitals. Therefore the most serious adverse drug reactions present to public hospitals. The Clinial Data Analyais and Reporting System (CDARS) has been capturing prescribing and dispensing data, as well as clinic and hospital attendance data, in Hong Kong for over ten years. This authentic source of information can link clinical events, at least the serious ones, with dispensing data. There are three recent examples of how we used these data. Firstly, contaminiated batches of generic valsartan were recalled in 2018 and we monitored the daily admissions due to stroke in the whole of Hong Kong. Fortunately, there was not the slightest signal of an abrupt rise in stroke. Secondly, we studied drug-related hypokalaemia (low plasma potassium) in the population. Of around 25000 hospitalisations due to hypokalaemia over a 10-year period, around 5000 were due to diuretics, one third of which were due to indapamide used for the treatment of hypertension. Lastly, we studied the incidence and risk factors for vancomycin-induced acute kidney injury. 1450 cases were identified over a five-year period. We identified trough vancomycin level, baseline renal function, organ dysfunction and concomitant drugs as risk factors These studies demonstrate the power of using computerised population medical data to investigate uncommon adverse effects of drugs. These big data can be used to complement the spontaneous reports from clinicians