A project, recently approved by the EU, aims to exploit data from electronic healthcare records (EHRs) and biomedical databases to foster the early detection of adverse drug reactions (ADRs).
The scheme follows a number of high profile cases where adverse drug reactions have been reported in patients too late, when millions of patients had already been exposed. The project, called ALERT, aims to develop a computerised system, better and faster than spontaneous reporting. ALERT will use clinical data from the EHRs of over 30 million patients from the Netherlands, Denmark, the United Kingdom and Italy. Text mining, epidemiological and other techniques will be employed to analyse the EHRs to detect suspected adverse events and combinations of drugs that warrant further study. The project will emphasise the detection of ADRs in children, using paediatric data from all the countries represented, given the relative lack of knowledge of such events. ALERT also aims to discriminate between signals pointing to a genuine ADR, and spurious signals which can create unease in both patients and physicians and can result in the removal of a useful drug from the market. The project, also known by its full title “Early Detection of Adverse Drug Events by Integrative Mining of Clinical Records and Biomedical Knowledge” is funded by a