Making better use of computerised clinical data for epidemiological research. (360G-Wellcome-082178_Z_07_Z)

The overall aims are to undertake two linked programmes of hypothesis driven research, one on inflammation and vascular disease and the other on drug effects. The two programmes share the aim of exploiting the enormous potential offered by the increasing computerisation of health records. Although distinct, the two themes are closely linked by common study designs and data sources, and by a focus on vascular disease. A thread of methodological development is also common to both themes. Studies will be based on anonymous computerised clinical data. New opportunities in data linkage will be exploited and innovative methods applied including the within person case- series and propensity score matched designs. Theme 1 - inflammation and acute vascular disease. Studies include examining the effects of a range of inflammatory stimuli on the risk of vascular events: a specific infectious agent (varicella zoster); dental procedures; and vaccination. Whether infection increases the risk of pre-eclampsia will also be established. Theme 2 - pharmacoepidemiology. Projects include a study of statins looking at a wide range of outcomes, an assessment of the effects of drug exposure in pregnancy on perinatal outcomes, and a case-only approach to assess the risk of stroke associated with antipsychotic drugs.