A prospective controlled trial of computerized decision support for lipid management in primary care

Objectives. This study aimed to assess the uptake and effect in primary care of a computerized decision support system (DSS) for the management of hyperlipidaemia. Method. A prospective controlled trial was conducted in 25 practices covering a population of 150 000 in the city of Birmingham. The Primed system, a specialist developed, rule based DSS for general practice, was introduced prospectively after a 3-month baseline data collection. The main outcome measures were nine months' data on prescribing of lipid lowering agents; use of laboratory tests; and referrals to secondary care for the investigation of hyperlipidaemia. Results. System use was lower than expected. A shift was observed towards requests for appropriate follow-up of previously abnormal lipid results and a greater emphasis on full lipid profiles, in line with the DSS guidelines. Referrals showed a 55% decrease on those expected (NS). The prescribing evaluation revealed a large variation between practices, but no significant alteration following system use. Views of users favoured decision support as a concept, but criticised technical problems with the system. Conclusions. Greater integration of DSS software and practice based data handling systems is needed. The mode of data capture, and hence both the content and form of knowledge representation, in DSS must take greater account of the primary care consultation process if such systems are to be of use to practitioners.