Use of electronic patient data overview with alerts in primary care increases prescribing of lipid-lowering medications in patients with type 2 diabetes

Abstract

Abstract Aims/hypothesis We aimed to assess whether general practices (GPs) using an electronic disease management program (DMP) with population overviews, including alerts when patients failed to receive guideline-recommended prescription medications, increased prescriptions of lipid-lowering drugs for patients with type 2 diabetes with no history of lipid-lowering treatment. Methods This observational study included 165 GPs that reached a high level of use of the DMP in 2012 and a control group of 135 GPs who reached a high level of use in 2013 and, hence, who were less exposed to the DMP throughout 2012. A binary measure for having been prescribed and filled lipid-lowering drugs at any time within a 12-month exposure period was derived for all patients with type 2 diabetes who did not receive a prescription for lipid-lowering drugs in the baseline year prior to the study period (i.e. 2011). Results were derived using ORs from multivariate logistic regression analyses. Subgroup stratification based on age, sex, diabetes duration, deprivation status and Charlson Comorbidity Index (CCI) score was conducted and assessed. Placebo tests were carried out to assess bias from selection to treatment. Results Patients who did not receive a prescription of lipid-lowering drugs in the year prior to being listed with GPs that used the DMP had statistically significant greater odds of receiving a prescription of lipid-lowering medications when compared with individuals who attended control GPs (OR 1.23 [95% CI 1.09, 1.38]). When the analysis period was shifted back by 2 years, no significant differences in lipid-lowering drug prescription between the two groups were found to occur, which indicates that these results were not driven by selection bias. Subgroup analyses showed that the increase in lipid-lowering drug prescriptions was primarily driven by changes among male participants (OR 1.32 [95% CI 1.12, 1.54]), patients aged 60–70 years (OR 1.40 [95% CI 1.13, 1.74]), patients with a diabetes duration of ≤5 years (OR 1.33 [95% CI 1.13, 1.56]), non-deprived patients (OR 1.25 [95% CI 1.08, 1.45]) and patients without comorbidities (CCI score = 0; OR 1.27 [95% CI 1.11, 1.45]). Conclusions/interpretation Access to population overviews using a DMP with alerts of clinical performance measures with regard to adhering to guideline-recommended prescription of medications can increase GP prescriptions of lipid-lowering drugs. Graphical abstract