TY - JOUR T1 - Scottish Keep Well health check programme: an interrupted time series analysis JF - Journal of Epidemiology and Community Health JO - J Epidemiol Community Health SP - 924 LP - 929 DO - 10.1136/jech-2015-206926 VL - 70 IS - 9 AU - Claudia Geue AU - James D Lewsey AU - Daniel F MacKay AU - Grace Antony AU - Colin M Fischbacher AU - Jill Muirie AU - Gerard McCartney Y1 - 2016/09/01 UR - http://jech.bmj.com/content/70/9/924.abstract N2 - Background Effective interventions are available to reduce cardiovascular risk. Recently, health check programmes have been implemented to target those at high risk of cardiovascular disease (CVD), but there is much debate whether these are likely to be effective at population level. This paper evaluates the impact of wave 1 of Keep Well, a Scottish health check programme, on cardiovascular outcomes.Methods Interrupted time series analyses were employed, comparing trends in outcomes in participating and non-participating practices before and after the introduction of health checks. Health outcomes are defined as CVD mortality, incident hospitalisations and prescribing of cardiovascular drugs.Results After accounting for secular trends and seasonal variation, coronary heart disease mortality and hospitalisations changed by 0.4% (95% CI −5.2% to 6.3%) and −1.1% (−3.4% to 1.3%) in Keep Well practices and by −0.3% (−2.7% to 2.2%) and −0.1% (−1.8% to 1.7%) in non-Keep Well practices, respectively, following the intervention. Adjusted changes in prescribing in Keep Well and non-Keep Well practices were 0.4% (−10.4% to 12.5%) and −1.5% (−9.4% to 7.2%) for statins; −2.5% (−12.3% to 8.4%) and −1.6% (−7.1% to 4.3%) for antihypertensive drugs; and −0.9% (−6.5% to 5.0%) and −2.4% (−10.1% to 6.0%) for antiplatelet drugs.Conclusions Any impact of the Keep Well health check intervention on CVD outcomes and prescribing in Scotland was very small. Findings do not support the use of the screening approach used by current health check programmes to address CVD. We used an interrupted time series method, but evaluation methods based on randomisation are feasible and preferable and would have allowed more reliable conclusions. These should be considered more often by policymakers at an early stage in programme design when there is uncertainty regarding programme effectiveness. ER -