METHODS: A systematic search was conducted through Pubmed, CINAHL, EMBASE and Cochrane Central Register of Controlled Trials. Additional articles were located through cross-checking of the references list and bibliography citations of the included studies and previous review papers. We included intervention studies with controlled or baseline comparison groups that were conducted in primary care practices or the community, targeted at adult populations (randomized controlled trials, non-randomized trials with controlled groups and pre- and post-intervention studies). The interventions were targeted either at individuals, communities, health care professionals or the health-care system. The main outcome of interest was the relative risk (RR) of screening uptake rates due to the intervention.
RESULTS: We included 21 studies in the meta-analysis. The risk of bias for randomization was low to medium in the randomized controlled trials, except for one, and high in the non-randomized trials. Two analyses were performed; optimistic (using the highest effect sizes) and pessimistic (using the lowest effect sizes). Overall, interventions were shown to increase the uptake of screening for CVD risk factors (RR 1.443; 95% CI 1.264 to 1.648 for pessimistic analysis and RR 1.680; 95% CI 1.420 to 1.988 for optimistic analysis). Effective interventions that increased screening participation included: use of physician reminders (RR ranged between 1.392; 95% CI 1.192 to 1.625, and 1.471; 95% CI 1.304 to 1.660), use of dedicated personnel (RR ranged between 1.510; 95% CI 1.014 to 2.247, and 2.536; 95% CI 1.297 to 4.960) and provision of financial incentives for screening (RR 1.462; 95% CI 1.068 to 2.000). Meta-regression analysis showed that the effect of CVD risk factors screening uptake was not associated with study design, types of population nor types of interventions.
CONCLUSIONS: Interventions using physician reminders, using dedicated personnel to deliver screening, and provision of financial incentives were found to be effective in increasing CVD risk factors screening uptake.
METHODS: The EBMQ was developed based on a qualitative study, literature review and an expert panel. Face and content validity was verified by the expert panel and piloted among 10 participants. Primary care physicians with or without EBM training who could understand English were recruited from December 2015 to January 2016. The EBMQ was administered at baseline and two weeks later. A higher score indicates better knowledge, better practice of EBM and less barriers towards the implementation of EBM. We hypothesized that the EBMQ would have three domains: knowledge, practice and barriers.
RESULTS: The final version of the EBMQ consists of 80 items: 62 items were measured on a nominal scale, 22 items were measured on a 5 point Likert-scale. Flesch reading ease was 61.2. A total of 343 participants were approached; of whom 320 agreed to participate (response rate = 93.2%). Factor analysis revealed that the EBMQ had eight domains after 13 items were removed: "EBM websites", "evidence-based journals", "types of studies", "terms related to EBM", "practice", "access", "patient preferences" and "support". Cronbach alpha for the overall EBMQ was 0.909, whilst the Cronbach alpha for the individual domain ranged from 0.657-0.940. The EBMQ was able to discriminate between doctors with and without EBM training for 24 out of 42 items. At test-retest, kappa values ranged from 0.155 to 0.620.
CONCLUSIONS: The EBMQ was found to be a valid and reliable instrument to assess the knowledge, practice and barriers towards the implementation of EBM among primary care physicians in Malaysia.