Adherence to guidelines on cervical cancer screening in general practice: programme elements of successful implementation

Background: There is still only limited understanding of whether and why in terventions to facilitate the implementation of guidelines for improving pr imary care are successful. It is therefore important to look inside the 'bl ack box' of the intervention, to ascertain which elements work well or less well. Aim: To assess the associations of key elements of a nationwide multifacete d prevention programme with the successful implement tation of cervical scr eening guidelines in general practice. Design of study: A nationwide prospective cohort study. Setting. A random sample of one-third of all 4758 general practices in The Netherlands (n = 1586). Method: General practitioners (GPs) in The Netherlands were exposed to a tw o-and-a-half-year nationwide multifaceted prevention programme to improve t he adherence to national guidelines for cervical cancer screening. Adherenc e to guidelines at baseline and after the intervention and actual exposure to programme elements were assessed in the sample using self-administered q uestionnaires. Results. Both baseline and post-measurement questionnaires were returned by 988 practices (response rate = 62%). No mayor differences in baseline prac tice characteristics between study population, non-responders, and all Neth erlands practices were observed. After the intervention all practices impro ved markedly (P<0.001) in their incorporation of nine out of 10 guideline i ndicators or effective cervical screening into practice. The most important elements for successful implementation were: specific software modules (od ds ratios and 95% confidence intervals for all nine indicators ranged from OR = 1.85 [95% CI = 1.24-2.77] to OR = 10.2 [95% CI = 7.38-14.1]); two or m ore 'practice visits' by outreach visitors (ORs and 95% CIs for six indicat ors ranged from OR = 1.46 [95% CI = 1.01-2.12] to OR = 2.35 [95% CI = 1.63- 3.38]); and an educational programme for practice assistants (ORS and 95% C Is for four indicators ranged from OR = 1.37 [95% CI 1.00-1.92] to OR = 1.9 0 [95% CI = 125-2.88]). Conclusion; A multifaceted programme targeting GPs, including facilitating software modules, outreach visits, and educational sessions for PAs, contri butes to the successful implementation of national guidelines for cervical screening.