Implementing targeted expectant management in fertility care using prognostic modelling: A cluster randomized trial with a multifaceted strategy

STUDY QUESTION What is the effectiveness of a multifaceted implementation strategy compared to usual care on improving the adherence to guideline recommendations on expectant management for couples with unexplained infertility? SUMMARY ANSWER The multifaceted implementation strategy did not significantly increase adherence to guideline recommendations on expectant management compared to care as usual. WHAT IS KNOWN ALREADY Intrauterine insemination (IUI) with or without ovarian hyperstimulation has no beneficial effect compared to no treatment for 6 months after the fertility work-up for couples with unexplained infertility and a good prognosis of natural conception. Therefore, various professionals and policy makers have advocated the use of prognostic profiles and expectant management in guideline recommendations. STUDY DESIGN, SIZE, DURATION A cluster randomized controlled trial in 25 clinics in the Netherlands was conducted between March 2013 and May 2014. Clinics were randomized between the implementation strategy (intervention, n = 13) and care as usual (control, n = 12). The effect of the implementation strategy was evaluated by comparing baseline and effect measurement data. Data collection was retrospective and obtained from medical record research and a patient questionnaire. PARTICIPANTS/MATERIALS, SETTING, METHODS A total of 544 couples were included at baseline and 485 at the effect measurement (247 intervention group/238 control group). MAIN RESULTS AND THE ROLE OF CHANCE Guideline adherence increased from 49 to 69% (OR 2.66; 95% CI 1.45-4.89) in the intervention group, and from 49 to 61% (OR 2.03; 95% CI 1.38-3.00) in the control group. Multilevel analysis with case-mix adjustment showed that the difference of 8% was not statistically significant (OR 1.31; 95% CI 0.67-2.59). The ongoing pregnancy rate within six months after fertility work-up did not significantly differ between intervention and control group (25% versus 27%: OR 0.72; 95% CI 0.40-1.27). LIMITATIONS REASONS FOR CAUTION There is a possible selection bias, couples included in the study had a higher socio-economic status than non-responders. How this affects guideline adherence is unclear. Furthermore, when powering for this study we did not take into account the unexpected improvement of adherence in the control group. WIDER IMPLICATIONS OF THE FINDINGS Generalization of our results to other countries with recommendations on expectant management might be questionable because barriers for expectant management can be very different in other countries. Furthermore, due to a large variation in improved adherence rate in the intervention group it will be interesting to further analyse the process of implementation in each clinic with a process evaluation on professionals and couples' exposure to and experiences with the strategy. STUDY FUNDING/COMPETING INTEREST(S) Supported by Netherlands Organisation for Health Research and Development (ZonMW, project number 171203005). No competing interests. TRIAL REGISTRATION NUMBER Dutch trial Register, www.trialregister.nl NTR3405. TRIAL REGISTRATION DATE 19 April 2012. DATE OF FIRST PATIENT'S ENROLMENT 10 July 2012.