Factors influencing women's preferences for subsequent management in the event of incomplete evacuation of the uterus after misoprostol treatment for miscarriage

STUDY QUESTION What affects women's treatment preferences in the management of an incomplete evacuation of the uterus after misoprostol treatment for a first-trimester miscarriage? SUMMARY ANSWER Women's treatment preferences in the management of an incomplete evacuation of the uterus after misoprostol treatment for miscarriage are most strongly influenced by 'the risk of a reduced fertility' followed by 'the probability of success'. WHAT IS KNOWN ALREADY Available treatment options in miscarriage are surgical, medical or expectant management. Treatment with misoprostol leads to an incomplete evacuation of the uterus and additional surgical treatment in 20-50% of women. To our knowledge, women's preferences for subsequent treatment of an incomplete evacuation of the uterus after misoprostol treatment for miscarriage have not been studied yet. STUDY DESIGN, SIZE, DURATION Between April 2014 and January 2015, we conducted a prospective nationwide multicentre discrete-choice experiment (DCE). DCEs have become the most frequently applied approach for studying patient preferences in health care. In our DCE, which considerers five attributes, a target sample size was calculated including 20 patients per attribute for the main analysis. We intended to include 25% more patients, i.e. a total of 125 thus enabling us to assess heterogeneity of treatment choices. PARTICIPANTS/MATERIALS, SETTING, METHODS All women visiting the outpatient clinic with first-trimester miscarriage or incomplete miscarriage were invited to participate in the study. Women under 18 years of age, women who were unable to understand the Dutch questionnaire or women who already had received a treatment for the current miscarriage were excluded. Women's preferences were assessed using a DCE. A literature review, expert opinions and interviews with women from the general population were used to define relevant treatment characteristics. Five attributes were selected: (i) certainty about the duration of convalescence; (ii) number of days of bleeding after treatment; (iii) probability of success (empty uterus after treatment); (iv) risk of reduced fertility and (v) risk of complications requiring more time or readmission to hospital. Fourteen scenarios using these attributes were selected in the DCE. Each of these scenarios presented two treatment options, while treatment characteristics varied between the 14 scenarios. For each scenario, respondents were asked to choose the preferred treatment option. The importance of each attribute was analysed, and preference heterogeneity was investigated through latent-class analysis. MAIN RESULTS AND THE ROLE OF CHANCE One hundred and eighty-six women were included of whom 128 completed the DCE (69% response rate). The two attributes with the greatest effect on their preference were, probability of success and risk of reduced fertility. The latent-class analysis revealed two subgroups of patients with different preference patterns. Forty per cent of women were more influenced by treatment success and 59% were more influenced by risk. LIMITATIONS, REASONS FOR CAUTION Most women were highly educated and were of Dutch origin, which limits the generalizability of our findings. Women with lower education levels, other cultural backgrounds and/or different previous experiences may differ from our findings. WIDER IMPLICATIONS OF THE FINDINGS Patients preferences should be addressed when counselling patients with an incomplete miscarriage after misoprostol treatment. STUDY FUNDING/COMPETING INTEREST(S) This study was embedded in the MisoREST trial, and funded by ZonMw, a Dutch organization for Health Research and Development, project number 80-82310-97-12066. There were no conflicts of interests. TRIAL REGISTRATION NUMBER Dutch Trial Register NTR3310, http://www.trialregister.nl TRIAL REGISTRATION DATE 27 February 2012. DATE OF FIRST PATIENT'S ENROLMENT 12 June 2012.