Women's preferences for attributes of first-trimester miscarriage management: a stated preference discrete-choice experiment

Abstract

Objective: To elicit women's preferences for attributes of alternative management options for first-trimester miscarriage.

Methods: A stated preference discrete-choice experiment was conducted among 1198 women with a confirmed pregnancy of less than 13 weeks gestation, who had been diagnosed with either an incomplete miscarriage or missed miscarriage/early fetal demise and who had been recruited as part of a randomized controlled trial (miscarriage treatment [MIST] trial) comparing expectant, medical, and surgical miscarriage. Six attributes, each with three or four levels, were used in the statistical design. An orthogonal main effects design was generated (i.e., a design where the attributes are independent of each other) and the choice sets devised according to the principles of minimum overlap and level balance. A cost attribute was included to allow estimation of willingness to pay (WTP) values. Three different questionnaires were designed such that women were asked their preferences for attributes of the two management options they had not been allocated to in the trial.

Results: A total of 630 women completed the stated preference discrete-choice survey questionnaires: 189 out of 398 women (47.5%) allocated to expectant management, 223 out of 398 women (56.0%) allocated to medical management, and 218 out of 402 women (54.2%) allocated to surgical management. For each of the three discrete-choice survey questionnaires, women expressed a clear preference for decreased levels of all six attributes (time spent at the hospital receiving treatment, level of pain experienced, number of days of bleeding after treatment, time taken to return to normal activities after treatment, cost of treatment to women, and chance of complications requiring more time or readmission to hospital). For each of the three discrete-choice survey questionnaires, the highest valued attribute in terms of WTP was for a reduction in pain levels followed by time taken to return to normal activities after treatment. On aggregate, surgical management was valued more highly than expectant and medical management by women allocated to medical and expectant management, respectively, and medical management was valued more highly than expectant management by women allocated to surgical management. This held true regardless of the application of either hypothetical data for each attribute generated by the pretrial-designed discrete-choice experiment questionnaires or actual data for each attribute observed in the MIST trial.

Conclusions: The preference results generated by this study suggest that many women undergoing management of first-trimester miscarriage would value being offered alternatives to expectant management. The data from this study should be considered by decision-makers in conjunction with the clinical and cost-effectiveness evidence base in this area as well as consideration of the budgets available to them for such services.