Reporting Quality of Marginal Rates of Substitution in Discrete Choice Experiments That Elicit Patient Preferences

Abstract

Background: Discrete choice experiments (DCEs) are commonly used to elicit patient preferences as marginal rates of substitution (MRSs) between treatment or health service attributes. Because these studies are increasing in importance, it is vital that uncertainty around MRS estimates is reported.Objective: To review recently published DCE studies that elicit patient preferences in relation to MRS reporting and to explore the accuracy of using other reported information to estimate the uncertainty of the MRSs.Methods: A systematic literature review of DCEs conducted with patients between 2014 and July 2019 was performed. The number of studies reporting coefficients, MRSs, standard errors (SEs), and confidence intervals was recorded. If all information was reported, studies were included in an analysis to determine the impact of estimating the SEs of MRSs using coefficients and assuming zero covariance, to determine the impact of this assumption.Results: Two hundred and thirty-two patient DCEs were identified in the review; 34.1% (n = 79) reported 1 or more MRS and, of these, only 62.0% (n = 49) provided an estimate of the uncertainty. Of these studies, 16 contained enough information for inclusion in the analysis, providing 116 datapoints. Actual SEs were smaller than estimated SEs in 75.0% of cases (n = 87), and estimated SEs were within 25% of the actual SE in 59.5% of cases (n = 69). Conclusion:Uncertainty of MRS estimates is unreported in a substantial proportion of recently published DCE studies. Estimating the SE of a MRS by solely using the SEs of the utility coefficients is likely to lead to biased estimates of the precision of patient trade-offs.

A Systematic and Critical Review of Discrete Choice Experiments in Asthma and Chronic Obstructive Pulmonary Disease

“…The number of studies reporting a behavioural outcome such as MRS, relative attribute importance, or predicted uptake probability with a measure of uncertainty of the estimate were counted. Such behavioural insights that take into account attribute trade-offs are one of the key outputs of preference research, and the underreporting of measures of uncertainty may lead to inaccurate policy advice [ 15 ].…”

“…Unlike other survey techniques, DCEs estimate participants' preferences and the trade-offs they are willing to make based on responses to several choice questions. Although non-indication-specific reviews are useful for understanding how methodological and reporting standards have developed over time [14,15], they do not inform researchers of the current state of knowledge within specific indications. A previous review of patient preference studies in asthma and COPD was conducted in 2015 [16], but it did not specifically review the DCE methodology nor report on the consistency in reporting of key study criteria or attributes included in the studies.…”

Utility Values Associated with Atypical Hemolytic Uremic Syndrome-Related Attributes: A Discrete Choice Experiment in Five Countries

“…Therefore, any framing effects that influence the importance of the overall survival attribute could have a substantial impact on the estimated disutilities. In addition, the covariance between attributes was not accounted for in the calculation of the confidence intervals around the MRS and utility estimates, which may bias the precision of these estimates [43].…”

Response to “Letter to the Editor Regarding: Patient Preferences for Glucagon-like Peptide-1 (GLP-1) Receptor Agonist Treatment of Type 2 Diabetes Mellitus in Japan: A Discrete Choice Experiment”

“…By contrast, providing the uncertainty (95% confidence intervals) around the predicted choice probability for the semaglutide 0.50 mg profile versus the dulaglutide 0.75 mg profile would have strengthened our conclusions and might have alleviated the author's concern. Uncertainty around estimates from patient preference studies should be provided to help interpret results and inform patient-centered benefit-risk assessments [3]. The confidence interval around the predicted choice probability was small (78% [95% confidence interval, 74-82%]), supporting the conclusion that the large majority of participants preferred the semaglutide 0.50 mg profile.…”