General Population vs. Patient Preferences in Anticoagulant Therapy: A Discrete Choice Experiment

Development of a disease-specific health utility score for chronic obstructive pulmonary disease from a discrete choice experiment patient preference study

OBJECTIVES

While patient input to health technology assessment (HTA) has traditionally been of a qualitative nature, there is increasing interest to integrate quantitative evidence from patient preference studies into HTA decision making. Preference data can be used to generate disease-specific health utility data. We generated a health utility score for patients with chronic obstructive pulmonary disease (COPD) and consider its use within HTAs.

METHODS

Based on qualitative research, six symptoms were identified as important to COPD patients: shortness of breath, exacerbations, chronic cough, mucus secretion, sleep disturbance, and urinary incontinence. We employed a discrete choice experiment (DCE) and the random parameter logistic regression technique to estimate utility scores for all COPD health states. The relationship between patients' COPD health utility scores, self-perceived COPD severity, and EQ-5D-3L utility scores was analyzed, with data stratified according to disease severity and comorbidity subgroups.

RESULTS

The COPD health utility score had face validity, with utility scores negatively correlated with patients' self-perceived COPD severity. The correlation between the COPD health utility scores and EQ-5D-3L values was only moderate. While patient EQ-5D-3L scores were impacted by comorbidities, the COPD health utility score was less impacted by comorbid conditions.

CONCLUSIONS

Our COPD utility measure, derived from a DCE, provides a patient-centered health utility score and is more sensitive to the COPD health of the individual and less sensitive to other comorbidities. This disease-specific instrument should be considered alongside generic health-related quality of life instruments when valuing new COPD therapies in submissions to licensing and reimbursement agencies.

Do preferences differ based on respondent experience of a health issue and its treatment? A case study using a public health intervention

OBJECTIVES

Preference information is increasingly being elicited to support decision-making. Although discrete choice experiments (DCEs) are commonly used, little is known about how respondents' relative experience of a health issue, and its treatment, might impact the results of preference studies. The aim of this study was to explore how preferences differ between groups of individuals with varying levels of experience of a health issue and its treatment, using a weight loss maintenance (WLM) programme as a case study.

METHODS

An online DCE survey was provided to four groups, each differing in their level of experience with weight loss and WLM programmes. One group was recruited from a randomised controlled trial of a WLM programme (ISRCTN14657176) and the other three from an online panel. Choice data were analysed using mixed logit models. Relative attribute importance scores and willingness-to-pay (WTP) estimates were estimated to enable comparisons between groups.

RESULTS

Preferences differed between the groups across different attributes. The largest differences related to the outcome (weight re-gain) and cost attributes, resulting in WTP estimates that were statistically significantly different. The most experienced group was willing to pay £0.35 (95% CI: £0.28, £0.42) to avoid a percentage point increase in weight re-gain, compared with £0.12 (95% CI: £0.08, £0.16) for the least experienced group.

CONCLUSION

This study provides evidence in a public health setting to suggest that preferences differ based on respondent experience of the health issue and its treatment. Health preference researchers should therefore carefully consider the appropriate composition of their study samples.

Method for Calculating the Simultaneous Maximum Acceptable Risk Threshold (SMART) from Discrete-Choice Experiment Benefit-Risk Studies

BACKGROUND

Medical decisions require weighing expected benefits of treatment against multiple adverse outcomes under uncertainty (i.e., risks) that must be accepted as a bundle. However, conventional maximum acceptable risk (MAR) estimates derived from discrete-choice experiment benefit-risk studies evaluate the acceptance of individual risks, assuming other risks are fixed, potentially leading decision makers to misinterpret levels of risk acceptance.

DESIGN

Using simulations and a published discrete-choice experiment, we demonstrate a method for identifying multidimensional risk-tolerance measures given a treatment level of benefit.

RESULTS

Simultaneous Maximum Acceptable Risk Thresholds (SMART) represents combinations of risks that would be jointly accepted in exchange for specific treatment benefits. The framework shows how the expectation of utility associated with treatments that involve multiple risks are related even when preferences for potential adverse events are independent. We find that the form of the marginal effects of adverse-event probabilities on the expected utility of treatment determines the magnitude of differences between SMART and conventional single-outcome MAR estimates.

LIMITATIONS

Preferences for potential adverse events not considered in a study or preferences for adverse-event attributes held constant in risk-tolerance calculations may affect estimated risk tolerance. Further research is needed to understand the right balance between realistically reflecting clinical treatments with many potential adverse events and the cognitive burden of evaluating risk-risk tradeoffs in research and in practice.

CONCLUSIONS AND IMPLICATIONS

SMART analysis should be considered in preference studies evaluating the joint acceptance of multiple potential adverse events.

HIGHLIGHTS

Conventional approaches to calculate maximum-acceptable risk (MAR) using discrete-choice experiment data account for 1 adverse-event risk at a time, requiring that decision makers infer the acceptability of treatments when patients are exposed to multiple risks simultaneously.The Simultaneous Maximum Acceptable Risk Threshold (SMART) maps combinations of adverse-event risks that would be jointly acceptable given a specific treatment benefit and provides a transparent and precise portrayal of acceptance of multiple risks.Risk levels that would be accepted using individual MAR estimates might not be acceptable when simultaneous risks are considered, especially when marginal expected disutility of risk is decreasing nonlinearly with risk probabilities.Preference researchers should calculate SMARTs in any discrete-choice study in which 2 or more adverse-event risks are presented, particularly if risk preferences are nonlinear.

Patient Preference Studies for Advanced Prostate Cancer Treatment Along the Medical Product Life Cycle: Systematic Literature Review

BACKGROUND

Patient preference studies can inform decision-making across all stages of the medical product life cycle (MPLC). The treatment landscape for advanced prostate cancer (APC) treatment has substantially changed in recent years. However, the most patient-relevant aspects of APC treatment remain unclear. This systematic review of patient preference studies in APC aimed to summarize the evidence on patient preferences and patient-relevant aspects of APC treatments, and to evaluate the potential contribution of existing studies to decision-making within the respective stages of the MPLC.

METHODS

We searched MEDLINE and EMBASE for studies evaluating patient preferences related to APC treatment up to October 2020. Two reviewers independently performed screening, data extraction and quality assessment in duplicate. We descriptively summarized the findings and analyzed the studies regarding their contribution within the MPLC using an analytical framework.

RESULTS

Seven quantitative preference studies were included. One study each was conducted in the marketing approval and the health technology assessment (HTA) and reimbursement stage, and five were conducted in the post-marketing stage of the MPLC. While almost all stated to inform clinical practice, the specific contributions to clinical decision-making remained unclear for almost all studies. Evaluated attributes related to benefits, harms, and other treatment-related aspects and their relative importance varied relevantly between studies. All studies were judged of high quality overall, but some methodological issues regarding sample selection and the definition of patient-relevant treatment attributes were identified.

CONCLUSION

The most patient-relevant aspects regarding the benefits and harms of APC treatment are not yet established, and it remains unclear which APC treatments are preferred by patients. Findings from this study highlight the importance of transparent reporting and discussion of study findings according to their aims and with respect to their stage within the MPLC. Future research may benefit from using the MPLC framework for analyzing or determining the aims and design of patient preference studies.

Patient Preferences of Low-Dose Aspirin for Cardiovascular Disease and Colorectal Cancer Prevention in Italy: A Latent Class Analysis

Background

Patients taking low-dose aspirin to prevent cardiovascular disease (CVD) may also benefit from a reduced risk of colorectal cancer (CRC).

Objective

The aim was to examine the preferences of people eligible for preventive treatment with low-dose aspirin and the trade-offs they are willing to make between CVD prevention, CRC prevention, and treatment risks.

Methods

A cross-sectional study using a discrete choice experiment (DCE) survey was conducted in Italy in 2019 to elicit preferences for three benefit attributes (prevention of ischemic stroke, myocardial infarction, and CRC) and four risk attributes (intracranial and gastrointestinal bleeding, peptic ulcer, and severe allergic reaction) associated with use of low-dose aspirin. Latent class logit models were used to evaluate variation in treatment preferences.

Results

The DCE survey was completed by 1005 participants eligible for use of low-dose aspirin. A four-class model had the best fit for the primary CVD prevention group (n = 491), and a three-class model had the best fit for the secondary CVD prevention group (n = 514). For the primary CVD prevention group, where classes differed on age, education level, type 2 diabetes, exercise, and low-dose aspirin use, the most important attributes were intracranial bleeding (two classes), myocardial infarction (one class), and CRC (one class). For the secondary CVD prevention group, where classes differed on various comorbidities, self-reported health, exercise, and CVD medication use, the most important attributes were intracranial bleeding (two classes), myocardial infarction (one class), and gastrointestinal bleeding (one class).

Conclusion

Patient preferences for the benefits and risks of low-dose aspirin differ significantly among people eligible for treatment as primary or secondary CVD prevention.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40271-021-00506-2.

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

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.