Patient Preferences for Hemophilia a Treatments: A Discrete Choice Experiment
Shawn X. Sun,Jing Zhao,Hongbo Yang,Eric Wu
DOI: https://doi.org/10.1182/blood-2020-134656
IF: 20.3
2020-01-01
Blood
Abstract:Background: Gene therapy is being investigated as a potential new treatment for patients with hemophilia A. This is designed to provide functional copies of the factor VIII (FVIII) gene via a 1-time infusion to facilitate sustained endogenous production of therapeutic FVIII levels. As part of the patient-focused development of new hemophilia treatments, gene therapy is expected to have a number of different characteristics from current hemophilia A treatments, which are likely to influence patient treatment preferences. Aims: To assess patients' preferences for different hemophilia A treatments, including emerging treatments such as gene therapy. Methods: Male patients aged ≥18 years with moderate or severe hemophilia A and no history of FVIII inhibitors, who had been receiving prophylaxis in the previous month, were recruited in the US from December 2019 to January 2020. Patients who provided informed consent to participate in the online survey were screened and those eligible took part in a discrete choice experiment to quantify the relative importance of different treatment attributes on patient treatment preferences. Patients were presented with 14 choice cards and were asked to select the preferred treatment profile from 2 treatments with varying attribute levels on each card. Attributes and their levels to describe hemophilia A treatments were selected on the basis of a targeted literature review and feedback from 15 patients interviewed by telephone. One attribute level for out-of-pocket cost (a cost not covered by medical insurance) was hypothetical (1-time cost of $3000 for 5 years). Attributes focused on efficacy, safety, and administration-related measures. A conditional logit regression model was used to analyze patient preferences. Odds ratios quantified the impact of specific attribute levels on patients' treatment choices. Estimates of relative attribute importance (RAI) quantified the influence of each attribute on patients' decision making. Results: The survey was completed by 95 patients, 20 (21.1%) with moderate and 75 (78.9%) with severe hemophilia A, who had received prophylactic treatment for a mean (standard deviation [SD]) 14.0 (10.2) years. Most patients (75/95 [78.9%]) were receiving recombinant or plasma-derived FVIII prophylaxis administered once per month, once every 2 weeks, once per week, twice per week, or every other day (2, 3, 4, 33, and 33 patients, respectively). Twenty patients (21.1%) received nonfactor prophylaxis once per month, once every 2 weeks, once per week, twice per week, and with regular use of concomitant FVIII (2, 8, 7, 2, and 1 patient, respectively). Mean (SD) patient age at the survey date was 36.0 (11.3) years. Administration-related measures were the most important class of attributes for patients when selecting treatments (71.5% RAI): route and frequency of administration (30.3% RAI), out-of-pocket cost (24.1% RAI), and place of administration, eg, home versus clinic (17.1% RAI; Figure 1A). The risk of serious side effects was the least important attribute (2.2% RAI), lacking a significant impact on patient treatment choice. The directionality of odds ratios (ORs) for all attribute levels aligned with expectations. Patients exhibited a clear preference for infusion treatments administered with low versus high frequency (OR 3.3; P<0.001) and for treatments administered at home versus the clinic (OR 2.0; P<0.001; Figure 1B). Conclusions: Frequency, route, and place of administration and out-of-pocket costs were the treatment attributes found to have the greatest impact on patient preferences for hemophilia A treatment. Study limitations include potential for selection bias associated with the use of an online panel-based survey. Exclusion of patients with FVIII inhibitors also reduced generalizability, possibly leading to the reduced influence of safety-related treatment attributes. In future studies, design of efficacy attribute levels may better reflect the observed variability in FVIII expression between patients and over time with gene therapy. Insights from this study improve our understanding of preferred treatment regimens for patients with hemophilia A, and could help physicians, stakeholders, and decision makers to evaluate patient acceptance of emerging treatments for hemophilia A, such as gene therapies. Disclosures Sun: Takeda Pharmaceutical Company Ltd.: Current Employment. Zhao:Analysis Group Inc.: Current Employment; Takeda Pharmaceutical Company Ltd: Research Funding. Yang:Analysis Group Inc.: Current Employment; Takeda Pharmaceutical Company Ltd: Research Funding. Wu:Analysis Group Inc.: Current Employment; Takeda Pharmaceutical Company Ltd: Research Funding.