The impact of external reference pricing on pharmaceutical costs and market dynamics.

Growth in the cost of prescription drugs in the US has generated significant interest in the use of external reference pricing (ERP) to tie prices paid for drugs to those in other countries. We used data from the Pricentric ONE™ database, an international drug pricing database, to examine product launch timing, launch price, and price changes from January 2010 - October 2021 in both ERP and non-ERP settings, with a focus on 100 high-priced drugs of interest to Medicare and Medicaid. We found that ERP policies were associated with a 73% reduction in the likelihood of drug launch within 9 months of regulatory approval relative to non-ERP settings. In addition, while ERP was associated with statistically significant reductions in annual drug price changes, such policies did not impact launch price. In addition, no single ERP feature (e.g., number of countries referenced, ERP calculation) was materially associated with the outcomes of interest. We conclude that ERP policies do not appear to impact drug launch price and may delay access to new therapies, raising questions about the utility of such policies in the US and potential consequences abroad.

Developing Criteria for Health Economic Quality Evaluation Tool.

OBJECTIVES: Because existing publication guidelines and checklists have limitations when used to assess the quality of cost-effectiveness analysis, we developed a novel quality assessment tool for cost-effectiveness analyses, differentiating methods and reporting quality and incorporating the relative importance of different quality attributes. METHODS: We defined 15 quality domains from a scoping review and identified 72 methods and reporting quality attributes (36 each). After designing a best-worst scaling survey, we fielded an online survey to researchers and practitioners to estimate the relative importance of the attributes in February 2021. We analyzed the survey data using a sequential conditional logit model. The final tool included 48 quality attributes deemed most important for assessing methods and reporting quality (24 each), accompanied by a free and web-based scoring system. RESULTS: A total of 524 participants completed the methodology section, and 372 completed both methodology and reporting sections. Quality attributes pertaining to the "modeling" and "data inputs and evidence synthesis" domains were deemed most important for methods quality, including "structure of the model reflects the underlying condition and intervention's impact" and "model validation is conducted." Quality attributes pertaining to "modeling" and "Intervention/comparator(s)" domains were considered most important for reporting quality, including "model descriptions are detailed enough for replication." Despite its growing prominence, "equity considerations" were not deemed as important as other quality attributes. CONCLUSIONS: The Criteria for Health Economic Quality Evaluation tool allows users to differentiate methods and reporting as well as quantifies the relative importance of quality attributes. Alongside other considerations, it could help assess and improve the quality of cost-effectiveness evidence to inform value-based decisions.

The Value of New: Consideration of Product Novelty in Health Technology Assessments of Pharmaceuticals.

BACKGROUND: Efforts to understand how treatments affect patients and society have broadened the criteria that health technology assessment (HTA) organizations apply to value assessments. We examined whether HTA agencies in eight countries consider treatment novelty in methods and deliberations. METHODS: We defined a novel pharmaceutical product to be one that offers a new approach to treatment (e.g., new mechanism of action), addresses an unmet need (e.g., targets a rare condition without effective treatments), or has a broader impact beyond what is typically measured in an HTA. We reviewed peer-reviewed publications and technical guidance materials from HTA organizations in Australia, Canada, England, France, The Netherlands, Norway, Sweden, and the United States (US). In addition, we explored how HTA organizations integrated novelty considerations into deliberations and recommendations related to two newer therapies-voretigene neparvovec for an inherited retinal disorder and ocrelizumab for multiple sclerosis. RESULTS: None of the HTA organizations acknowledge treatment novelty as an explicit value criterion in their assessments of pharmaceutical products. However, drugs that have novel characteristics are given special consideration, particularly when they address an unmet need. Several organizations document a willingness to expend more resources and accept greater evidence uncertainty for such treatments. Qualitative deliberations about the additional unquantified potential benefits of treatment may also influence HTA recommendations. CONCLUSION: Major HTA organizations do not recognize novelty as an explicit value criterion, although drugs with novel characteristics may receive special consideration. There is an opportunity for organizations to codify their approach to evaluating novelty in value assessment.

The effectiveness and value of aducanumab for Alzheimer's disease.

DISCLOSURES: Funding for this summary was contributed by Arnold Ventures, The Donaghue Foundation, Harvard Pilgrim Health Care, and Kaiser Foundation Health Plan to the Institute for Clinical and Economic Review (ICER), an independent organization that evaluates the evidence on the value of health care interventions. ICER's annual policy summit is supported by dues from AbbVie, America's Health Insurance Plans, Anthem, Alnylam, AstraZeneca, Biogen, Blue Shield of CA, Boehringer-Ingelheim, Cambia Health Services, CVS, Editas, Evolve Pharmacy, Express Scripts, Genentech/Roche, GlaxoSmithKline, Harvard Pilgrim, Health Care Service Corporation, HealthFirst, Health Partners, Humana, Johnson & Johnson (Janssen), Kaiser Permanente, LEO Pharma, Mallinckrodt, Merck, Novartis, National Pharmaceutical Council, Pfizer, Premera, Prime Therapeutics, Regeneron, Sanofi, Sun Life Financial, uniQure, and United Healthcare. Whittington, Rind, and Pearson are employed by ICER. Through their affiliated institutions, Synnott and Lin received funding from ICER for the work described in this summary. In addition, Synnott reports support from Biogen for the Tufts Medical Center Cost-Effectiveness Analysis Registry, which is maintained by the Center for the Evaluation of Value and Risk in Health.

Bridging the gap: aligning economic research with disease burden.

INTRODUCTION: Cost-effectiveness analysis (CEA) is critical for identifying high-value interventions that address significant unmet need. This study examines whether CEA study volume is proportionate to the burden associated with 21 major disease categories. METHODS: We searched the Tufts Medical Center CEA and Global Health CEA Registries for studies published between 2010 and 2019 that measured cost per quality-adjusted life-year or cost per disability-adjusted life-year (DALY). Stratified by geographical region and country income level, the relationship between literature volume and disease burden (as measured by 2019 Global Burden of Disease estimates of population DALYs) was analysed using ordinary least squares linear regression. Additionally, the number of CEAs per intervention deemed 'essential' for universal health coverage by the Disease Control Priorities Network was assessed to evaluate how many interventions are supported by cost-effectiveness evidence. RESULTS: The results located below the regression line but with relatively high burden suggested disease areas that were 'understudied' compared with expected study volume. Understudied disease areas varied by region. Higher-income and upper-middle-income country (HUMIC) CEA volume for non-communicable diseases (eg, mental/behavioural disorders) was 100-fold higher than that in low-income and lower-middle-income countries (LLMICs). LLMIC study volume remained concentrated in HIV/AIDS as well as other communicable and neglected tropical diseases. Across 60 essential interventions, only 33 had any supporting CEA evidence, and only 21 had a decision context involving a low-income or middle-income country. With the exception of one intervention, available CEA evidence revealed the 21 interventions to be cost-effective, with base-case findings less than three times the GDP per capita. CONCLUSION: Our analysis highlights disease areas that require significant policy attention. Research gaps for highly prevalent, lethal or disabling diseases, as well as essential interventions may be stifling potential efficiency gains. Large research disparities between HUMICs and LLMICs suggest funding opportunities for improving allocative efficiency in LLMIC health systems.

The Effectiveness and Value of Rivaroxaban and Icosapent Ethyl as Additive Therapies for Cardiovascular Disease.

DISCLOSURES: Funding for this summary was contributed by Arnold Ventures, Commonwealth Fund, California Health Care Foundation, National Institute for Health Care Management (NIHCM), New England States Consortium Systems Organization, Blue Cross Blue Shield of Massachusetts, Harvard Pilgrim Health Care, Kaiser Foundation Health Plan, and Partners HealthCare to the Institute for Clinical and Economic Review (ICER), an independent organization that evaluates the evidence on the value of health care interventions. ICER's annual policy summit is supported by dues from Aetna, America's Health Insurance Plans, Anthem, Allergan, Alnylam, AstraZeneca, Biogen, Blue Shield of CA, Cambia Health Services, CVS, Editas, Express Scripts, Genentech/Roche, GlaxoSmithKline, Harvard Pilgrim, Health Care Service Corporation, Health Partners, Johnson & Johnson (Janssen), Kaiser Permanente, LEO Pharma, Mallinckrodt, Merck, Novartis, National Pharmaceutical Council, Premera, Prime Therapeutics, Regeneron, Sanofi, Spark Therapeutics, and United Healthcare. Pearson is employed by ICER; Synnott was employed by ICER at the time of this report. Ollendorf, Campbell, and McQueen received grants from ICER for work on this review. Ollendorf also reports advisory board, consulting, and other fees from Sarepta Therapeutics, DBV Technologies, EMD Serono, Gerson Lehman Group, The CEA Registry Sponsors, Autolus, Analysis Group, Amgen, AbbVie, Cytokinetics, Aspen Institute/University of Southern California, and University of Colorado, unrelated to this review.

The Effectiveness and Value of Siponimod for Secondary Progressive Multiple Sclerosis.

DISCLOSURES: Funding for this summary was contributed by Arnold Ventures, Commonwealth Fund, California Health Care Foundation, National Institute for Health Care Management (NIHCM), New England States Consortium Systems Organization, Blue Cross Blue Shield of Massachusetts, Harvard Pilgrim Health Care, Kaiser Foundation Health Plan, and Partners HealthCare to the Institute for Clinical and Economic Review (ICER), an independent organization that evaluates the evidence on the value of health care interventions. ICER's annual policy summit is supported by dues from Aetna, America's Health Insurance Plans, Anthem, Allergan, Alnylam, AstraZeneca, Biogen, Blue Shield of CA, Cambia Health Services, CVS, Editas, Express Scripts, Genentech/Roche, GlaxoSmithKline, Harvard Pilgrim, Health Care Service Corporation, Health Partners, Johnson & Johnson (Janssen), Kaiser Permanente, LEO Pharma, Mallinckrodt, Merck, Novartis, National Pharmaceutical Council, Premera, Prime Therapeutics, Regeneron, Sanofi, Spark Therapeutics, and United Healthcare. Synnott and Pearson are employed by ICER. Bloudek and Carlson report a research agreement between the University of Washington and ICER; Bloudek reports consulting fees from Allergan, Seattle Genetics, Dermira, Sunovion, TerSera Therapeutics, Cook Regentech, and Mallinckrodt Pharmaceuticals; and Carlson reports personal fees from Bayer, unrelated to this report. Sharaf reports consulting fees from ICER.

The Effectiveness and Value of Biologic Therapies for the Treatment of Uncontrolled Asthma.

DISCLOSURES: Funding for this summary was contributed by the Laura and John Arnold Foundation, Blue Shield of California, and California Health Care Foundation to the Institute for Clinical and Economic Review (ICER), an independent organization that evaluates the evidence on the value of health care interventions. ICER's annual policy summit is supported by dues from Aetna, AHIP, Anthem, Blue Shield of California, CVS Caremark, Express Scripts, Harvard Pilgrim Health Care, Cambia Health Solutions, United Healthcare, Kaiser Permanente, Premera Blue Cross, AstraZeneca, Genentech, GlaxoSmithKline, Johnson & Johnson, Merck, National Pharmaceutical Council, Prime Therapeutics, Sanofi, Spark Therapeutics, Health Care Service Corporation, Editas, Alnylam, Regeneron, Mallinkrodt, Biogen, HealthPartners, and Novartis. Synnott, Kumar, Adair, Rind, and Pearson are employees of ICER, which provided grants to the University of California, San Francisco, and the University of Colorado to perform work for these analyses. Tice and Walsh are employed by the University of California, San Francisco, and Campbell and Whittington are employed by the University of Colorado.

Impact of Abuse Deterrent Formulations of Opioids in Patients With Chronic Pain in the United States: A Cost-Effectiveness Model.

OBJECTIVE: Opioid abuse is a significant public health problem in the United States. We evaluate the clinical effectiveness and economic impact of abuse-deterrent formulations (ADF) of opioids relative to non-ADF opioids in preventing abuse. METHODS: We developed a cost-effectiveness model simulating 2 cohorts of 100 000 noncancer, chronic-pain patients newly prescribed either ADF or non-ADF extended-release (ER) opioids and followed them over 5 years, tracking new events of opioid abuse and opioid-related overdose deaths in addition to tracking 5-year cumulative costs of therapeutic use and abuse of ADF and non-ADF opioids. Patients in each cohort entered the model for therapeutic opioid use from where they could continue in that pathway, discontinue opioid use, or abuse opioids or die of opioid overdose-related or unrelated causes. In addition, one-way sensitivity and scenario analysis were conducted. RESULTS: Over a 5-year time period, using ADF opioids prevented an additional 2300 new cases of opioid abuse at an additional cost of approximately $535 million to the healthcare sector. Threshold analyses showed that a 40% decrease in ADF opioid costs was required to attain cost neutrality between the 2 cohorts, whereas a 100% effectiveness in abuse reduction still did not result in cost neutrality. A 43% decrease in diversion with ADFs relative to non-ADFs was required to attain cost neutrality. Including a societal perspective produced results directionally similar to the base-case analysis findings. CONCLUSION: ADF opioids have the potential to prevent new cases of opioid abuse, but at substantially higher costs to the health system.

Cost Utility of Voretigene Neparvovec for Biallelic RPE65-Mediated Inherited Retinal Disease.

OBJECTIVE: The gene therapy voretigene neparvovec (VN) is the first Food and Drug Administration-approved treatment for vision loss owing to the ultra-rare RPE65-mediated inherited retinal disorders. We modeled the cost-utility of VN compared with standard of care (SoC). STUDY DESIGN: A 2-state Markov model, alive and dead, with a lifetime horizon. METHODS: Visual acuity (VA) and visual field (VF) were tracked to model quality-adjusted life-years (QALYs). VN led to an improvement in VA and VF that we assumed was maintained for 10 years followed by a 10-year waning period. The cost of VN was $850 000, and other direct medical costs for depression and trauma were included for a US healthcare system perspective. A modified societal perspective also included direct nonmedical costs and indirect costs. RESULTS: VN provided an additional 1.3 QALYs over the remaining lifetime of an individual. The average total lifetime direct medical cost for individuals treated with VN was $1 039 000 compared with $213 400 for SoC, leading to an incremental cost-effectiveness ratio (ICER) of $643 800/QALY from the US healthcare system perspective. Direct nonmedical costs totalled $1 070 900 for VN and $1 203 300 for SoC, and indirect costs totalled $405 400 for VN and $482 900 for SoC, leading to an ICER of $480 100/QALY from the modified societal perspective. CONCLUSIONS: At the current price, VN was unlikely to reach traditional cost-effectiveness standards compared with SoC. VN has important implications for both development and pricing of future gene therapies; therefore clinical and economic analyses must be carefully considered.

Assessing the Value of Sarilumab Monotherapy for Adults with Moderately to Severely Active Rheumatoid Arthritis: A Cost-Effectiveness Analysis.

BACKGROUND: Rheumatoid arthritis is associated with a societal burden greater than $39 billion annually. Novel treatments, known as targeted immune modulators (TIMs), are expensive but effective, producing improvements in response rates compared with conventional disease-modifying antirheumatic drugs (cDMARDs). Sarilumab, a TIM approved in 2017, shows superior improvements compared with cDMARDs and produced significantly greater likelihood of achieving response and improvement in the Health Assessment Questionnaire Disability Index than adalimumab monotherapy. Although sarilumab monotherapy has shown improvements over cDMARDs and the TIM market leader adalimumab, treatment with sarilumab is costly, with an annual wholesale acquisition cost of $39,000. OBJECTIVE: To estimate the lifetime cost-effectiveness of starting treatment with sarilumab monotherapy for adult patients with moderately to severely active rheumatoid arthritis who have had an inadequate response to cDMARDs. METHODS: A sequential treatment cohort model followed a hypothetical cohort from initiation of sarilumab monotherapy until death. The model allowed patients to switch therapies up to 3 times due to effectiveness or adverse events. The first switch was to a TIM within the same treatment category; the second switch was to a TIM within a different treatment category; and the third switch was to a cDMARD. Sarilumab monotherapy was compared with a cDMARD (methotrexate) and the TIM market leader (adalimumab monotherapy). Key risk and benefit evidence came from clinical studies and network meta-analyses of data on radiographic progression and response. We used a lifetime time horizon and the U.S. health sector payer perspective assuming therapy net pricing. We also incorporated loss of productivity to reflect a restricted societal perspective. RESULTS: Over a lifetime time horizon, a treatment pathway starting with sarilumab resulted in 17.16 life-years and 13.66 quality-adjusted life-years (QALYs). Treatment pathways starting with the cDMARD resulted in 16.54 life-years and 11.77 QALYs; treatment pathways starting with adalimumab resulted in 17.05 life-years and 13.35 QALYs. Total costs for sarilumab ($492,000 for payer perspective, $634,000 for societal perspective) were less than total costs for adalimumab ($536,000 for payer perspective, $689,000 for societal perspective) but higher than total costs for the cDMARD ($63,000 for payer perspective, $272,000 for societal perspective). When compared with cDMARD therapy, sarilumab resulted in a cost-effectiveness estimate of $227,000 per QALY gained from the payer perspective and $191,000 per QALYs gained from the societal perspective. When compared with adalimumab, sarilumab was dominant from both perspectives. CONCLUSIONS: Sarilumab resulted in better health outcomes than conventional therapy alone. However, its additional cost with assumed class-level net prices led to cost-effectiveness estimates above commonly cited thresholds. When compared with the market leader, sarilumab achieved favorable value. This evaluation informs stakeholders of the value of sarilumab and its alternatives to promote high value practices in health care. DISCLOSURES: Funding for this research was contributed by the Institute for Clinical and Economic Review (ICER). Ollendorf, Chapman, Kumar, Synnott, and Agboola are employees of ICER, an independent organization that evaluates the evidence on the value of health care interventions, which is funded by grants from the Laura and John Arnold Foundation, Blue Shield of California Foundation, and the California HealthCare Foundation. The organization's annual policy summit is supported by dues from Aetna, AHIP, Anthem, Blue Shield of California, CVS Caremark, Express Scripts, Harvard Pilgrim Health Care, Omeda Rx, United Healthcare, Kaiser Permanente, Premera Blue Cross, AstraZeneca, Genentech, GlaxoSmithKline, Johnson & Johnson, Merck, National Pharmaceutical Council, Takeda, Pfizer, Novartis, Lilly, and Humana. This work is an extension of an analysis presented at the New England Comparative Effectiveness Public Advisory Council on March 24, 2017, where the authors received public feedback on the analysis, results, and effect of a value assessment for targeted immune modulators. At the time of presentation, sarilumab was still an investigational product; therefore, a price was not known, so cost-effectiveness estimates were not generated. Since the presentation of that material, additional evidence for sarilumab has become available. The additional evidence has been incorporated into this analysis to present cost-effectiveness estimates for sarilumab.

The Authors Respond: Reframing the Value of Treatments for Relapsed Refractory Multiple Myeloma.

DISCLOSURES: Funding for the Carlson et al. study was provided in part by the Institute for Clinical and Economic Review. Ollendorf, Synnott, Chapman, and Pearson disclosed grants from Blue Shield of California Foundation, California Health Care Foundation, Laura and John Arnold Foundation, Aetna, AHIP, Anthem, Blue Shield of California, CVS Caremark, Express Scripts, Harvard Pilgrim Health Care, OmedaRx, United Healthcare, Kaiser Permanente, Premera, AstraZeneca, Genentech, GlaxoSmithKline, Johnson & Johnson, Merck, National Pharmaceutical Council, Takeda, Pfizer, Novartis, Lilly, Spark Therapeutics, Sanofi, Prime Therapeutics, and Health Care Service Corporation. Carlson disclosed grants from the Institute for Clinical and Economic Review and personal fees from Seattle Genetics, Genentech, and Pfizer. Russo, Guzauskas, Liu, and Brouwer have nothing to disclose.

Cost-effectiveness of Drugs to Treat Relapsed/Refractory Multiple Myeloma in the United States.

BACKGROUND: New 3-drug regimens have been developed and approved to treat multiple myeloma (MM). The absence of direct comparative data and the high cost of treatment support the need to assess the relative clinical and economic outcomes across all approved regimens. OBJECTIVE: To evaluate the cost-effectiveness of treatments for relapsed and/or refractory MM from a U.S. health system perspective. METHODS: We developed a partition survival model with 3 health states (progression-free, progression, and death) to evaluate the following regimens: carfilzomib (CFZ), elotuzumab (ELO), ixazomib (IX), daratumumab (DAR), and panobinostat (PAN) in combination with lenalidomide (LEN) or bortezomib (BOR) plus dexamethasone (DEX) in the second and/or third line of therapy. To estimate relative treatment effects, we developed a network meta-analysis and applied progression-free survival hazard ratios to baseline parametric progression-free survival functions derived from pooled data on LEN+DEX. We estimated overall survival using data on the relationship between progression-free survival and overall survival from a large meta-analysis of MM patients. Modeled costs included those related to drug treatment, administration, monitoring, adverse events, and progression. Utilities were from publicly available data and manufacturer data, if published sources were unavailable. RESULTS: Model results showed that regimens containing DAR yielded the highest expected life years (DAR range: 6.71-7.38 vs. non-DAR range: 3.25-5.27) and quality-adjusted life-years (QALY; DAR range: 4.38-5.44 vs. non-DAR range: 2.04-3.46), with DAR+BOR+DEX (second line) and PAN+BOR+DEX (third line) as the most cost-effective options (incremental cost-effectiveness ratio: $50,700 and cost saving, respectively). The applicability of the PAN+BOR+DEX result may be challenging, however, because of ongoing toxicity concerns. In the probabilistic sensitivity analysis, second-line DAR+BOR+DEX and third-line PAN+BOR+DEX had an 89% and 87% probability of being cost-effective at the $150,000 per QALY threshold, respectively. CONCLUSIONS: The introduction of newer drugs and regimens to treat second- and third-line relapsed/refractory MM appears to provide clinical benefits by lengthening progression-free and overall survival and improving quality of life. However, only the addition of DAR or PAN may be considered cost-effective options according to commonly cited thresholds, and PAN+BOR+DEX results require cautious interpretation. Achieving levels of value more closely aligned with patient benefit would require substantial discounts from the remaining agents evaluated. DISCLOSURES: Funding for this work was provided in part by the Institute for Clinical and Economic Review, which collaborated on the design, conduct, and reporting of this evaluation. During the conduct of this study, Ollendorf, Synnott, Chapman, and Pearson report grants from Blue Shield of California Foundation, California Health Care Foundation, and Laura and John Arnold Foundation and also report other grants from Aetna, AHIP, Anthem, Blue Shield of California, CVS Caremark, Express Scripts, Harvard Pilgrim Health Care, OmedaRx, United Healthcare, Kaiser Permanente, Premera, AstraZeneca, Genentech, GlaxoSmithKline, Johnson & Johnson, Merck, National Pharmaceutical Council, Takeda, Pfizer, Novartis, Lilly, Spark Therapeutics, Sanofi, Prime Therapeutics, and Health Care Service Corporation outside the submitted work. Carlson reports grants from the Institute for Clinical and Economic Review during the conduct of the study and personal fees from Seattle Genetics, Genentech, and Pfizer outside the submitted work. Russo, Guzauskas, Liu, and Brouwer have nothing to disclose. Study concept and design were contributed by Carlson, Guzauskas, and Ollendorf. Guzauskas, Chapman, Synnott, and Liu collected the data, and Carlson, Guzauskas, Chapman, and Ollendorf contributed to data analysis, along with Synnott and Liu. The manuscript was written by Carlson, Guzauskas, and Brouwer, along with Chapman, Synnott, and Ollendorf, and revised by Carlson, Brouwer, and Guzauskas, along with Chapman, Synnott, and Ollendorf.

Evaluating the Importance of Heterogeneity of Treatment Effect: Variation in Patient Utilities Can Influence Choice of the "Optimal" Oral Anticoagulant for Atrial Fibrillation.

OBJECTIVES: To investigate heterogeneity of treatment effect (HTE) for anticoagulants in atrial fibrillation across subgroups defined by 1) clinical characteristics and 2) variation in patient utilities for benefits and harms of treatment. METHODS: We reanalyzed aggregate data from a published network meta-analysis that compared four anticoagulants for atrial fibrillation (apixaban, dabigatran, edoxaban, and rivaroxaban) as well as warfarin. Event rates for stroke/systemic embolism (SE) and major bleeding were generated for each agent across seven subgroups, and rankings were developed on the basis of clinical performance. Utilities were derived from a national catalog and then applied to generate summary measures of benefit. The choice between any two agents was examined across a range of plausible utility values, defined as the interquartile range for stroke/SE and major bleeding. RESULTS: Little HTE was apparent in clinical and utility-adjusted analyses. Dabigatran 150 mg produced the lowest rates of stroke/SE, and edoxaban 30 mg had the lowest rate of major bleeding. Greater HTE was observed when utilities were varied across a plausible utility range. For example, among patients 75 years and older, dabigatran 150 mg would be preferred over edoxaban 30 mg when mean utility estimates are used. The preferred agent, however, would change at plausible utility thresholds of 0.6 and 0.7 for major bleeding and stroke/SE, respectively. Nearly 25% of all possible comparisons would see a change in preferred treatment within the plausible utility range. CONCLUSIONS: The optimal choice of anticoagulant in atrial fibrillation differs across subgroups defined by clinical characteristics and reasonable ranges of utilities.