Value in Health最新文献

Objectives: To systematically review empirical evidence on the prevalence and influence of conflicts of interest (COIs) among members and public speakers of US Food and Drug Administration (FDA) advisory committees.

Methods: Following PRISMA guidelines, we searched MEDLINE, PubMed, and Cochrane Library from inception to November 2024 for peer-reviewed studies reporting quantitative data on COIs in FDA advisory committees. Eligible studies examined prevalence, type, or impact of COIs among voting members or public speakers. Data extraction and quality assessment were conducted independently by two reviewers using the Joanna Briggs Institute checklist for cross-sectional studies.

Results: Eighteen studies published between 2006 and 2022 were included, covering committee activity from 1997 to 2022. COIs among voting members ranged from 15% to over 70% of meetings, while 25% of public speakers disclosed financial COIs, most commonly consulting fees, research funding, and stock ownership. Evidence linking member COIs to voting outcomes was mixed, with some studies finding no significant association. In contrast, public speakers with financial COIs were consistently more likely to deliver favorable testimony, with odds ratios ranging from 3.0 to 6.1. Several studies suggested a decline in member COI prevalence after the 2007 FDA Amendments Act, but no similar trend was observed among public speakers.

Conclusions: COIs remain prevalent in FDA advisory committees, particularly among public speakers, where they are strongly associated with favorable testimony. These findings underscore the need for stronger and more consistent COI disclosure and management policies that include both committee members and public speakers to protect decision-making integrity.

In late 2025, the White House announced new Most Favored Nation (MFN) pricing agreements for the glucagon-like peptide-1 receptor agonist (GLP-1RA) class, including three semaglutide products, establishing substantially lower prices for Medicare and Medicaid. Shortly after, the Centers for Medicare and Medicaid Services (CMS) released Maximum Fair Prices (MFPs) for selected drugs under IPAY 2027, revealing semaglutide prices that differed from the MFN prices and from earlier assumptions used in prior economic evaluations, including our prior paper. Using previously published forecasting methods, we updated our ten-year (2026-2035) Medicare spending estimates for semaglutide across all FDA-approved indications under both the newly announced MFP and MFN price structures. Incorporating revised 30-day MFPs for Ozempic, Rybelsus, and Wegovy, as well as patient cost-sharing assumptions and future generic entry, we now estimate Medicare savings of $463 million under base-case MFP conditions, with alternative uptake scenarios producing $328-$599 million in savings and up to $1.78 billion with loss-of-exclusivity assumptions. Using the lower MFN price of $245 per month and a $600 annual patient copay, estimated Medicare savings increase substantially to $1.76 billion, ranging from $1.03 to $2.50 billion across uptake scenarios and reaching $2.63 billion with generic entry. These findings highlight the significant fiscal impact of recent price negotiations and underscore uncertainties regarding the durability and future scope of MFN-based drug pricing arrangements.

Objectives: Patients' time spent receiving care incurs an opportunity cost, which ought to be considered when conducting an economic evaluation from a societal perspective. Instruments for capturing time-related costs are presently lacking, especially for children and young people (CYP). To address this gap, we developed and pre-tested the Children and Young People's Time-Use Questionnaire for use in Economic Evaluation (CYP-TUQEE), producing versions for direct completion by CYP aged 11-17 years, and proxy completion by parents/carers of CYP aged up to 10 years.

Methods: The CYP-TUQEE was developed using an iterative process involving scoping reviews, consultation with a Working Group of experts, and pre-testing through think aloud interviews with 20 CYP and nine parents/carers. This process aimed to produce a comprehensive, adaptable questionnaire that is not onerous to complete by CYP or parents/carers within the target age ranges.

Results: Participants engaged well with the think aloud process, and provided feedback to inform the development of a novel, standardised instrument to facilitate the collection and inclusion of time-use data for paediatric economic evaluations. Feedback indicates that the CYP-TUQEE is easy to complete, clear, and ready for additional validation.

Conclusions: The CYP-TUQEE addresses a prominent gap by providing an accessible tool for data collection, tailored to CYP. Inclusion of patient time costs can assist in decision-making and ensure prioritisation of interventions respectful of patients' time. Future research will involve additional testing of the CYP-TUQEE in a real-world setting for further validation and refinement, and elicitation of a 'unit cost' (value) for CYP's time.

Objectives: The study objectives were (a) to demonstrate the feasibility of constructing a stated-preference evidence base and its use to quantify patients' consensus risk tolerance for treatment efficacy and (b) to use the evidence base to inform a new, parsimonious choice experiment to test an hypothesis for which there is no evidence-base information.

Methods: Nine original datasets from 5 discrete-choice-experiment studies that included inflammatory bowel disease symptom-remission and serious-infection risk attributes were obtained, totaling 2,247 respondents and 25,017 choice questions. All 9 datasets were pooled and fused in a single scale-adjusted, random-parameters logit, latent-class model describing risk-tolerant and risk-averse class preferences plus a statistically uninformative class. We used a 7-dataset fusion model to predict maximum acceptable risk for 2 holdout datasets.

Results: Class-membership probabilities for the risk-tolerant, risk-averse, and statistically uninformative classes were 0.53, 0.35, and 0.12, respectively. Consensus maximum acceptable 1-year risks of serious infection for 1 month of symptom remission were 9.5% (8.5, 10.6) and 5.8% (4.5, 7.1) for the risk-tolerant and risk-averse preference classes, respectively. The 7-dataset fusion model performed well for combined IBD out-of-sample predictions but predicted disease-specific values less accurately.

Conclusions: Maturation of the stated-preference literature offers opportunities to treat multiple quantitative preference studies similar to how multiple clinical studies are evaluated to estimate consensus effect sizes. There is significant value in developing and utilizing stated-preference evidence bases to provide benefit-transfer values as well as to identify information gaps and inform efficient de novo study designs to close those gaps.

Objective: This systematic review assessed the scope, reporting quality, and methodological risk of bias of health economic evaluations (HEEs) of medical artificial intelligence (AI) technologies, alongside the technological maturity of the AI systems assessed.

Methods: Following PRISMA 2020 guidelines, six databases were searched through April 2025 for studies reporting economic outcomes of AI applications in healthcare. Reporting quality was evaluated using the CHEERS-AI checklist, methodological risk of bias using the ECOBIAS framework, and AI maturity using the Clinical Machine Learning Readiness Level (CMLRL; 1-9). Inclusion of implementation and operational costs was examined, as well as their association with AI maturity.

Results: A total of 117 studies were included, with most published after 2021. Reporting quality was generally suboptimal, and ECOBIAS assessments highlight recurring risks of bias, particularly regarding incomplete cost inclusion, limited data transparency, inadequate uncertainty analysis, and insufficient model validation. Most studies evaluated AI tools at early development stages (63% at CMLRL 4-5), with limited real-world validation. While the majority of studies reported cost savings or cost-effectiveness, key cost categories were frequently omitted: only 28% included implementation costs and 57% reported operational costs.

Conclusions: Despite frequent claims of economic benefit, current HEEs of medical AI are constrained by limited reporting quality, risk of bias, and evaluations of immature technologies. Future HEEs should explicitly report technological maturity, incorporate full cost components, and employ rigorous methods. Robust evaluations conducted at higher readiness levels are also needed to generate reliable evidence for policy-making , reimbursement decisions, and responsible implementation.

Objectives: Health economic models for type 1 diabetes (T1D) typically require utilities or disutilities associated with diabetes-related complications. We conducted a systematic review of studies reporting utilities and disutilities associated with T1D-related complications, and assessed their methodological quality to identify a set of preferred disutilities.

Methods materials-methods: We searched six databases from inception to 30 April 2024. Data were extracted on study design, participant characteristics, complications, utility measurement methods, and reported values. Study quality was assessed based on sample size, population representativeness, appropriateness of the value sets, and statistical methods. Preferred disutilities for economic evaluations were selected from higher-quality studies.

Results: From 14,122 records identified, 25 were included for data extraction. Most studies identified complications via self-reporting (n = 12) or clinical assessment (n = 9). Of 22 studies analysing health utilities derived from MAUIs, only eight used value sets from the same countries as the study cohorts, and 14 did not report the value sets used. We derived disutilities for 66 complications/conditions. Fifteen studies used statistical models to estimate disutilities for 44 complications. Disutilities for several complications varied widely, e,g., stroke (-0.470 to -0.015), end-stage renal disease (-0.340 to -0.021), and diabetic neuropathy (-0.358 to -0.045). Quality assessment yielded preferred disutilities for 26 complications.

Conclusions: This review provides a comprehensive database of utilities and disutilities for T1D complications and a recommended set of disutilities for economic evaluations. Due to methodological and patient heterogeneity, these values should be used cautiously, with careful alignment between modelled health states and source study characteristics.

Background: Heterogeneous treatment effects (HTEs) refer to differences in how individual patients or subgroups respond to the same treatment. Estimating HTEs helps target care to those most likely to benefit, improving outcomes and avoiding unnecessary interventions. Machine learning (ML) enables the use of real-world data (RWD) to estimate HTEs when randomized controlled trials are not feasible. However, practical guidance for applying these methods in health economics is lacking.

Purpose: To support method selection, we identified and categorized ML approaches to estimating HTEs in RWD and assessed the methodological quality of studies applying them.

Methods: We conducted a scoping review following PRISMA-ScR guidelines. PubMed, Scopus, Web of Science, EBSCO, and MEDLINE were searched for studies published between 2014 and 2025 that applied ML to estimate HTEs from RWD. Methodological quality was assessed using a standardized checklist.

Findings: Of 1743 records screened, 74 met the inclusion criteria. We grouped the included studies into three categories: those using prediction-only approaches unsuited to HTE estimation (n=8), those applying outcome modelling (n=9), and those using customized conditional average treatment effect (CATE) estimation (n=58). Most innovations originated in the ML and statistics communities, with minimal uptake in health economics. Methodological quality was inconsistent and requires improvement.

Conclusion: ML methods for HTE estimation are increasingly applied to RWD. Tree-based models are most common, and interest in customized CATE approaches is growing. Better evaluation standards and more transparent reporting are needed for these methods to become reliable tools for health economics research.