This study examines how patient-level utility data are analysed within NICE Technology Appraisals (TAs) and assesses the extent to which current practice aligns with methodological guidance identified in the published literature. Patient-level utility analysis is critical for estimating health-state utility values used in cost-utility modelling; however, no NICE Technical Support Document currently provides guidance on its appropriate implementation. Three complementary reviews were undertaken. A landscape review of the published literature identified recommended methods for analysing patient-level utility data. A targeted review was undertaken to supplement the findings of the landscape review. A review of recent NICE TAs assessed how utility data were collected and analysed, including the methods used, and issues raised by External Assessment Groups (EAGs) and NICE Committees. Twenty studies met inclusion criteria for the literature review, providing recommendations on core analytical methods but limited guidance on addressing complex data features. Targeted literature searches identified 32 studies, and two further seminal papers were added through expert knowledge. They provided recommendations on approaches to handling repeated measures, missing data, model specification and covariate selection. The TA review included 125 appraisals. Although mixed-effects models were the most common analytical approach, substantial heterogeneity in methods remained. While some best practices have been adopted in NICE submissions, variation and limited transparency persist in the analysis of patient-level utility data. The three methodological challenges most frequently identified as issues were capturing sub-response treatment benefits, modelling the long-term evolution of utility beyond progression or treatment discontinuation and capturing acute disease or treatment features. They are also among the areas least well covered by existing methodological guidance. The development of best-practice guidance for patient-level utility analysis should therefore prioritise these three areas to support more consistent and robust utility estimation for economic evaluation.
Genetic tests predict an individual’s risk of developing neurodegenerative diseases. A portion of the value provided by genetic tests can be attributed to the individual’s value of early knowledge of their risk, or the “value of knowing.” We developed a discrete choice experiment survey and administered it to a nationally representative sample of US adults. Respondents were presented with 8 choice tasks, each with two hypothetical tests and an opt-out option, where tests varied in cost, type, and accuracy. Respondents were randomly assigned to scenarios with varied disease severity, perceived risk, treatment availability, and whom the test was for. We used a generalized multinomial logit model to estimate preferences and willingness to pay (WTP). The final sample included 1,034 respondents. The alternative-specific constant implied a baseline model-derived WTP of 2,954 (95
Estimating and interpreting treatment effects (TE) for rare or delayed clinical outcomes is often challenging. To address this, researchers may incorporate additional evidence sources, including historical trial data and concurrent information from intermediate outcomes. In this article, we present Bayesian dynamic borrowing (BDB) as a principled framework for integrating such data while maintaining control of bias and Type I error. Using hypothetical trials of a novel high-efficacy therapy for multiple sclerosis, we provide a step-by-step demonstration of how BDB can be used to combine an imprecise TE estimate for a final outcome with a prediction derived from historical data and information on a concurrent intermediate outcome. Our illustration includes calibration of BDB to meet desired Type I error and power properties, and sensitivity analyses to assess robustness to assumption violations. We also discuss key considerations for applying BDB in regulatory decision making and health technology assessment contexts.
Providing informal care to a person with dementia can substantially affect carers’ quality of life. Existing generic preference-weighted measures (e.g. EQ-5D-5L) may not fully capture the caring experience, while carer-specific measures (e.g. CarerQol), although preference weighted, cannot be used to derive quality-adjusted life-years for cost-utility analysis. The COCOON project aimed to develop a new preference-weighted quality-of-life measure for carers of people living with dementia that is suitable for cost-utility analysis. Development of the descriptive system was informed by: (1) a literature review; (2) qualitative interviews with carers of people with dementia to assess face validity of candidate items; (3) an online survey of carers to evaluate psychometric performance (including missing data, distributions, known-group validity, convergent validity, exploratory factor analysis and item response theory analysis); and (4) consultation with an advisory group. An initial pool of 60 candidate items was reduced to 44 items for qualitative testing. Following interviews, the item set was refined to 35 items. Based on psychometric analyses and final advisory group consultation, the final descriptive system comprised seven items: emotional health, physical health, loss and grief, loss of freedom, loneliness, support and financial burden. Each item is measured using a five-point frequency scale from ‘never’ to ‘always’. The COCOON measure is grounded in the lived experiences of carers of people living with dementia, capturing the multidimensional facets of dementia caregiving and offers a conceptually robust and policy-relevant measure for assessing carers’ quality of life in economic evaluations.
Several quantitative approaches are available for synthesizing evidence from economic evaluations (EEs). This scoping review mapped the landscape of such studies, identified quantitative synthesis methods, and examined their frequency, settings, and study characteristics. Searches were conducted in PubMed, Embase, Epistemonikos, EconLit, and the DARE database from inception to October 2025. Eligible studies included those using quantitative methods to combine findings from multiple EEs. Studies were classified by synthesis method, with frequencies and distributions analyzed descriptively and summarized narratively. A total of 63 studies were included. Most were from high- and middle-income countries, covering diverse health areas, with pharmacological interventions representing the largest proportion. Eight quantitative synthesis methods were identified, with meta-analysis of incremental net benefit (INB) being the most commonly applied (n = 44; 69.8
Missing data are common in trial-based cost-effectiveness analyses (CEAs) and often lead to biased estimates. Earlier reviews suggested that the use of appropriate statistical approaches to handle the missing data in trial-based CEAs has been patchy. Since then, several methodological guidelines and tutorials in this area have been published, but it remains unclear whether these have improved practice. This paper provides a contemporary picture of missing data methods used in trial-based CEAs conducted in the UK and investigates the extent to which published guidance has permeated practice. We reviewed trial-based CEAs published in the Health Technology Assessment journal between 2022 and 2024, and 63 studies were identified. Missing data remains pervasive in trial-based CEAs; the median proportion of individuals with complete cost-effectiveness data was 61
Health economic models, many of which rely on risk equations, often play an important role in informing local healthcare decision making for the management of obesity. Risk equations for predicting cardiovascular outcomes have previously been derived from SELECT trial data. Validation of risk equations in real-world populations can provide confidence in the accuracy and reliability of health economic models. The aim of this study was to validate the predicted outcomes of the novel, SELECT trial-derived cardiovascular risk equations in a real-world UK population with overweight or obesity and established CVD without diabetes using the Clinical Practice Research Datalink (CPRD) database. The SELECT risk equations for acute coronary syndrome (ACS) and stroke were externally validated in a CPRD cohort; records were assessed between 24 October 2008 and 29 March 2021 for patients aged ≥ 45 years with body mass index (BMI) ≥ 27 kg/m2 and established CVD without diabetes. The risk equations were validated both in their original form, and after re-calibration to the CPRD analysis population. Discrimination and calibration were evaluated to assess the predictive performance of the risk equations. The SELECT risk equations, without recalibration, were also compared with published risk equations in terms of their discrimination and calibration. The discrimination C-indices at 4 years were 0.65 (95
Understanding the financial and health consequences if economic evaluation assumptions prove incorrect is essential for managing the risk associated with benefit package decisions, particularly in resource-constrained and overburdened healthcare systems. Yet these are also the settings that face the greatest challenges in conducting comprehensive uncertainty analysis, owing to a range of factors including limited skilled staff, data constraints, and short timelines to generate evidence in time to influence policy. This paper takes a pragmatic approach to support health technology assessment agencies in these settings to generate policy-relevant uncertainty analysis, drawing on good practice literature and the authors’ collective experience conducting economic evaluation for policy across resource-constrained settings. For each step of the economic evaluation process, we outline the main sources of uncertainty, principles for deciding which uncertainty analysis to prioritise, and approaches to overcome some of the common challenges faced when dealing with constrained timelines, data, and skilled staff. The overarching goal is to support better-informed decisions, by targeting uncertainty analysis to factors that actually affect decisions and by effectively communicating this decision-relevant uncertainty to policymakers.
Clinical trials have demonstrated the benefits of early detection of lung cancer (LC); however, the implementation of national LC screening programs remains under consideration. Economic evaluations provide evidence-based insights to support informed policy decisions. This systematic review aims to comprehensively synthesize the health economic evidence on LC screening and examine the methodological challenges. We searched six databases from inception to July 31, 2025. Articles were included if evaluating LC screening using a full economic evaluation and/or budget impact analysis (BIA). The Criteria for Health Economic Quality Evaluation (CHEQUE) tool was used for quality assessment of full economic evaluations. The International Society for Pharmacoeconomics and Outcomes Research (ISPOR) guidelines were modified as a quality checklist for BIA studies. Narrative synthesis was conducted to summarize the included studies. Ninety-five articles were identified, representing 93 full economic evaluations and nine BIAs. Seventy-nine full economic evaluations supported the cost-effectiveness of LC screening, while five indicated it was not cost-effective. Seventy studies evaluated low-dose computed tomography (LDCT) screening compared with no screening. Twelve studies considered emerging technologies either as standalone screening modalities or as adjuncts to LDCT screening, including artificial intelligence, genomics, and biomarkers. All BIAs assessed an LDCT-based screening program. Eight studies found that it would increase the budget, while one reported the opposite. All included studies demonstrated substantial methodological variation, including differences in the definition of eligible populations, modeling approaches, assumed uptake rates, cost components, sources of utility values, and equity consideration. Our review suggests that LDCT screening is generally cost-effective in high-risk populations, although implementing LC screening programs may entail a substantial budgetary impact. Emerging technologies, novel therapies, and combined screening have the potential to improve screening efficiency. Future economic evaluations are needed to assess these advances, while addressing existing methodological challenges to improve alignment between evidence. This review provides up-to-date economic evidence on LC screening and offers policymakers a clearer understanding of its financial implications.
Several user-written Stata codes exist for trial-based economic evaluations, but they lack assessment and guidance. This study aimed to identify and compare publicly available user-written Stata codes for trial-based economic evaluations. A focused literature search of Ovid Medline, SSC Archive, The Stata Journal and Google Scholar was conducted to identify relevant codes to June 2025. Codes were applied to data from two clinical trials, both featuring missing data and covariate adjustment. Codes were compared in terms of their ability to estimate key economic parameters and produce graphical outputs and functionality in handling four common statistical challenges: correlated costs and effects, covariate adjustment, skewed costs and effects, and missing data. We identified eight codes reported in four publications: codes for assessing health economic agreement (Gallacher et al.), sampling uncertainty for cost-effectiveness analysis (Glick et al.) and codes addressing missing data (Mutubuki et al. and Faria et al.). Gallacher et al.’s codes reported lower incremental quality-adjusted life years (QALYs) and net monetary benefit than Glick’s et al.’s codes. Mutubuki et al. and Faria et al.’s codes produced comparable incremental costs and QALYs, though Faria et al. yielded wider confidence intervals in cost estimates. Differences in estimates across statistical approaches show that code choice can influence economic evaluation results. Some codes were better suited for generating basic economic outputs, whereas others provide more comprehensive analyses or address specific statistical challenges including missing data. However, no single code provided all key outputs while addressing the main statistical challenges.
OFF episodes in Parkinson’s disease (PD), where symptoms worsen despite symptomatic treatment, are associated with significant burden on patients, carers and healthcare providers. There is an unmet need for an effective, on-demand treatment for OFF episodes that provides fast-acting symptom relief. This study estimated the cost-effectiveness of inhaled levodopa (LD) compared with relevant alternatives from a UK National Health Service and Personal Social Services perspective. A Markov model was developed comparing inhaled LD with subcutaneous (SC) apomorphine, sublingual (SL) apomorphine, dispersible LD and no on-demand treatment (no-ODT) in adults with advanced PD treated with LD and carbidopa and experiencing OFF episodes. A lifetime time horizon was adopted. Within-trial and beyond-trial models were applied for initial treatment and subsequent therapy, respectively. Model structure comprised 12 health states: ten ‘Off’ states, an ‘On’ state and death. Clinical trials and a network meta-analysis informed clinical inputs, with natural history progression applied from year 3. UK national databases and published literature informed costs (2025) and utility data. A 3.5
Neurodevelopmental disorders are lifelong conditions with onset in early childhood that impair cognitive, social, behavioural and physical functioning. Their consequences extend beyond healthcare, shaping education, social and community services, and family life. Economic evaluations can guide payers by quantifying the value for money of interventions, but standard methods require adaptation for this context. This paper offers practical guidance for pharmacoeconomic evaluations of interventions for children with neurodevelopmental disorders, synthesising methodological literature and applied experience to articulate core challenges, recommend solutions and highlight emerging opportunities. Evaluations must address five inter-related issues: defining and justifying an analytic perspective that captures multi-sectoral costs and benefits; identifying and valuing direct, indirect and intangible costs across health, education and social domains; selecting outcomes sensitive to changes in child functioning and health-related quality of life that can be economically valued; incorporating spillover effects for caregivers and families; and modelling long-term developmental trajectories amid data scarcity and uncertainty. We recommend adopting a societal perspective; applying comprehensive context-appropriate costing; combining generic and condition-specific outcome measures; explicitly including caregiver and family impacts; integrating equity analyses; and using long time horizons supported by robust modelling such as microsimulation or agent-based approaches. Promising directions include capability-based frameworks and digital biomarkers to enable real-world measurement of function and quality of life. Tailored pharmacoeconomic methods are essential for credible policy-relevant evaluations of neurodevelopmental disorder interventions. Transparent reporting and alignment with international standards will improve comparability, support uptake by decision makers, and promote equitable resource allocation for children and families affected by neurodevelopmental disorders.
The National Institute for Health and Care Excellence (NICE) regularly issues optimised technology appraisal (TA) recommendations, whereby medicines are recommended for a subset of their licensed population. These decisions play a growing role in shaping patient access, yet the underlying drivers and implications of optimisation remain poorly understood. We conducted a mixed‑methods analysis of all optimised NICE TAs published in 2023 and 2024. Fifty‑six appraisals met our inclusion criteria for qualitative assessment of committee documentation, while 30 provided sufficient data to estimate the degree of patient access using the established M‑score method. Three case studies were developed to illustrate how optimisation rationales translate into real‑world implications for patients. Of 131 positive decisions in 2023–2024, 50
Ischemic strokes are one of the main causes for death and disability worldwide and pose a substantial economic burden on healthcare systems. Currently, endovascular treatment is considered one of the most effective therapeutic options for the acute treatment of ischemic strokes. Furthermore, endovascular treatment has increasingly become the focus of health economic evaluations, in which decision modeling is used for the estimation of long-term costs and effects of endovascular treatment. This systematic review aims to assess the methodological quality regarding decision modeling-based health economic evaluations of endovascular treatment for acute ischemic strokes. A systematic search was conducted from January 2008 to December 2024 in PubMed, Econ-Lit, CDSR (Cochrane Database of Systematic Reviews), DARE (Database of Abstracts of Reviews of Effectiveness), and NHS EED/HTA (UK National Health Service Economic Evaluation Database/Health Technology Assessment) in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (search date: January 2025). The review was registered at PROSPERO (international prospective register of systematic reviews, identifier: CRD420251148065). Articles were eligible if they included a cost-effectiveness analysis of endovascular treatment compared to standard care for acute ischemic strokes using discrete event simulation or Markov modeling, and if they were considered original research and were written in English or German. General study characteristics, information on model design, parameter inputs, uncertainty, and validation were extracted and synthesized in tables. Methodological quality assessment was supported using a framework by Ramos et al. Forty-nine articles were finally included. Health states were defined using a stroke disability rating scale (modified Rankin scale) in all articles, although individual modified Rankin scale levels were merged into broader categories in 15 articles. All articles used a short-term decision tree for treatment effectiveness up to 90 days after stroke, of which one article subsequently employed discrete event simulation and one article employed Markov microsimulation to model lifetime outcomes. The remaining 47 articles employed Markov cohort models capturing health state stability and deterioration for long-term results, of which 13 articles additionally modeled an intermediate recovery phase (up to 1 year after stroke), in which health state improvement was possible. Recurrences were modeled in 40 articles. While transition probabilities were primarily taken from high-quality evidence studies (e.g., randomized controlled trials for treatment effectiveness in 44 articles), costs and/or utilities were not country specific in 32 articles. Sensitivity analyses including probabilistic or deterministic/scenario analyses were conducted in almost all articles (45 and 47 articles, respectively). Some form of validation was evident in all articles, although internal validation was reported in only ten articles. Future modelers should incorporate recovery up to the first year after the stroke, consider recurrences in the models, and define health states by using individual modified Rankin scale levels. Furthermore, costs and utilities plugged into the models should be country specific, probabilistic and deterministic sensitivity and scenario analyses should be assessed for costs, utilities, and transition probabilities, and the model should be internally validated.
Multi-indication medicines challenge health technology assessment processes as health technology assessment relies on repeated indication-based assessments to determine cost-effective pricing and reimbursement. The reimbursement of programmed death-ligand 1 inhibitors is an international example of this. In 2025, Australia’s Pharmaceutical Benefits Advisory Committee adopted a novel broad cancer listing on the Pharmaceutical Benefits Schedule for nivolumab, ipilimumab and pembrolizumab. This agreement consolidates all advanced and metastatic cancer indications under a single broad listing per medicine, using a single weighted-average price and utilisation-based rebate caps. This article analyses the economic, policy and clinical practice implications of such an agreement. The broad listing agreement features several characteristics identified in the literature as desirable for a multi-indication reimbursement framework, including facilitating timeliness of access, potentially reducing the health technology assessment workload, mitigation of arbitrage risk and the partial preservation of indication-based pricing principles. However, the agreement also includes some contestable trade-offs and introduces unresolved challenges, such as weakening indication-level value signalling, constraining routine utilisation monitoring and distorting comparator pricing in future economic evaluations. Further, the agreement may alter evidence generation and shift decision-making responsibilities to other actors within the health system. We suggest that while a broad listing represents a pragmatic response to the growing prevalence of multi-indication medicines, it entails important trade-offs between access, transparency and value alignment. These design considerations are likely to be relevant for other jurisdictions considering similar reimbursement reforms.
Dose administration aids (DAAs) support medication taking, yet little is known about how pharmacists perceive DAA benefits and challenges. This study aimed to explore community pharmacists’ preferences for attributes of a DAA service and how these influence their decision to provide this service. A discrete choice experiment (DCE) was conducted to elicit pharmacists’ preferences. Participants completed 12 choice questions, each presenting three alternatives: two DAA services and no DAA service. The questions included seven attributes: patient type, patient control over their medicines, impact on patient medication adherence, staff time, patient co-payment, government reimbursement fee and presence of a cap on government funded DAA services per pharmacy. A total of 615 community pharmacists completed the DCE survey. All attributes, except patient type and co-payment, significantly influenced preferences. Pharmacists were more likely to provide a DAA service (compared with no DAA service) when it was associated with higher government funding (pharmacy owners: OR 1.054, 95
