Interpreting cost-effectiveness ratios in a cost-effectiveness analysis of risk-tailored prostate screening: A critique of Callender et al

Callender et al. recently published a model-based cost-effectiveness analysis of a risk-tailored approach to prostate cancer screening. It considers the costs and effects of prostate cancer screening offered to all men aged 55-69 without any risk selection and, alternatively, over a range of risk-tailored strategies in which screen eligibility is determined by a varying threshold of disease risk. The analysis finds that the strategy of screening men once they reach a 10-year absolute risk of disease of 5% or more is cost-effective in a UK context. I believe there are several problems with the study, mostly stemming from an incorrect interpretation of the cost-effectiveness estimates. I show that one reinterpretation of their results indicates that screening is much less cost-effective than the original analysis suggests, indicating that screening should be restricted to a much smaller group of higher risk men. More broadly, I explain the challenges of attempting to meaningfully reinterpret the originally published results due to the simulation of non-mutually exclusive intervention strategies. Finally, I consider the relevance of considering sufficient alternative screening intensities. This critique highlights the need for appropriate interpretation of cost-effectiveness results for policymakers, especially as risk stratification within screening becomes increasingly feasible.

The Joint Committee on Vaccination and Immunisation's Advice on Extending Human Papillomavirus Vaccination to Boys: Were Cost-Effectiveness Analysis Guidelines Bent to Achieve a Politically Acceptable Decision?

In July 2018, the UK Minister of Public Health announced that human papillomavirus vaccination would be extended to 12-year-old boys. This decision was informed by updated evidence from the Joint Committee on Vaccination and Immunisation (JCVI) published earlier that month. Vaccination of boys had been found not to be cost-effective in a series of analyses conducted for the JCVI, including the most recent assessment prior to the minister's announcement. These analyses were conducted under the standard methods for cost-effectiveness analysis recommended by the JCVI, which are primarily based on guidelines from the National Institute of Health and Care Excellence. Although the JCVI concluded they were unable to advise extending vaccination on the basis of standard appraisal methods, their most recent round of assessment also considered analyses using nonstandard appraisal methods. In particular, the JCVI noted that vaccination of boys was likely to be cost-effective when a lower discount rate of 1.5% is applied to costs and health effects, as opposed to the 3.5% rate usually employed. The JCVI stated that they were supportive of applying such alternative methods, and on this basis, they would advise extending vaccination to boys. This commentary explains the JCVI's application of nonstandard appraisal methods and considers whether it was justified. We conclude that the JCVI was not justified in applying the lower discount rate. We voice concerns that a willingness to endorse a politically popular intervention may have driven the JCVI to depart from a fair and consistent application of healthcare rationing.

Preeclampsia Prevention Using Routine Versus Screening Test-Indicated Aspirin in Low-Risk Women

The objective was to evaluate whether routine aspirin 75 mg is more cost-effective than the Fetal Medicine Foundation screen-and-treat approach for preeclampsia prevention in low-risk nulliparous women. A health economic decision analytical model was devised to estimate the discounted net health and cost outcomes of routine aspirin versus Fetal Medicine Foundation screening test-indicated aspirin for a cohort of 100 000 low-risk nulliparous women. Both strategies were compared with no intervention. A subanalysis also compared disaggregated components of the algorithm. The analysis used data from hospital administration, literature, and a randomized controlled trial. Sensitivity analyses assessed the impact of aspirin adherence, test cost, and accuracy on study results. Presumed rates of preeclampsia were 3.75% with no intervention versus 0.45% with aspirin use. Results found that routine aspirin was the preferred strategy, in terms of greater health gains and larger cost savings. It provided 163 quality-adjusted life-years relative to no intervention, whereas the screen-and-treat policy achieved 108 quality-adjusted life-years. Routine aspirin would result in an estimated cost saving of €14.9 million annually relative to no intervention, whereas screen-and-treat approach would result in a smaller cost saving of €3.1 million. When the analysis was extended to consider alternative screen-and-treat strategies, routine aspirin remained the optimally cost-effective approach. In conclusion, routine aspirin use in low-risk nulliparous women has a greater health gain and cost saving compared with both the Fetal Medicine Foundation and other screen-and-treat approaches.

Systematic Review of Costs and Effects of Self-Management Interventions for Chronic Musculoskeletal Pain: Spotlight on Analytic Perspective and Outcomes Assessment

BACKGROUND

Evidence for the cost-effectiveness of self-management interventions for chronic musculoskeletal pain (CMP) lacks consensus, which may be due to variability in the costing methods employed.

PURPOSE

The purposes of the study were to identify how costs and effects have been assessed in economic analysis of self-management interventions for CMP and to identify the effect of the chosen analytical perspective on cost-effectiveness conclusions.

DATA SOURCES

Five databases were searched for all study designs using relevant terms.

STUDY SELECTION

Two independent researchers reviewed all titles for predefined inclusion criteria: adults (≥18 years of age) with CMP, interventions with a primary aim of promoting self-management, and conducted a cost analysis.

DATA EXTRACTION

Descriptive data including population, self-management intervention, analytical perspective, and costs and effects measured were collected by one reviewer and checked for accuracy by a second reviewer.

DATA SYNTHESIS

Fifty-seven studies were identified: 65% (n = 37) chose the societal perspective, of which 89% (n = 33) captured health care utilization, 92% (n = 34) reported labor productivity, 65% (n = 24) included intervention delivery, and 59% (n = 22) captured patient/family costs. Types of costs varied in all studies. Eight studies conducted analyses from both health service and societal perspectives; cost-effectiveness estimates varied with perspective chosen, but in no case was the difference sufficient to change overall policy recommendations.

LIMITATIONS

Chronic musculoskeletal pain conditions where self-management is recommended, but not as a primary treatment, were excluded. Gray literature was excluded.

CONCLUSION

Substantial heterogeneity in the cost components captured in the assessment of self-management for CMP was found; this was independent of the analytic perspective used. Greater efforts to ensure complete and consistent costings are required if reliable cost-effectiveness evidence of self-management interventions is to be generated and to inform the most appropriate perspective for economic analyses in this field.

Health-related quality of life as measured by the EQ-5D in the prevention, screening and management of cervical disease: A systematic review

PURPOSE

Cost-effectiveness analyses (CEAs) of screening can be highly sensitive to the health-related quality of life (HRQoL) effects of screen tests and subsequent treatment. Accordingly, accurate assessment of HRQoL is essential. We reviewed the literature regarding HRQoL in cervical prevention and management in order to appraise the current evidence regarding this important input to CEA.

METHODS

We searched the MEDLINE, Scopus and EconLit databases for studies that estimated HRQoL in cervical cancer prevention and management published January 1995-December 2015. The primary inclusion criterion was for studies that assess HRQoL using the EQ-5D. Data were abstracted from eligible studies on setting, elicitation group, sample size, elicitation instruments, health state valuations, study design and follow-up. We assessed the quality and comparability of the studies with a particular focus on the HRQoL reported across states and groups.

RESULTS

Fifteen papers met the inclusion criteria. Most used patient elicitation groups (n = 11), 2 used the general public and 2 used a mix of both. Eight studies were cross-sectional and seven were longitudinal. Six studies used both the EQ-5D-3L and the EQ-VAS together with other measures of overall HRQoL or condition-specific instruments. Extensive heterogeneity was observed across study characteristics.

CONCLUSIONS

Our results reveal the challenges of sourcing reliable estimates of HRQoL for use in CEAs of cervical cancer prevention and treatment. The EQ-5D appears insufficiently sensitive for some health states. A more general problem is the paucity of HRQoL estimates for many health states and their change over time.

Discounting the Recommendations of the Second Panel on Cost-Effectiveness in Health and Medicine

Twenty years ago, the "Panel on Cost-effectiveness in Health and Medicine" published a landmark text setting out appropriate methods for conducting cost-effectiveness analyses of health technologies. In the two decades since, the methods used for economic evaluations have advanced substantially. Recently, a "second panel" (hereafter "the panel") was convened to update the text and its recommendations were published in November 2016. The purpose of this paper is to critique the panel's updated guidance regarding the discounting of costs and health effects. The advances in discounting methodology since the first panel include greater theoretical clarity regarding the specification of discount rates, how these rates vary with the analytical perspective chosen, and whether the healthcare budget is constrained. More specifically, there has been an important resolution of the debate regarding the conditions under which differential discounting of costs and health effects is appropriate. We show that the panel's recommendations are inconsistent with this recent literature. Importantly, the panel's departures from previously published findings do not arise from an alternative interpretation of theory; rather, we demonstrate that this is due to fundamental errors in methodology and logic. The panel also failed to conduct a formal review of relevant empirical evidence. We provide a number of suggestions for how the panel's recommendations could be improved in future.

Evaluation of the Effectiveness and Cost-Effectiveness of Personalized Surveillance After Colorectal Adenomatous Polypectomy

Lifetime risk of developing colorectal cancer is 5%, and 5-year survival at early stage is 92%. Individuals with precancerous lesions removed at primary screening are typically recommended surveillance colonoscopy. Because greater benefits are anticipated for those with higher risk of colorectal cancer, scope for risk-specific surveillance recommendations exists. This review assesses published cost-effectiveness estimates of postpolypectomy surveillance to consider the potential for personalized recommendations by risk group. Meta-analyses of incidence of advanced neoplasia postpolypectomy for low-risk cases were comparable to those without adenoma, with both rates under the lifetime risk of 5%. This group may not benefit from intensive surveillance, which risks unnecessary harm and inefficient use of often scarce colonoscopy capacity. Therefore, greater personalization through deintensified strategies for low-risk individuals could be beneficial. The potential for noninvasive testing, such as fecal immunochemical tests, combined with primary prevention or chemoprevention may reserve colonoscopy for targeted use in personalized risk-stratified surveillance. This review appraised evidence supporting a program of personalized surveillance in patients with colorectal adenoma according to risk group and compared the effectiveness of surveillance colonoscopy with alternative prevention strategies. It assessed trade-offs among costs, benefits, and adverse effects that must be considered in a decision to adopt or reject personalized surveillance.

The Irish Cost-Effectiveness Threshold: Does it Support Rational Rationing or Might it Lead to Unintended Harm to Ireland's Health System?

Ireland is one of the few countries worldwide to have an explicit cost-effectiveness threshold. In 2012, an agreement between government and the pharmaceutical industry that provided substantial savings on existing medications set the threshold at €45,000/quality-adjusted life-year (QALY). This replaced a previously unofficial threshold of €20,000/QALY. According to the agreement, drugs within the threshold will be granted reimbursement, whereas those exceeding it may still be approved following further negotiation. A number of drugs far exceeding the threshold have been approved recently. The agreement only applies to pharmaceuticals. There are four reasons for concern regarding Ireland's threshold. The absence of an explicit threshold for non-drug interventions leaves it unclear if there is parity in willingness to pay across all interventions. As the threshold resembles a price floor rather than a ceiling, in principle it only offers a weak barrier to cost-ineffective interventions. It has no empirical basis. Finally, it is probably too high given recent estimates of a threshold for the UK based on the cost effectiveness of services forgone of approximately £13,000/QALY. An excessive threshold risks causing the Irish health system unintended harm. The lack of an empirically informed threshold means the policy recommendations of cost-effectiveness analysis cannot be considered as fully evidence- based rational rationing. Policy makers should consider these issues and recent Irish legislation that defined cost effectiveness in terms of the opportunity cost of services forgone when choosing what threshold to apply once the current industry agreement expires at the end of 2015

Dealing with Time in Health Economic Evaluation: Methodological Issues and Recommendations for Practice

Time is an important aspect of health economic evaluation, as the timing and duration of clinical events, healthcare interventions and their consequences all affect estimated costs and effects. These issues should be reflected in the design of health economic models. This article considers three important aspects of time in modelling: (1) which cohorts to simulate and how far into the future to extend the analysis; (2) the simulation of time, including the difference between discrete-time and continuous-time models, cycle lengths, and converting rates and probabilities; and (3) discounting future costs and effects to their present values. We provide a methodological overview of these issues and make recommendations to help inform both the conduct of cost-effectiveness analyses and the interpretation of their results. For choosing which cohorts to simulate and how many, we suggest analysts carefully assess potential reasons for variation in cost effectiveness between cohorts and the feasibility of subgroup-specific recommendations. For the simulation of time, we recommend using short cycles or continuous-time models to avoid biases and the need for half-cycle corrections, and provide advice on the correct conversion of transition probabilities in state transition models. Finally, for discounting, analysts should not only follow current guidance and report how discounting was conducted, especially in the case of differential discounting, but also seek to develop an understanding of its rationale. Our overall recommendations are that analysts explicitly state and justify their modelling choices regarding time and consider how alternative choices may impact on results.

HIQA's CEA of Breast Screening: Pragmatic Policy Recommendations are Welcome, but ACERs Reported as ICERs are Not

The Health Information and Quality Authority (HIQA) is Ireland's statutory cost-effectiveness analysis (CEA) agency. It recently published a CEA of screening strategies for women at elevated risk of breast cancer. Although the strategies recommended by HIQA exceed Ireland's cost-effectiveness threshold, they can reasonably be welcomed as a pragmatic response to constraints on disinvestment and are expected to improve screening cost-effectiveness. What is not welcome, however, is HIQA's reporting of average cost-effectiveness ratios (ACERs) as incremental cost-effectiveness ratios (ICERs). The distinction between ACERs and ICERs is well understood in CEA, as is the fact that ICERs not ACERs are the appropriate metric to determine cost-effectiveness. This article critiques HIQA's reporting, considering the implications for the particular case of breast cancer screening and the broader context of consistency of and confidence in CEA as a guide to resource allocation in Ireland. The reporting of ACERs as ICERs is unlikely to be of any great significance in the particular case of screening women at elevated risk of breast cancer, given likely constraints on disinvestment. Despite this, ICERs still need to be reported correctly. If thresholds are exceeded in certain cases, then it is important that decision makers appreciate by how much. More generally, using ACERs in some cases and ICERs in others raises concerns that methods are being applied inconsistently, which risks compromising confidence in CEA in Ireland. As Ireland's statutory CEA authority, HIQA has a special onus of responsibility to ensure established methods are applied correctly.

Beware of Kinked Frontiers: A Systematic Review of the Choice of Comparator Strategies in Cost-Effectiveness Analyses of Human Papillomavirus Testing in Cervical Screening

OBJECTIVES

To systematically review the choice of comparator strategies in cost-effectiveness analyses (CEAs) of human papillomavirus testing in cervical screening.

METHODS

The PubMed, Web of Knowledge, and Scopus databases were searched to identify eligible model-based CEAs of cervical screening programs using human papillomavirus testing. The eligible CEAs were reviewed to investigate what screening strategies were chosen for analysis and how this choice might have influenced estimates of the incremental cost-effectiveness ratio (ICER). Selected examples from the reviewed studies are presented to illustrate how the omission of relevant comparators might influence estimates of screening cost-effectiveness.

RESULTS

The search identified 30 eligible CEAs. The omission of relevant comparator strategies appears likely in 18 studies. The ICER estimates in these cases are probably lower than would be estimated had more comparators been included. Five of the 30 studies restricted relevant comparator strategies to sensitivity analyses or other subanalyses not part of the principal base-case analysis. Such exclusion of relevant strategies from the base-case analysis can result in cost-ineffective strategies being identified as cost-effective.

CONCLUSIONS

Many of the CEAs reviewed appear to include insufficient comparator strategies. In particular, they omit strategies with relatively long screening intervals. Omitting relevant comparators matters particularly if it leads to the underestimation of ICERs for strategies around the cost-effectiveness threshold because these strategies are the most policy relevant from the CEA perspective. Consequently, such CEAs may not be providing the best possible policy guidance and lead to the mistaken adoption of cost-ineffective screening strategies.

Assessing cost-effectiveness of HPV vaccines with decision analytic models: what are the distinct challenges of low- and middle-income countries? A protocol for a systematic review

BACKGROUND

Cervical cancer poses a huge health burden, both to developed and developing nations, making prevention and control strategies necessary. However, the challenges of designing and implementing prevention strategies differ for low- and middle-income countries (LMICs) as compared to countries with fully developed health care systems. Moreover, for many LMICs, much of the data needed for decision analytic modelling, such as prevalence, will most likely only be partly available or measured with much larger uncertainty. Lastly, imperfect implementation of human papillomavirus (HPV) vaccination may influence the effectiveness of cervical cancer prevention in unpredictable ways. This systematic review aims to assess how decision analytic modelling studies of HPV cost-effectiveness in LMICs accounted for the particular challenges faced in such countries. Specifically, the study will assess the following: (1) whether the existing literature on cost-effectiveness modelling of HPV vaccines acknowledges the distinct challenges of LMICs, (2) how these challenges were accommodated in the models, (3) whether certain parameters systemically exhibited large degrees of uncertainty due to lack of data and how influential were these parameters on model-based recommendations, and (4) whether the choice of modelling herd immunity influences model-based recommendations, especially when coverage of a HPV vaccination program is not optimal.

METHODS

We will conduct a systematic review to identify suitable studies from MEDLINE (via PubMed), EMBASE, NHS Economic Evaluation Database (NHS EED), EconLit, Web of Science, and CEA Registry. Searches will be conducted for studies of interest published since 2006. The searches will be supplemented by hand searching of the most relevant papers found in the search. Studies will be critically appraised using Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement checklist. We will undertake a descriptive, narrative, and interpretative synthesis of data to address the study objectives.

DISCUSSION

The proposed systematic review will assess how the cost-effectiveness studies of HPV vaccines accounted for the distinct challenges of LMICs. The gaps identified will expose areas for additional research as well as challenges that need to be accounted for in future modelling studies.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42015017870.

Some inconsistencies in NICE's consideration of social values

The UK's National Institute for Health and Care Excellence (NICE) recently proposed amendments to its methods for the appraisal of health technologies. Previous amendments in 2009 and 2011 placed a greater value on the health of patients at the "end of life" and in cases where "treatment effects are both substantial in restoring health and sustained over a very long period". Drawing lessons from these previous amendments, we critically appraise NICE's proposals. The proposals repeal "end of life" considerations but add consideration of the "proportional" and "absolute" quality-adjusted life-year (QALY) loss from illness. NICE's cost-effectiveness threshold may increase from £20,000 to £50,000 per QALY on the basis of these and four other considerations: the "certainty of the ICER [incremental cost-effectiveness ratio]"; whether health-related quality of life is "inadequately captured"; the "innovative nature" of the technology; and "non-health objectives of the NHS". We demonstrate that NICE's previous amendments are flawed; they contain logical inconsistencies which can result in different values being placed on health gains for identical patients, and they do not apply value weights to patients bearing the opportunity cost of NICE's recommendations. The proposals retain both flaws and are also poorly justified. Applying value weights to patients bearing the opportunity cost would lower NICE's threshold, in some cases to below £20,000 per QALY. Furthermore, this baseline threshold is higher than current estimates of the opportunity cost. NICE's proposed threshold range is too high, for empirical and methodological reasons. NICE's proposals will harm the health of unidentifiable patients, whilst privileging the identifiable beneficiaries of new health technologies.

NICE's selective application of differential discounting: ambiguous, inconsistent, and unjustified

The National Institute for Health and Clinical Excellence (NICE) recently recommended differential discounting of costs and health effects in the economic appraisal of health care interventions in certain circumstances. The recommendation was published in an amendment to NICE's Guide to the Methods of Technology Appraisal. The amendment states that differential discounting should be applied where "treatment effects are both substantial in restoring health and sustained over a very long period (normally at least 30 years)." Renewed interest in differential discounting from NICE is welcome; however, the recommendation's selective application of differential discounting raises a number of concerns. The stated criteria for applying differential discounting are ambiguous. The rationale for the selective application of differential discounting has not been articulated by NICE and is questionable. The selective application of differential discounting leads to several inconsistencies, the most concerning of which is the lower valuation of health gains for those with less than 30 years remaining life expectancy, which can be interpreted as age discrimination. Furthermore, the discount rates chosen by NICE do not appear to be informed by recent advances in the theoretical understanding of differential discounting. NICE's apparent motivation for recommending differential discounting was to ensure a favorable cost-effectiveness ratio for a pediatric oncology drug. While flexibility may be appropriate to allow some interventions that exceed conventional cost-effectiveness thresholds to be adopted, the selective adjustment of appraisal methods is problematic and without justification.

The influence of disease risk on the optimal time interval between screens for the early detection of cancer: a mathematical approach

The intervals between screens for the early detection of diseases such as breast and colon cancer suggested by screening guidelines are typically based on the average population risk of disease. With the emergence of ever more biomarkers for cancer risk prediction and the development of personalized medicine, there is a need for risk-specific screening intervals. The interval between successive screens should be shorter with increasing cancer risk. A risk-dependent optimal interval is ideally derived from a cost-effectiveness analysis using a validated simulation model. However, this is time-consuming and costly. We propose a simplified mathematical approach for the exploratory analysis of the implications of risk level on optimal screening interval. We develop a mathematical model of the optimal screening interval for breast cancer screening. We verified the results by programming the simplified model in the MISCAN-Breast microsimulation model and comparing the results. We validated the results by comparing them with the results of a full, published MISCAN-Breast cost-effectiveness model for a number of different risk levels. The results of both the verification and validation were satisfactory. We conclude that the mathematical approach can indicate the impact of disease risk on the optimal screening interval.

A mathematical approach for evaluating Markov models in continuous time without discrete-event simulation

Markov models are a simple and powerful tool for analyzing the health and economic effects of health care interventions. These models are usually evaluated in discrete time using cohort analysis. The use of discrete time assumes that changes in health states occur only at the end of a cycle period. Discrete-time Markov models only approximate the process of disease progression, as clinical events typically occur in continuous time. The approximation can yield biased cost-effectiveness estimates for Markov models with long cycle periods and if no half-cycle correction is made. The purpose of this article is to present an overview of methods for evaluating Markov models in continuous time. These methods use mathematical results from stochastic process theory and control theory. The methods are illustrated using an applied example on the cost-effectiveness of antiviral therapy for chronic hepatitis B. The main result is a mathematical solution for the expected time spent in each state in a continuous-time Markov model. It is shown how this solution can account for age-dependent transition rates and discounting of costs and health effects, and how the concept of tunnel states can be used to account for transition rates that depend on the time spent in a state. The applied example shows that the continuous-time model yields more accurate results than the discrete-time model but does not require much computation time and is easily implemented. In conclusion, continuous-time Markov models are a feasible alternative to cohort analysis and can offer several theoretical and practical advantages.

Multicohort models in cost-effectiveness analysis: why aggregating estimates over multiple cohorts can hide useful information

BACKGROUND

Models used in cost-effectiveness analysis (CEA) of screening programs may include 1 or many birth cohorts of patients. As many screening programs involve multiple screens over many years for each birth cohort, the actual implementation of screening often involves multiple concurrent recipient cohorts. Consequently, some advocate modeling all recipient cohorts rather than 1 birth cohort, arguing it more accurately represents actual implementation. However, reporting the cost-effectiveness estimates for multiple cohorts on aggregate rather than per cohort will fail to account for any heterogeneity in cost-effectiveness between cohorts. Such heterogeneity may be policy relevant where there is considerable variation in cost-effectiveness between cohorts, as in the case of cancer screening programs with multiple concurrent recipient birth cohorts, each at different stages of screening at any one point in time.

OBJECTIVE

The purpose of this study is to illustrate the potential disadvantages of aggregating cost-effectiveness estimates over multiple cohorts, without first considering the disaggregate estimates. Analysis. We estimate the cost-effectiveness of 2 alternative cervical screening tests in a multicohort model and compare the aggregated and per-cohort estimates. We find instances in which the policy choices suggested by the aggregate and per-cohort results differ. We use this example to illustrate a series of potential disadvantages of aggregating CEA estimates over cohorts.

CONCLUSIONS

Recent recommendations that CEAs should consider the cost-effectiveness of more than just a single cohort appear justified, but the aggregation of estimates across multiple cohorts into a single estimate does not.

Practical implications of differential discounting in cost-effectiveness analyses with varying numbers of cohorts

OBJECTIVE

To call attention to the influence of the number of birth-cohorts used in cost-effectiveness analysis (CEA) models on incremental cost-effectiveness ratios (ICERs) under differential discounting.

METHODS

The consequences of increasing the number of birth-cohorts are demonstrated using a CEA of cervical cancer prevention as an example. The cost-effectiveness of vaccinating 12-year-old girls against the human papillomavirus is estimated with the MISCAN microsimulation screening analysis model for 1, 10, 20, and 30 birth-cohorts. Costs and health effects are discounted with equal rates of 4% and alternatively with differential rates of 4% and 1.5% respectively. The effects of increasing the number of cohorts are shown by comparing the ICERs under equal and differential discounting.

RESULTS

The ICER decreases as the number of cohorts increases under differential discounting, but not under equal discounting.

CONCLUSIONS

The variation of ICERs with the number of cohorts under differential discounting prompts questions regarding the appropriate specification of CEA models and interpretation of their results. In particular, it raises concerns that arbitrary variation in study specification leads to arbitrary variation in results. Such variations could lead to erroneous policy decisions. These findings are relevant to CEA guidance authorities, CEA practitioners, and decision makers. Our results do not imply a problem with differential discounting per se, yet they highlight the need for practical guidance for its use.