Value Of Partial Perfect Information Research Articles

Prioritizing Future Research on Allopurinol and Febuxostat for the Management of Gout: Value of Information Analysis.

The aim of this study was to quantify the value of conducting additional research and reducing uncertainty regarding the cost effectiveness of allopurinol and febuxostat for the management of gout. We used a previously developed Markov model that evaluated the cost effectiveness of nine urate-lowering strategies: no treatment, allopurinol-only fixed dose (300mg), allopurinol-only dose escalation (up to 800mg), febuxostat-only fixed dose (80mg), febuxostat-only dose escalation (up to 120mg), allopurinol-febuxostat sequential therapy fixed dose, allopurinol-febuxostat sequential therapy dose escalation, febuxostat-allopurinol sequential therapy fixed dose, and febuxostat-allopurinol sequential therapy dose escalation. Each strategy was evaluated over the lifetime of a hypothetical gout patient. We calculated population expected value of perfect information (EVPI). We used a linear regression meta-modeling approach to calculate population expected value of partial perfect information (EVPPI), and a Gaussian approximation to calculate the population expected value of sample information for parameters (EVSI) and the expected net benefit of sampling (ENBS) for four potential study designs: (1) an allopurinol efficacy trial; (2) a febuxostat efficacy trial; (3) a prospective observational study evaluating health utilities; and (4) a comprehensive study evaluating the efficacy of allopurinol and febuxostat and health utilities. A 5-year decision time horizon was used in the base-case analysis. EVPI varied by a decision maker's willingness-to-pay (WTP) per quality-adjusted life-year (QALY) and was $US900 million for WTP of $US60,000 per QALY. Population EVPPI was highest across all WTP values for study design #4. For study design #4 and a WTP of $US60,000 per QALY, the optimal sample size was 735 patients per study arm. Future studies are needed to evaluate the effectiveness of allopurinol and febuxostat dose escalation.

A Review of Methods for Analysis of the Expected Value of Information.

In recent years, value-of-information analysis has become more widespread in health economic evaluations, specifically as a tool to guide further research and perform probabilistic sensitivity analysis. This is partly due to methodological advancements allowing for the fast computation of a typical summary known as the expected value of partial perfect information (EVPPI). A recent review discussed some approximation methods for calculating the EVPPI, but as the research has been active over the intervening years, that review does not discuss some key estimation methods. Therefore, this paper presents a comprehensive review of these new methods. We begin by providing the technical details of these computation methods. We then present two case studies in order to compare the estimation performance of these new methods. We conclude that a method based on nonparametric regression offers the best method for calculating the EVPPI in terms of accuracy, computational time, and ease of implementation. This means that the EVPPI can now be used practically in health economic evaluations, especially as all the methods are developed in parallel with R functions and a web app to aid practitioners.

Some Health States Are Better Than Others: Using Health State Rank Order to Improve Probabilistic Analyses.

Probabilistic sensitivity analyses (PSA) may lead policy makers to take nonoptimal actions due to misestimates of decision uncertainty caused by ignoring correlations. We developed a method to establish joint uncertainty distributions of quality-of-life (QoL) weights exploiting ordinal preferences over health states. Our method takes as inputs independent, univariate marginal distributions for each QoL weight and a preference ordering. It establishes a correlation matrix between QoL weights intended to preserve the ordering. It samples QoL weight values from their distributions, ordering them with the correlation matrix. It calculates the proportion of samples violating the ordering, iteratively adjusting the correlation matrix until this proportion is below an arbitrarily small threshold. We compare our method with the uncorrelated method and other methods for preserving rank ordering in terms of violation proportions and fidelity to the specified marginal distributions along with PSA and expected value of partial perfect information (EVPPI) estimates, using 2 models: 1) a decision tree with 2 decision alternatives and 2) a chronic hepatitis C virus (HCV) Markov model with 3 alternatives. All methods make tradeoffs between violating preference orderings and altering marginal distributions. For both models, our method simultaneously performed best, with largest performance advantages when distributions reflected wider uncertainty. For PSA, larger changes to the marginal distributions induced by existing methods resulted in differing conclusions about which strategy was most likely optimal. For EVPPI, both preference order violations and altered marginal distributions caused existing methods to misestimate the maximum value of seeking additional information, sometimes concluding that there was no value. Analysts can characterize the joint uncertainty in QoL weights to improve PSA and value-of-information estimates using Open Source implementations of our method.

Estimating the expected value of partial perfect information in health economic evaluations using integrated nested Laplace approximation.

The Expected Value of Perfect Partial Information (EVPPI) is a decision‐theoretic measure of the ‘cost’ of parametric uncertainty in decision making used principally in health economic decision making. Despite this decision‐theoretic grounding, the uptake of EVPPI calculations in practice has been slow. This is in part due to the prohibitive computational time required to estimate the EVPPI via Monte Carlo simulations. However, recent developments have demonstrated that the EVPPI can be estimated by non‐parametric regression methods, which have significantly decreased the computation time required to approximate the EVPPI. Under certain circumstances, high‐dimensional Gaussian Process (GP) regression is suggested, but this can still be prohibitively expensive. Applying fast computation methods developed in spatial statistics using Integrated Nested Laplace Approximations (INLA) and projecting from a high‐dimensional into a low‐dimensional input space allows us to decrease the computation time for fitting these high‐dimensional GP, often substantially. We demonstrate that the EVPPI calculated using our method for GP regression is in line with the standard GP regression method and that despite the apparent methodological complexity of this new method, R functions are available in the package BCEA to implement it simply and efficiently. © 2016 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

A systematic review, evidence synthesis and meta-analysis of quantitative and qualitative studies evaluating the clinical effectiveness, the cost-effectiveness, safety and acceptability of interventions to prevent postnatal depression.

Postnatal depression (PND) is a major depressive disorder in the year following childbirth, which impacts on women, their infants and their families. A range of interventions has been developed to prevent PND. To (1) evaluate the clinical effectiveness, cost-effectiveness, acceptability and safety of antenatal and postnatal interventions for pregnant and postnatal women to prevent PND; (2) apply rigorous methods of systematic reviewing of quantitative and qualitative studies, evidence synthesis and decision-analytic modelling to evaluate the preventive impact on women, their infants and their families; and (3) estimate cost-effectiveness. We searched MEDLINE, EMBASE, Science Citation Index and other databases (from inception to July 2013) in December 2012, and we were updated by electronic alerts until July 2013. Two reviewers independently screened titles and abstracts with consensus agreement. We undertook quality assessment. All universal, selective and indicated preventive interventions for pregnant women and women in the first 6 postnatal weeks were included. All outcomes were included, focusing on the Edinburgh Postnatal Depression Scale (EPDS), diagnostic instruments and infant outcomes. The quantitative evidence was synthesised using network meta-analyses (NMAs). A mathematical model was constructed to explore the cost-effectiveness of interventions contained within the NMA for EPDS values. From 3072 records identified, 122 papers (86 trials) were included in the quantitative review. From 2152 records, 56 papers (44 studies) were included in the qualitative review. The results were inconclusive. The most beneficial interventions appeared to be midwifery redesigned postnatal care [as shown by the mean 12-month EPDS score difference of -1.43 (95% credible interval -4.00 to 1.36)], person-centred approach (PCA)-based and cognitive-behavioural therapy (CBT)-based intervention (universal), interpersonal psychotherapy (IPT)-based intervention and education on preparing for parenting (selective), promoting parent-infant interaction, peer support, IPT-based intervention and PCA-based and CBT-based intervention (indicated). Women valued seeing the same health worker, the involvement of partners and access to several visits from a midwife or health visitor trained in person-centred or cognitive-behavioural approaches. The most cost-effective interventions were estimated to be midwifery redesigned postnatal care (universal), PCA-based intervention (indicated) and IPT-based intervention in the sensitivity analysis (indicated), although there was considerable uncertainty. Expected value of partial perfect information (EVPPI) for efficacy data was in excess of £150M for each population. Given the EVPPI values, future trials assessing the relative efficacies of promising interventions appears to represent value for money. In the NMAs, some trials were omitted because they could not be connected to the main network of evidence or did not provide EPDS scores. This may have introduced reporting or selection bias. No adjustment was made for the lack of quality of some trials. Although we appraised a very large number of studies, much of the evidence was inconclusive. Interventions warrant replication within randomised controlled trials (RCTs). Several interventions appear to be cost-effective relative to usual care, but this is subject to considerable uncertainty. Several interventions appear to be cost-effective relative to usual care, but this is subject to considerable uncertainty. Future research conducting RCTs to establish which interventions are most clinically effective and cost-effective should be considered. This study is registered as PROSPERO CRD42012003273. The National Institute for Health Research Health Technology Assessment programme.

Decisions on Further Research for Predictive Biomarkers of High-Dose Alkylating Chemotherapy in Triple-Negative Breast Cancer: A Value of Information Analysis

ObjectivesTo inform decisions about the design and priority of further studies of emerging predictive biomarkers of high-dose alkylating chemotherapy (HDAC) in triple-negative breast cancer (TNBC) using value-of-information analysis. MethodsA state transition model compared treating women with TNBC with current clinical practice and four biomarker strategies to personalize HDAC: 1) BRCA1-like profile by array comparative genomic hybridization (aCGH) testing; 2) BRCA1-like profile by multiplex ligation-dependent probe amplification (MLPA) testing; 3) strategy 1 followed by X-inactive specific transcript gene (XIST) and tumor suppressor p53 binding protein (53BP1) testing; and 4) strategy 2 followed by XIST and 53BP1 testing, from a Dutch societal perspective and a 20-year time horizon. Input data came from literature and expert opinions. We assessed the expected value of partial perfect information, the expected value of sample information, and the expected net benefit of sampling for potential ancillary studies of an ongoing randomized controlled trial (RCT; NCT01057069). ResultsThe expected value of partial perfect information indicated that further research should be prioritized to the parameter group including “biomarkers’ prevalence, positive predictive value (PPV), and treatment response rates (TRRs) in biomarker-negative patients and patients with TNBC” (€639 million), followed by utilities (€48 million), costs (€40 million), and transition probabilities (TPs) (€30 million). By setting up four ancillary studies to the ongoing RCT, data on 1) TP and MLPA prevalence, PPV, and TRR; 2) aCGH and aCGH/MLPA plus XIST and 53BP1 prevalence, PPV, and TRR; 3) utilities; and 4) costs could be simultaneously collected (optimal size = 3000). ConclusionsFurther research on predictive biomarkers for HDAC should focus on gathering data on TPs, prevalence, PPV, TRRs, utilities, and costs from the four ancillary studies to the ongoing RCT.

Graduated compression stockings for the prevention of deep-vein thrombosis in postoperative surgical patients: a systematic review and economic model with a value of information analysis.

Deep-vein thrombosis (DVT) can occur in surgical patients. Routine prophylaxis can be pharmacological and/or mechanical [e.g. graduated compression stockings (GCSs)]. GCSs are available in knee length or thigh length. To establish the expected value of undertaking additional research addressing the relative effectiveness of thigh-length GCSs versus knee-length GCSs, in addition to pharmacoprophylaxis, for prevention of DVT in surgical patients. Systematic review and economic model, including value of information (VOI) analysis. Randomised controlled trials (RCTs) assessing thigh- or knee-length GCSs in surgical patients were eligible for inclusion. The primary outcome was incidence of DVT. DVT complications and GCSs adverse events were assessed. Random-effects meta-analysis was performed. To draw on a wider evidence base, a random-effects network meta-analysis (NMA) was undertaken for the outcome DVT. A review of trials and observational studies of patient adherence was also conducted. A decision-analytic model was developed to assess the cost-effectiveness of thigh- and knee-length GCSs and the VOI. Twenty-three RCTs were included in the review of effectiveness. There was substantial variation between trials in terms of the patient characteristics, interventions and methods of outcome assessment. Five trials comparing knee-length with thigh-length GCSs with or without pharmacoprophylaxis were pooled; the summary estimate of effect indicated a non-significant trend favouring thigh-length GCSs [odds ratio (OR) 1.48, 95% confidence interval (CI) 0.80 to 2.73]. Thirteen trials were included in the NMA. In the base-case analysis, thigh-length GCSs with pharmacoprophylaxis were more effective than knee-length GCSs with pharmacoprophylaxis (knee vs. thigh OR 1.76, 95% credible interval 0.82 to 3.53). Overall, thigh-length stockings with pharmacoprophylaxis was the most effective treatment, with a 0.73 probability of being the most effective treatment in a new trial of all the treatments. Patient adherence was generally higher with knee-length GCSs, and patients preferred knee-length GCSs. Thigh-length GCSs were found to be cost-effective in all but the subgroup with the lowest baseline risk, although the absolute differences in costs and effects were relatively small. The expected value of perfect information ranged from £0.2M to £178.0M depending on the scenario and subgroup. The relative effect parameters had the highest expected value of partial perfect information and ranged from £2.0M to £39.4M. The value of further research was most evident in the high-risk subgroups. There was substantial variation across the included trials in terms of patient and intervention characteristics. Many of the included trials were old and poorly reported, which reduces the reliability of the results of the review. Given that the results from both the standard meta-analysis and the NMA lacked precision (CIs were wide) owing to the heterogeneous evidence base, a new definitive trial in high-risk patients may be warranted. However, the efficiency of any further research (i.e. whether this represents value for money) is dependent on several factors, including the acquisition price of GCSs, expected compliance with thigh-length GCSs wear, and whether or not uncertainty can be resolved around possible effect modifiers, as well as the feasibility and actual cost of undertaking the proposed research. This study is registered as PROSPERO CRD42014007202. The National Institute for Health Research Health Technology Assessment programme.

H-TArget model: Early technology assessment for ext generation sequencing in oncology.

Background: Next Generation Sequencing (NGS) promises to find mutations (targets) in individual cancer patients, to subsequently prescribe targeted therapy. Currently, NGS is in development, the effects on choice of therapy and prognosis are still unclear, and the costs for targeted therapies are high. To accelerate the reimbursement process of NGS and potential new targeted therapies, and have a NGS-panel available for patients in the earliest possible stage, early Technology Assessment (TA) is performed to inform policy making around NGS in the Netherlands. One of the aims of the TA was to conduct a cost-effectiveness analysis. Methods: We constructed a target-based decision model (H-TArget) to estimate the cost-effectiveness of NGS versus single- and no testing. Standard- and targeted therapies in first and second line for 9 targets (BRAF, KRAS, NRAS, EGFR, ERBB/HER, MET, ROS, ALK, RET) over 3 tumor types (melanoma, non-small-cell lung cancer (NSCLC), colorectal cancer (CRC)) were incorporated. A Dutch healthcare perspective and a 5-year time horizon were adopted. Outcomes were incremental cost-effectiveness ratios (ICER) expressed in {euro}/quality adjusted life year (QALY). The threshold for cost-effectiveness is 80k in the Netherlands, which means that the concerning technology is cost-effective if the ICER is below this threshold. Expected Value of Partial Perfect Information (EVP(P)I) was calculated to quantify the value of further research into particular subsets of uncertain parameters. Results: The expected ICER was {euro}65k/QALY for melanoma, {euro}188k/QALY for NSCLC, and {euro}103k/QALY for CRC. As a weighted average to the three populations, the overall ICER yielded {euro}160k/QALY. The EVPI was {euro}25M for melanoma, the subsets of parameters to focus on in future research were: {euro}2M together for failures, prevalence, survival, and {euro}23M for costs. Conclusions: At the moment, using NGS is only cost-effective for melanoma. This is mostly due to the high costs of targeted therapies and the fact that the effects are still small. Based on our findings, industry should strive for a significant cost reduction of targeted therapies or achieve a spectacular improvement in effectiveness, which could improve the cost-effectiveness.

Use of value of information in UK health technology assessments.

The aim of this study was to identify and critically appraise the use of Value of Information (VOI) analyses undertaken as part of health technology assessment (HTA) reports in England and Wales. A systematic review of National Institute for Health Research (NIHR) funded HTA reports published between 2004 and 2013 identified the use of VOI methods and key analytical details in terms of: (i) types of VOI methodology used; (ii) parameters and key assumptions; and (iii) conclusions drawn in terms of the need for further research. A total of 512 HTA reports were published during the relevant timeframe. Of these, 203 reported systematic review and economic modeling studies and 25 of these had used VOI method(s). Over half of the twenty-five studies (n = 13) conducted both EVPI (Expected Value of Perfect Information) and EVPPI (Expected Value of Partial Perfect Information) analyses. Eight studies conducted EVPI analysis, three studies conducted EVPI, EVPPI, and EVSI (Expected Value of Sampling Information) analyses and one study conducted EVSI analysis only. The level of detail reporting the methods used to conduct the VOI analyses varied. This review has shown that the frequency of the use of VOI methods is increasing at a slower pace compared with the published volume of HTA reports. This review also suggests that analysts reporting VOI method(s) in HTA reports should aim to describe the method(s) in sufficient detail to enable and encourage decision-makers guiding research prioritization decisions to use the potentially valuable outputs from quantitative VOI analyses.

Expected Value of Partial Perfect Information for the Disability Progression Efficacy of Teriflunomide and Fingolimod in the Treatment of Relapsing-Remitting Multiple Sclerosis.

Expected Value of Partial Perfect Information for the Disability Progression Efficacy of Teriflunomide and Fingolimod in the Treatment of Relapsing-Remitting Multiple Sclerosis.

Network vs. Pairwise Meta-Analyses: A Case Study of the Impact of an Evidence-Synthesis Paradigm on Value of Information Outcomes

To evaluate the impact of using two evidence-synthesis paradigms, pairwise meta-analysis (PMA) vs. network meta-analysis (NMA), on the expected value of information (EVI) outcomes, using pharmacotherapy of chronic obstructive pulmonary disease as a case study. Bayesian random-effects PMAs were performed for each pharmacotherapy vs. placebo, and a Bayesian random-effects NMA was performed combining both placebo-controlled and head-to-head trials. Both provided comparative rate ratio (RR) estimates between each pharmacotherapy vs. placebo. A Markov model was developed to project costs and quality-adjusted life-years of five commonly used treatments for chronic obsructive pulmonary disorder. RRs for the treatment effect compared with placebo derived using PMA and NMA were used alongside values from the literature to populate the model. In addition to standard cost-effectiveness outputs, we calculated and compared the expected value of perfect information (EVPI) and the expected value of partial perfect information (EVPPI) for treatment effects, for comparisons that included all or a subset of treatments. The network of evidence included five different treatments, compared in 19 randomized controlled trials (RCTs), which in total included 28,172 individuals. The cost-effectiveness outcomes were similar between the two evidence-synthesis paradigms. The individual EVPI for all treatments was Can$1,262 for PMA-based analyses and Can$572 for NMA-based analyses. For all comparisons involving two, three, or four treatments, the comparison with the highest EVPI was different between the two methods. Similarly, the choice of PMA or NMA had resulted in substantially different EVPPI rankings. Our case study shows that the choice of PMA or NMA can have significant effects on the EVI results. Under comparable conditions, the incorporation of more evidence in the NMA most likely increases the precision of estimates and therefore is likely to result in lower EVI outcomes. As our study demonstrates, the difference in EVI outcomes can be substantial, potentially affecting the decision to conduct research and the design of future research.

Strategies for Efficient Computation of the Expected Value of Partial Perfect Information

Expected value of information methods evaluate the potential health benefits that can be obtained from conducting new research to reduce uncertainty in the parameters of a cost-effectiveness analysis model, hence reducing decision uncertainty. Expected value of partial perfect information (EVPPI) provides an upper limit to the health gains that can be obtained from conducting a new study on a subset of parameters in the cost-effectiveness analysis and can therefore be used as a sensitivity analysis to identify parameters that most contribute to decision uncertainty and to help guide decisions around which types of study are of most value to prioritize for funding. A common general approach is to use nested Monte Carlo simulation to obtain an estimate of EVPPI. This approach is computationally intensive, can lead to significant sampling bias if an inadequate number of inner samples are obtained, and incorrect results can be obtained if correlations between parameters are not dealt with appropriately. In this article, we set out a range of methods for estimating EVPPI that avoid the need for nested simulation: reparameterization of the net benefit function, Taylor series approximations, and restricted cubic spline estimation of conditional expectations. For each method, we set out the generalized functional form that net benefit must take for the method to be valid. By specifying this functional form, our methods are able to focus on components of the model in which approximation is required, avoiding the complexities involved in developing statistical approximations for the model as a whole. Our methods also allow for any correlations that might exist between model parameters. We illustrate the methods using an example of fluid resuscitation in African children with severe malaria.

An Integrated Approach to Evaluating Alternative Risk Prediction Strategies: A Case Study Comparing Alternative Approaches for Preventing Invasive Fungal Disease

ObjectivesThis article proposes an integrated approach to the development, validation, and evaluation of new risk prediction models illustrated with the Fungal Infection Risk Evaluation study, which developed risk models to identify non-neutropenic, critically ill adult patients at high risk of invasive fungal disease (IFD). MethodsOur decision-analytical model compared alternative strategies for preventing IFD at up to three clinical decision time points (critical care admission, after 24 hours, and end of day 3), followed with antifungal prophylaxis for those judged “high” risk versus “no formal risk assessment.” We developed prognostic models to predict the risk of IFD before critical care unit discharge, with data from 35,455 admissions to 70 UK adult, critical care units, and validated the models externally. The decision model was populated with positive predictive values and negative predictive values from the best-fitting risk models. We projected lifetime cost-effectiveness and expected value of partial perfect information for groups of parameters. ResultsThe risk prediction models performed well in internal and external validation. Risk assessment and prophylaxis at the end of day 3 was the most cost-effective strategy at the 2% and 1% risk threshold. Risk assessment at each time point was the most cost-effective strategy at a 0.5% risk threshold. Expected values of partial perfect information were high for positive predictive values or negative predictive values (£11 million–£13 million) and quality-adjusted life-years (£11 million). ConclusionsIt is cost-effective to formally assess the risk of IFD for non-neutropenic, critically ill adult patients. This integrated approach to developing and evaluating risk models is useful for informing clinical practice and future research investment.

How much is another randomized trial of lymph node dissection in endometrial cancer worth? A value of information analysis

ObjectiveThis study aimed to assess the value of a randomized controlled trial (RCT) of lymph node dissection (LND) at the time of hysterectomy for high-risk subsets of women with endometrial cancer. MethodsA modified Markov decision model compared routine LND to no LND for women with grade 3 or grades 2–3 endometrial cancer. Inputs were modeled as distributions for Monte Carlo probabilistic sensitivity and value of information (VOI) analyses. Survival without LND was modeled from Surveillance, Epidemiology and End Results program data. A hazard ratio (HR) describing survival in the high-risk group undergoing LND (estimate 0.9, 95% CI 0.6–1.1), adverse event rates, probability and type of adjuvant therapy were modeled from published RCTs. Costs were obtained from national reimbursement data. VOI estimated the value of reducing uncertainty regarding the survival benefit of LND. ResultsFor grade 3, LND had an incremental cost-effectiveness ratio of $40,183/quality-adjusted life year (QALY) compared to no LND. Acceptability curves revealed considerable uncertainty, with an expected value of perfect information of $4,195 per patient at societal willingness to pay of $50,000/QALY. The estimated value of partial perfect information regarding the HR was $3,702 per patient. Assuming 8,000 individuals annually with grade 3 endometrial cancer in the US, the upper limit of VOI for the HR was $29.6million annually. For grades 2 and 3 combined, analysis revealed a much lower likelihood of finding LND cost-effective. ConclusionA clinical trial defining the survival effect of LND in women with grade 3 endometrial cancer is a worthwhile use of resources.

Need for Speed: An Efficient Algorithm for Calculation of Single-Parameter Expected Value of Partial Perfect Information

BackgroundThe expected value of partial perfect information (EVPPI) is a theoretically justifiable and informative measure of uncertainty in decision-analytic cost-effectiveness models, but its calculation is computationally intensive because it generally requires two-level Monte Carlo simulation. We introduce an efficient, one-level simulation method for the calculation of single-parameter EVPPI. ObjectiveWe show that under mild regularity assumptions, the expectation-maximization-expectation sequence in EVPPI calculation can be transformed into an expectation-maximization-maximization sequence. By doing so, calculations can be performed in a single-step expectation by using data generated for probabilistic sensitivity analysis. We prove that the proposed estimator of EVPPI converges in probability to the true EVPPI. Methods and ResultsThe performance of the new method was empirically demonstrated by using three exemplary decision models. Our proposed method seems to achieve remarkably higher accuracy than the two-level method with a fraction of its computation costs, though the achievement in accuracy was not uniform and varied across the parameters of the models. Software is provided to calculate single-parameter EVPPI based on the probabilistic sensitivity analysis data. ConclusionsThe new method, though applicable only to single-parameter EVPPI, is fast, accurate, and easy to implement. Further research is needed to evaluate the performance of this method in more complex scenarios and to extend such a concept to similar measures of decision uncertainty.

The Role of Value-of-Information Analysis in a Health Care Research Priority Setting

The Dutch reimbursement procedure for expensive drugs requires the submission of a baseline cost-effectiveness (CE) analysis and a research plan for the period of temporary reimbursement to estimate the real-life cost-effectiveness after 4 years. The Dutch guidelines recommend a value-of-information analysis to identify the critical parameters to be studied in such an outcome study. Identify situations where sensitivity analyses are sufficient to establish the need for additional data collection and priority setting. We used a hypothetical Markov model with 3 groups of parameters. We performed deterministic and probabilistic sensitivity analyses (PSA) and analyzed the expected value of partial perfect information (EVPPI), for different configurations of input parameters and a range of threshold incremental cost-effectiveness ratios (λ). We introduced a multivariate (deterministic) sensitivity analysis and a partial PSA. Deterministic, partial PSA, and EVPPI analyses came to the same ranking of priorities for future research in most cases, irrespective of the place of the results on the CE plane. Rankings differed only when the statistical metrics that we calculated for each method were close together. When a clear ranking can be established, all methods lead to the same priority setting. If there is no clear ranking, we regard the parameters as equally important. Priority setting for future research depends on λ and the location of results on the CE plane. The EVPPI is needed to estimate the value of doing additional research, but to prioritize parameters for further research, extensive (partial probabilistic) sensitivity analyses and expected value of perfect information are often sufficient.

PRM47 Cost-Effectiveness Uncertainty Analysis Methods: A Comparison of One-Way Sensitivity, Analysis of Covariance, and Expected Value of Partial Perfect Information

PRM47 Cost-Effectiveness Uncertainty Analysis Methods: A Comparison of One-Way Sensitivity, Analysis of Covariance, and Expected Value of Partial Perfect Information

Development of an economic model to assess the cost-effectiveness of hawthorn extract as an adjunct treatment for heart failure in Australia

ObjectiveAn economic model was developed to evaluate the cost-effectiveness of hawthorn extract as an adjunctive treatment for heart failure in Australia.MethodsA Markov model of chronic heart failure was developed to...

Value of Information in the Osteoarthritis Setting

Recent National Institute for Health and Clinical Excellence (NICE) guidance recommended that when traditional NSAIDs or cyclo-oxygenase (COX)-2 selective inhibitors are used by people with osteoarthritis (OA), they should be prescribed along with a proton pump inhibitor (PPI). However, specific recommendations about the type of NSAID or COX-2 could not be made due to high levels of uncertainty in the economic evaluation. To investigate the value of obtaining further evidence to inform the economic evaluation of NSAIDs, COX-2s and PPIs for people with OA. An economic evaluation with an expected value of perfect information (EVPI) analysis was conducted, using a Markov model with data identified from a systematic review. The base-case model used adverse event data from the three largest randomized trials of COX-2 inhibitors, and we repeated the analysis using observational adverse event data. The model was run for a hypothetical population of people with OA, and subgroup analyses were conducted for people with raised gastrointestinal (GI) and cardiovascular (CV) risk. The EVPI was based upon the OA population in England - approximately 2.8 million people. Of these, 50% were assumed to use NSAIDs or COX-2 selective inhibitors for 3 months per year and 56% of these were assumed to be patients with raised GI and CV risk. The value of further information for this decision problem was very high. Population-level EVPI was £85.1 million in the low-risk group and £179.5 million in the high-risk group (2007-8 values). Expected value of partial perfect information (EVPPI) analysis showed that the groups of parameters for which further evidence was likely to be of most value were CV adverse event risks and all adverse event rates associated with the specific drugs celecoxib and ibuprofen. The value of perfect information remained high even when observational adverse event data were used. There is a very high value associated with obtaining further information on uncertain parameters for the economic evaluation of NSAIDs, COX-2 selective inhibitors and PPIs for people with OA. Obtaining further randomized or observational information on CV risks is likely to be particularly cost effective.

Exploring the Research Decision Space: The Expected Value of Information for Sequential Research Designs

To investigate the expected value of partial perfect information (EVPPI) and the research decisions it can address. Expected value of information (EVI) analysis assesses the expected gain in net benefit from further research. Where the expected value of perfect information (EVPI) exceeds the costs of additional research, EVPPI can be used to identify parameters that contribute most to the EVPI and parameters with no EVPPI that may be disregarded as targets for further research. Recently, it was noted that parameters with low EVPPI for a one-off may be associated with high EVPPI when considered as part of a sequential design. This article examines the characteristics and role of conditional and sequential EVPPI in EVI analysis. The calculation of EVPPI is demonstrated for single parameters, groups of parameters, and conditional and sequential EVPPI. Conditional EVPPI is the value of perfect information about one parameter, conditional on having obtained perfect information about another. Sequential EVPPI is the value of perfect information for a sequential research design to investigate first one parameter, then another. Conditional EVPPI differs from the individual EVPPI for a single parameter. Sequential EVPPI includes elements from the joint EVPPI for the parameters and the EVPPI for the first parameter in sequence. Sequential designs allow abandonment of research on the second parameter on the basis of additional information obtained on the first. The research decision space addressed by EVI analyses can be widened by incorporating sequential EVPPI to assess sequential research designs.