Discretely Integrated Condition Event (DICE) Simulation for Pharmacoeconomics

De Novo Cost-Effectiveness Model Framework for Nonalcoholic Steatohepatitis-Modeling Approach and Validation

BACKGROUND

Nonalcoholic steatohepatitis (NASH) is a chronic liver disease associated with hepatic morbidity and mortality and extra-hepatic comorbidities. Published NASH cost-effectiveness models (CEMs) are heterogeneous and consistently omit comorbid conditions that frequently co-exist alongside NASH. We aimed to develop a de novo CEM framework that incorporates extra-hepatic disease states and outcomes alongside hepatic components to enable future estimation of the cost-effectiveness of NASH interventions.

METHODS

Patient-level simulation and cohort-level Markov models were implemented in the same framework. Model inputs included fibrosis progression; late-stage liver disease outcomes; comorbidity outcomes for cardiovascular disease, type 2 diabetes, and obesity; mortality; health-related quality of life; and direct medical costs. The prototype analysis assessed the cost-effectiveness of obeticholic acid versus standard of care from a US payer perspective over a lifetime horizon with costs and effects discounted at 3% per annum. However, the CEM was designed for easy adaptation to other countries, time horizons, and other considerations. Efficacy and adverse event parameters were obtained from the 18-month interim analysis of the REGENERATE trial. Outputs include total and incremental costs, total life years, and quality-adjusted life years.

RESULTS

In this model, total costs, total life years, and quality-adjusted life years were all higher with obeticholic acid compared with standard of care. Cross-validation of this model with the 2016 and 2020 Institute for Clinical and Economic Review models revealed marked differences, mainly driven by mortality inputs, transition probability estimates, and incorporation of the effect of treatment and comorbidities.

CONCLUSION

This is the first CEM in NASH to incorporate the clinical consequences of several comorbidities. The flexible yet standardized framework permits estimation of the cost-effectiveness of NASH interventions in a variety of settings. The model currently includes several assumptions and will be further developed as more relevant data become available.

A Systematic Review of Cost-Effectiveness Studies of Newer Non-Insulin Antidiabetic Drugs: Trends in Decision-Analytical Models for Modelling of Type 2 Diabetes Mellitus

BACKGROUND

We performed a systematic overview of the cost-effectiveness analyses (CEAs) comparing Non-insulin antidiabetic drugs (NIADs) with other NIADs for the treatment of type 2 diabetes mellitus (T2DM), using decision-analytical modelling (DAM), focusing on both the economic results and the underlying methodological choices.

METHODS

Eligible studies were CEAs using DAM to compare NIADs within the glucagon-like peptide-1 (GLP1) receptor agonists, sodium-glucose cotransporter-2 (SGLT2) inhibitors, or dipeptidyl peptidase-4 (DPP4) inhibitor classes with other NIADs within those classes for the treatment of T2DM. The PubMed, Embase and Econlit databases were searched from 1 January 2018 to 15 November 2022. Two reviewers screened the studies for relevance by titles and abstracts and then for eligibility via full-text screening, extracted the data from the full texts and appendices, and then stored the data in a spreadsheet.

RESULTS

The search yielded 890 records and 50 studies were eligible for inclusion. The studies were mainly based on a European setting (60%). Industry sponsorship was found in 82% of studies. The CORE diabetes model was used in 48% of the studies. GLP1 and SGLT2 products were the main comparators in 31 and 16 studies, respectively, while one study had DPP4 and two had no easily discernible main comparator. Direct comparison between SGLT2 and GLP1 occurred in 19 studies. At a class level, SGLT2 dominated GLP1 in six studies and was cost effective against GLP1 once as part of a treatment pathway. GLP1 was cost effective in nine studies and not cost effective against SGLT2 in three studies. At a product level, oral and injectable semaglutide, and empagliflozin, were cost effective against other within-class products. Injectable and oral semaglutide were more frequently found cost effective in these comparisons, with some conflicting results. Most of the modelled cohorts and treatment effects were sourced from randomised controlled trials. The following model assumptions varied depending on the class of the main comparator: choice of and reasoning behind risk equations, the time until the treatment switch, and how often the comparators were discontinued. Diabetes-related complications were emphasised on par with quality-adjusted life-years as model outputs. The main quality issues were regarding the description of alternatives, the perspective of analysis, the measurement of costs and consequences, and patient subgroups.

CONCLUSION

The included CEAs using DAMs have limitations that hinder their ability to inform decision makers on the cost-effective choice: lack of updated reasoning behind the choice of key model assumptions, over-reliance on risk equations based on older treatment practices, and sponsorship bias. The question of which NIAD is cost effective for the treatment of which T2DM patient is a pressing one and the answer remains unclear.

Cost-Effectiveness Analysis of HPV Extended versus Partial Genotyping for Cervical Cancer Screening in Singapore

Human papillomavirus (HPV) partial genotyping (PGT) identifies HPV16 and HPV18 individually, alongside 12 other high-risk HPV genotypes (hrHPV) collectively. HPV extended genotyping (XGT) identifies four additional hrHPV individually (HPV31, 45, 51, and 52), and reports the remaining eight in three groups (HPV33|58; 56|59|66; 35|39|68). Quality-adjusted life years (QALY), health care resource use, and costs of XGT were compared to PGT for cervical cancer screening in Singapore using DICE simulation. Women with one of the three hrHPV identified by XGT (HPV35|39|68; 56|59|66; 51), and atypical squamous cells of undetermined significance (ASCUS) on cytology, are recalled for a repeat screening in one year, instead of undergoing an immediate colposcopy with PGT. At the repeat screening, the colposcopy is performed only for persistent same-genotype infections in XGT, while with PGT, all the women with persistent HPV have a colposcopy. Screening 500,122 women, aged 30-69, with XGT, provided an incremental cost-effectiveness ratio (ICER) versus PGT of SGD 16,370/QALY, with 7130 (19.4%) fewer colposcopies, 6027 (7.0%) fewer cytology tests, 9787 (1.6%) fewer clinic consultations, yet 2446 (0.5%) more HPV tests. The XGT ICER remains well below SGD 100,000 in sensitivity analyses, (-SGD 17,736/QALY to SGD 50,474/QALY). XGT is cost-effective compared to PGT, utilizes fewer resources, and provides a risk-based approach as the primary cervical cancer screening method.

Cost-effectiveness of second-line empagliflozin versus liraglutide for type 2 diabetes in the United States

AIM

To estimate the cost-effectiveness of sequential use of the sodium-glucose co-transporter-2 inhibitor empagliflozin and glucagon-like peptide-1 receptor agonist liraglutide after metformin in patients with type 2 diabetes (T2D) from the US payer perspective.

MATERIALS AND METHODS

An economic simulation model with a lifetime horizon was developed to estimate T2D-related complications (including cardiovascular [CV] death, myocardial infarction, stroke, and renal outcomes) using EMPA-REG OUTCOME data or UK Prospective Diabetes Study risk equations, in patients with or without a history of cardiovascular disease (CVD), respectively. Evidence synthesis methods were used to provide effectiveness inputs for empagliflozin and liraglutide. Population characteristics, adverse event rates, treatment escalation, costs ($2019), and utilities (both discounted 3%/year) were taken from US sources.

RESULTS

Compared with second-line liraglutide in the overall T2D population, second-line empagliflozin was dominant as it was associated with lower total lifetime cost ($11 244/patient less) and resulted in a quality-adjusted life-year (QALY) gain (0.32/patient). Second-line empagliflozin was associated with reductions in CV death (by 5%) and lower cumulative complication rates in patients with CVD (by 2%), relative to second-line liraglutide. These findings were consistent among patients with co-morbid CVD, with gains in incremental QALYs (0.43/patient) and lower lifetime cost (by $10 175/patient) relative to second-line liraglutide. Scenario analyses consistently showed dominance for second-line empagliflozin.

CONCLUSION

For patients with T2D, use of second-line empagliflozin combined with metformin was a dominant strategy for US payers, associated with extended survival, improved QALYs, and lower costs compared with second-line liraglutide.

Predicting Hospital Resource Use During COVID-19 Surges: A Simple but Flexible Discretely Integrated Condition Event Simulation of Individual Patient-Hospital Trajectories

OBJECTIVES

To assist with planning hospital resources, including critical care (CC) beds, for managing patients with COVID-19.

METHODS

An individual simulation was implemented in Microsoft Excel using a discretely integrated condition event simulation. Expected daily cases presented to the emergency department were modeled in terms of transitions to and from ward and CC and to discharge or death. The duration of stay in each location was selected from trajectory-specific distributions. Daily ward and CC bed occupancy and the number of discharges according to care needs were forecast for the period of interest. Face validity was ascertained by local experts and, for the case study, by comparing forecasts with actual data.

RESULTS

To illustrate the use of the model, a case study was developed for Guy's and St Thomas' Trust. They provided inputs for January 2020 to early April 2020, and local observed case numbers were fit to provide estimates of emergency department arrivals. A peak demand of 467 ward and 135 CC beds was forecast, with diminishing numbers through July. The model tended to predict higher occupancy in Level 1 than what was eventually observed, but the timing of peaks was quite close, especially for CC, where the model predicted at least 120 beds would be occupied from April 9, 2020, to April 17, 2020, compared with April 7, 2020, to April 19, 2020, in reality. The care needs on discharge varied greatly from day to day.

CONCLUSIONS

The DICE simulation of hospital trajectories of patients with COVID-19 provides forecasts of resources needed with only a few local inputs. This should help planners understand their expected resource needs.

Evaluating Montréal's harm reduction interventions for people who inject drugs: protocol for observational study and cost-effectiveness analysis

INTRODUCTION

The main harm reduction interventions for people who inject drugs (PWID) are supervised injection facilities, needle and syringe programmes and opioid agonist treatment. Current evidence supporting their implementation and operation underestimates their usefulness by excluding skin, soft tissue and vascular infections (SSTVIs) and anoxic/toxicity-related brain injury from cost-effectiveness analyses (CEA). Our goal is to conduct a comprehensive CEA of harm reduction interventions in a setting with a large, dispersed, heterogeneous population of PWID, and include prevention of SSTVIs and anoxic/toxicity-related brain injury as measures of benefit in addition to HIV, hepatitis C and overdose morbidity and mortalities averted.

METHODS AND ANALYSIS

This protocol describes how we will develop an open, retrospective cohort of adult PWID living in Québec between 1 January 2009 and 31 December 2020 using administrative health record data. By complementing this data with non-linkable paramedic dispatch records, regional monthly needle and syringe dispensation counts and repeated cross-sectional biobehavioural surveys, we will estimate the hazards of occurrence and the impact of Montréal's harm reduction interventions on the incidence of drug-use-related injuries, infections and deaths. We will synthesise results from our empirical analyses with published evidence to simulate infections and injuries in a hypothetical population of PWID in Montréal under different intervention scenarios including current levels of use and scale-up, and assess the cost-effectiveness of each intervention from the public healthcare payer's perspective.

ETHICS AND DISSEMINATION

This study was approved by McGill University's Institutional Review Board (Study Number: A08-E53-19B). We will work with community partners to disseminate results to the public and scientific community via scientific conferences, a publicly accessible report, op-ed articles and open access peer-reviewed journals.

Overview and Use of Tools for Selecting Modelling Techniques in Health Economic Studies

The availability and use of tools to guide the choice of modelling technique are not well understood. Our study aims to review existing tools and explore the use of those tools in health economic models. Two reviews and one case study were conducted. Review 1 aimed to identify tools based on expert opinion and citation searching and explore the value of the tools for health economic models. Review 2, based on citation searching, aimed to describe how those tools have been used in health economic models. Both reviews were conducted using Web of Science and Scopus. Two independent reviewers selected studies for inclusion. A case study, focused on economic evaluations of antipsychotic medication in schizophrenia, was conducted to compare the modelling techniques used by existing models with modelling techniques recommended by identified tools. Seven tools were identified, of which the revised Brennan's toolkit, was assessed to be the most appropriate for health economic models. The seven tools were cited 126 times in publications reporting health economic models. Only 17 of these (13.5%) reported that they used the tool(s) to guide the choice of modelling technique. Application of these tools suggested discrete event simulation is most appropriate for modelling antipsychotic medication in schizophrenia, but discrete event simulation was only used by 17% of existing models. There is considerable inconsistency between the modelling techniques used by existing models and modelling techniques recommended by tools. It is recommended that for future modelling studies the choice of modelling technique should be justified, this can be achieved by the application of model selection tools, such as the revised Brennan's toolkit. Future research is required to explore the barriers to using model selection tools in health economic models and to update existing tools and make them easier to use.

Economic evaluation of betibeglogene autotemcel (Beti-cel) gene addition therapy in transfusion-dependent β-thalassemia

Background: Standard of care (SoC) for transfusion-dependent β-thalassemia (TDT) requires lifelong, regular blood transfusions as well as chelation to reduce iron accumulation. Objective: This study investigates the cost-effectiveness of betibeglogene autotemcel ('beti-cel'; LentiGlobin for β-thalassemia) one-time, gene addition therapy compared to lifelong SoC for TDT. Study design: Microsimulation model simulated the lifetime course of TDT based on a causal sequence in which transfusion requirements determine tissue iron levels, which in turn determine risk of iron overload complications that increase mortality. Clinical trial data informed beti-cel clinical parameters; effects of SoC on iron levels came from real-world studies; iron overload complication rates and mortality were based on published literature. Setting: USA; commercial payer perspective Participants: TDT patients age 2-50 Interventions: Beti-cel is compared to SoC. Main outcome measure: Incremental cost-effectiveness ratio (ICER) utilizing quality-adjusted life-years (QALYs) Results: The model predicts beti-cel adds 3.8 discounted life years (LYs) or 6.9 QALYs versus SoC. Discounted lifetime costs were $2.28 M for beti-cel ($572,107 if excluding beti-cel cost) and $2.04 M for SoC, with a resulting ICER of $34,833 per QALY gained. Conclusion: Beti-cel is cost-effective for TDT patients compared to SoC. This is due to longer survival and cost offset of lifelong SoC.

Modeling long-term health and economic implications of new treatment strategies for Parkinson's disease: an individual patient simulation study

Background: Simulation modeling facilitates the estimation of long-term health and economic outcomes to inform healthcare decision-making. Objective: To develop a framework to simulate progression of Parkinson's disease (PD), capturing motor and non-motor symptoms, clinical outcomes, and associated costs over a lifetime. Methods: A patient-level simulation was implemented accounting for individual variability and interrelated changes in common disease progression scales. Predictive equations were developed to model progression for newly diagnosed patients and were combined with additional sources to inform long-term progression. Analyses compared a hypothetical disease-modifying therapy (DMT) with a standard of care to explore the drivers of cost-effectiveness. Results: The equations captured the dependence between the various measures, leveraging prior values and rates of change to obtain realistic predictions. The simulation was built upon several interrelated equations, validated by comparison with observed values for the Movement Disorder Society Unified PD Rating Scale (MDS-UPDRS) and UPDRS subscales over time. In a case study, disease progression rates, patient utilities, and direct non-medical costs were drivers of cost-effectiveness. Conclusions: The developed equations supported the simulation of early PD. This model can support conducting simulations to inform internal decision-making, trial design, and strategic planning early in the development of new DMTs entering clinical trials.

Comparison of Decision Modeling Approaches for Health Technology and Policy Evaluation

We discuss tradeoffs and errors associated with approaches to modeling health economic decisions. Through an application in pharmacogenomic (PGx) testing to guide drug selection for individuals with a genetic variant, we assessed model accuracy, optimal decisions, and computation time for an identical decision scenario modeled 4 ways: using 1) coupled-time differential equations (DEQ), 2) a cohort-based discrete-time state transition model (MARKOV), 3) an individual discrete-time state transition microsimulation model (MICROSIM), and 4) discrete event simulation (DES). Relative to DEQ, the net monetary benefit for PGx testing (v. a reference strategy of no testing) based on MARKOV with rate-to-probability conversions using commonly used formulas resulted in different optimal decisions. MARKOV was nearly identical to DEQ when transition probabilities were embedded using a transition intensity matrix. Among stochastic models, DES model outputs converged to DEQ with substantially fewer simulated patients (1 million) v. MICROSIM (1 billion). Overall, properly embedded Markov models provided the most favorable mix of accuracy and runtime but introduced additional complexity for calculating cost and quality-adjusted life year outcomes due to the inclusion of "jumpover" states after proper embedding of transition probabilities. Among stochastic models, DES offered the most favorable mix of accuracy, reliability, and speed.

Cost-effectiveness analysis of empagliflozin versus sitagliptin as second-line therapy for treatment in patients with type 2 diabetes in the United States

AIM

To estimate the cost-effectiveness of sequential addition of empagliflozin versus sitagliptin after metformin in patients with type 2 diabetes (T2D) with or without cardiovascular disease (CVD) from the perspective of the US healthcare payer.

METHODS

An individual simulation model predicted lifetime diabetes-related complications, using UKPDS-OM2 equations in patients without CVD, and EMPA-REG OUTCOME equations in patients with CVD. Additional US-based sources informed inputs for population characteristics, adverse events, non-CV death, treatment escalation, quality of life and costs. Costs and quality-adjusted life-years (QALYs) were discounted 3.0% annually.

RESULTS

The incremental cost-effectiveness ratio (ICER) for second-line empagliflozin versus sitagliptin in the overall T2D population was $6967/QALY. Empagliflozin led to longer CVD-free survival (0.07 years) and an 11% reduction in CV death in patients with CVD compared with sitagliptin. Empagliflozin resulted in greater benefits with greater costs in patients with versus without baseline CVD, yielding ICERs of $3589/QALY versus $12 577/QALY, respectively. Results were consistent across a range of deterministic and probabilistic sensitivity analyses and scenarios.

CONCLUSION

Compared with sitagliptin, empagliflozin was cost-effective (at $50 000/QALY US threshold) as a second-line treatment to metformin for T2D patients with or without CVD in the United States. Our findings lend additional support for more widespread adoption of guidelines by healthcare decision-makers for T2D treatment.

Cost-effectiveness of valbenazine compared with deutetrabenazine for the treatment of tardive dyskinesia

AIMS

To evaluate clinical and economic outcomes associated with valbenazine compared with deutetrabenazine in patients with tardive dyskinesia (TD) using a model that accounts for multiple dimensions of patient health status.

MATERIALS AND METHODS

A discretely integrated condition event model was developed to evaluate the cost-effectiveness of treatment with valbenazine and deutetrabenazine in a synthetic cohort of 1,000 patients with TD who were receiving antipsychotic medication to treat an underlying psychiatric disorder. Clinical inputs were derived from relevant clinical trials or from publicly available sources. Patients were assessed over 1 year using ≥50% improvement from baseline in Abnormal Involuntary Movement Scale (AIMS) total score as the primary definition of response. Response at 1 year using Clinical Global Impression of Change (CGIC) score ≤2 was also assessed. Health outcomes included quality-adjusted life years (QALYs), life years, proportion responding to treatment at 1 year, and number of psychiatric relapses.

RESULTS

Regardless of the definition used for response, patients treated with valbenazine were more likely to have responded to treatment at 1 year, lived longer, and accrued more QALYs than patients who received deutetrabenazine. Using the AIMS response criterion, the incremental cost-effectiveness ratio was $9,951/QALY for valbenazine compared with deutetrabenazine. By comparison, using the CGIC response criterion valbenazine dominated deutetrabenazine with valbenazine-treated patients accumulating more QALYs (3.4 vs 3.3 years) and incurring lower lifetime costs ($252,311 vs $283,208) than deutetrabenazine-treated patients.

LIMITATIONS

There are no head-to-head trials of valbenazine and deutetrabenazine, so probabilities of response used in the model were calculated based on an indirect treatment comparison of results from individual trials with one drug or the other, using only those metrics reported across trials.

CONCLUSIONS

In patients with TD, treatment with valbenazine is highly cost-effective compared with deutetrabenazine.

Using Discretely Integrated Condition Event Simulation To Construct Quantitative Benefit-Risk Models: The Example of Rotavirus Vaccination in France

PURPOSE

Although quantitative benefit-risk models (qBRms) are indisputably valuable tools for gaining comprehensive assessments of health care interventions, they are not systematically used, probably because they lack an integrated framework that provides methodologic structure and harmonization. An alternative that allows all stakeholders to design operational models starting from a standardized framework was recently developed: the discretely integrated condition event (DICE) simulation. The aim of the present work was to assess the feasibility of implementing a qBRm in DICE, using the example of rotavirus vaccination.

METHODS

A model of rotavirus vaccination was designed using DICE and implemented in spreadsheet software with 3 worksheets: Conditions, Events, and Outputs. Conditions held the information in the model; this information changed at Events, and Outputs were special Conditions that stored the results collected during the analysis. A hypothetical French birth cohort was simulated for the assessment of rotavirus vaccination over time. The benefits were estimated for up to 5 years, and the risks in the 7 days following rotavirus vaccination versus no vaccination were assessed, with the results expressed as benefit-risk ratios.

FINDINGS

This qBRm model required 8 Events, 38 Conditions, and 9 Outputs. Two Events cyclically updated the rates of rotavirus gastroenteritis (RVGE) and intussusception (IS) according to age. Vaccination occurred at 2 additional Events, according to the vaccination scheme applied in France, and affected the occurrence of the other Events. Outputs were the numbers of hospitalizations related to RVGE and to IS, and related deaths. The entire model was specified in a small set of tables contained in a 445-KB electronic workbook. Analyses showed that for each IS-related hospitalization or death caused, 1613 (95% credible interval, 1001-2800) RVGE-related hospitalizations and 787 (95% credible interval, 246-2691) RVGE-related deaths would be prevented by vaccination. These results are consistent with those from a published French study using similar inputs but a very different modeling approach.

IMPLICATIONS

A limitation of the DICE approach was the extended run time needed for completing the sensitivity analyses when implemented in the electronic worksheets. DICE provided a user-friendly integrated framework for developing qBRms and should be considered in the development of structured approaches to facilitate benefit-risk assessment.

Cabozantinib and vandetanib for unresectable locally advanced or metastatic medullary thyroid cancer: a systematic review and economic model

BACKGROUND

Medullary thyroid cancer (MTC) is a rare form of cancer that affects patients' health-related quality of life (HRQoL) and survival. Cabozantinib (Cometriq^®; Ipsen, Paris, France) and vandetanib (Caprelsa^®; Sanofi Genzyme, Cambridge, MA, USA) are currently the treatment modality of choice for treating unresectable progressive and symptomatic MTC.

OBJECTIVES

(1) To evaluate the clinical effectiveness and safety of cabozantinib and vandetanib. (2) To estimate the incremental cost-effectiveness of cabozantinib and vandetanib versus each other and best supportive care. (3) To identify key areas for primary research. (4) To estimate the overall cost of these treatments in England.

DATA SOURCES

Peer-reviewed publications (searched from inception to November 2016), European Public Assessment Reports and manufacturers' submissions.

REVIEW METHODS

A systematic review [including a network meta-analysis (NMA)] was conducted to evaluate the clinical effectiveness and safety of cabozantinib and vandetanib. The economic analysis included a review of existing analyses and the development of a de novo model.

RESULTS

The systematic review identified two placebo-controlled trials. The Efficacy of XL184 (Cabozantinib) in Advanced Medullary Thyroid Cancer (EXAM) trial evaluated the efficacy and safety of cabozantinib in patients with unresectable locally advanced, metastatic and progressive MTC. The ZETA trial evaluated the efficacy and safety of vandetanib in patients with unresectable locally advanced or metastatic MTC. Both drugs significantly improved progression-free survival (PFS) more than the placebo (p < 0.001). The NMA suggested that, within the symptomatic and progressive MTC population, the effects on PFS were similar (vandetanib vs. cabozantinib: hazard ratio 1.14, 95% credible interval 0.41 to 3.09). Neither trial demonstrated a significant overall survival benefit for cabozantinib or vandetanib versus placebo, although data from ZETA were subject to potential confounding. Both cabozantinib and vandetanib demonstrated significantly better objective response rates and calcitonin (CTN) and carcinoembryonic antigen (CEA) response rates than placebo. Both cabozantinib and vandetanib produced frequent adverse events, often leading to dose interruption or reduction. The assessment group model indicates that, within the EU-label population (symptomatic and progressive MTC), the incremental cost-effectiveness ratios (ICERs) for cabozantinib and vandetanib are > £138,000 per quality-adjusted life-year (QALY) gained. Within the restricted EU-label population (symptomatic and progressive MTC with CEA/CTN doubling times of ≤ 24 months), the ICER for vandetanib is expected to be > £66,000 per QALY gained. The maximum annual budget impact within the symptomatic and progressive population is estimated to be ≈£2.35M for cabozantinib and ≈£5.53M for vandetanib. The costs of vandetanib in the restricted EU-label population are expected to be lower.

LIMITATIONS

The intention-to-treat populations of the EXAM and ZETA trials are notably different. The analyses of ZETA subgroups may be subject to confounding as a result of differences in baseline characteristics and open-label vandetanib use. Attempts to statistically adjust for treatment switching were unsuccessful. No HRQoL evidence was identified for the MTC population.

CONCLUSIONS

The identified trials suggest that cabozantinib and vandetanib improve PFS more than the placebo; however, significant OS benefits were not demonstrated. The economic analyses indicate that within the EU-label population, the ICERs for cabozantinib and vandetanib are > £138,000 per QALY gained. Within the restricted EU-label population, the ICER for vandetanib is expected to be > £66,000 per QALY gained.

FUTURE RESEARCH PRIORITIES

(1) Primary research assessing the long-term effectiveness of cabozantinib and vandetanib within relevant subgroups. (2) Reanalyses of the ZETA trial to investigate the impact of adjusting for open-label vandetanib use using appropriate statistical methods. (3) Studies assessing the impact of MTC on HRQoL.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42016050403.

FUNDING

The National Institute for Health Research Health Technology Assessment programme.

Leveraging DICE (Discretely-Integrated Condition Event) Simulation to Simplify the Design and Implementation of Hybrid Models

OBJECTIVES

Using an example of an existing model constructed by the National Institute for Health and Care Excellence (NICE) to inform a real health technology assessment, this study seeks to demonstrate how a discretely integrated condition event (DICE) simulation can improve the implementation of Markov models.

METHODS

Using the technical report and spreadsheet, the original model was translated to a standard DICE simulation without making any changes to the design. All original analyses were repeated and the results were compared. Aspects that could have improved the original design were then considered.

RESULTS

The original model consisted of 32 copies (8 risk strata × 4 treatments) of the Markov structure, containing more than 6000 Microsoft Excel® formulas (18 MB files). Three aspects (nonadherence, scheduled treatment stop, and end of fracture risk) were handled by incorporating weighted averages into the cycle-specific calculations. The DICE implementation used 3 conditions to represent the states and a single transition event to apply the probabilities; 3 additional events processed the special aspects, and profiles handled the 8 strata (0.12 MB file). One replication took 16 seconds. The original results were reproduced but extensive additional sensitivity analyses, including structural analyses, were enabled.

CONCLUSION

Implementing a real Markov model using DICE simulation both preserves the advantages of the approach and expands the available tools, improving transparency and ease of use and review.

Economic Evaluation of Sequences of Biological Treatments for Patients With Moderate-to-Severe Rheumatoid Arthritis and Inadequate Response or Intolerance to Methotrexate in France

OBJECTIVES

Biologic disease-modifying antirheumatic drugs (bDMARDs) are prescribed sequentially in the treatment of rheumatoid arthritis (RA). Healthcare decision makers continue to debate their use, mainly because of their high costs. Our aim was to perform an economic evaluation for France of bDMARD sequences for treatment of moderate-to-severe RA after inadequate response or intolerance to conventional DMARDs (eg, methotrexate).

METHODS

A discretely integrated condition event simulation was developed to track the course of patients from first bDMARD through switches to further lines in a sequence. The model included 11 events, 91 conditions, and 21 controlling equations. Inputs were obtained from a meta-analysis of clinical trials, a French registry, national drug lists, and databases. Survival, time with minimal activity, quality-adjusted life-years (QALYs), and total costs were output. Structural and probabilistic sensitivity analyses were conducted.

RESULTS

Sequences starting with etanercept biosimilars (ETB) cost less, with ETB-abatacept-infliximab the least expensive: the mean lifetime discounted total cost was €116 912 per patient, with a mean of 11.166 QALYs. Most other strategies were dominated or led to small QALY gains (0.0008-0.0329). Only ETB-tocilizumab-abatacept made it onto the efficiency frontier, but at €955 778 per QALY gained. These results were confirmed in several scenarios and uncertainty analyses.

CONCLUSION

Given minor differences in QALYs gained between bDMARD sequences with large cost differences, starting with biosimilars was more efficient than starting with branded products. Our model and findings should provide French and other decision makers with useful tools to address the challenges of comparing sequences of treatments for RA.

A Multidimensional Array Representation of State-Transition Model Dynamics

Cost-effectiveness analyses often rely on cohort state-transition models (cSTMs). The cohort trace is the primary outcome of cSTMs, which captures the proportion of the cohort in each health state over time (state occupancy). However, the cohort trace is an aggregated measure that does not capture information about the specific transitions among health states (transition dynamics). In practice, these transition dynamics are crucial in many applications, such as incorporating transition rewards or computing various epidemiological outcomes that could be used for model calibration and validation (e.g., disease incidence and lifetime risk). In this article, we propose an alternative approach to compute and store cSTMs outcomes that capture both state occupancy and transition dynamics. This approach produces a multidimensional array from which both the state occupancy and the transition dynamics can be recovered. We highlight the advantages of the multidimensional array over the traditional cohort trace and provide potential applications of the proposed approach with an example coded in R to facilitate the implementation of our method.

Cooking Up a Transparent Model Following a DICE Recipe

When a model is built, it tends to be an iterative process. The ideal model building process starts with constructing a top-level model (that works) and then digging deeper into the sections that need more details. A good model tends to have a long life and be both adapted and updated. A perfectly transparent first model (in whatever platform or programming language it might be) therefore tends to get increasingly convoluted. Yet, it still needs to be submitted and reviewed since it will likely be used as support to facilitate possibly multi-million Euro decisions. So, any modelling method should be able to accept substantial changes and updates of the model without compromising (any hopefully existing) transparency. One method that is gaining traction has its heritage in one of the oldest types of structured algorithms we have-the cooking recipe. Regardless of whether you are looking at a recipe for Swedish meatballs or a recipe to prepare the poisonous Fugu fish, the structure of the recipe is the same. It begins with a list of ingredients with the cooking steps to follow. Combining the 'recipe' idea with a common software like Excel^® (which tends to be available on most computers), and the need to 'cook' all kinds of different model types (Markov, micro-simulation, discrete-event simulation, partitioned survival), leads to the DICE method (Caro in Pharmacoeconomics 34(7):655-672, 2016). It starts by listing the initial take on the ingredients needed for the model (in DICE called 'conditions'). Just like writing a recipe, it is easy to add ingredients to this list later as needed. Then the actual 'cooking' is specified as a list of the events needed to fully represent the disease and interventions that will take place, followed by the instructions for each event. Using this approach, it becomes as easy to look at different models as it is to look at different recipes. The simplicity, flexibility and standardized structure, but especially its transparency, are the main features of DICE.

Treatment sequences for advanced renal cell carcinoma: A health economic assessment

Objective

Advanced renal cell carcinoma (RCC) is commonly treated with vascular endothelial growth factor or mammalian target of rapamycin inhibitors. As new therapies emerge, interest grows in gaining a deeper understanding of treatment sequences. Recently, we developed a patient-level, discretely integrated condition event (DICE) simulation to estimate survival and lifetime costs for various cancer therapies, using a US payer perspective. Using this model, we explored the impact of treatments such as nivolumab and cabozantinib, and compared the clinical outcomes and cost consequences of commonly used treatment algorithms for patients with advanced RCC.

Methods

Included treatment sequences were pazopanib or sunitinib as first-line treatment, followed by nivolumab, cabozantinib, axitinib, pazopanib or everolimus. Efficacy inputs were derived from the CheckMate 025 trial and a network meta-analysis based on available literature. Safety and cost data were obtained from publicly available sources or literature.

Results

Based on our analysis, the average cost per life-year (LY) was lowest for sequences including nivolumab (sunitinib → nivolumab, $75,268/LY; pazopanib → nivolumab, $84,459/LY) versus axitinib, pazopanib, everolimus and cabozantinib as second-line treatments. Incremental costs per LY gained were $49,592, $73,927 and $30,534 for nivolumab versus axitinib, pazopanib and everolimus-containing sequences, respectively. The model suggests that nivolumab offers marginally higher life expectancy at a lower cost versus cabozantinib-including sequences.

Conclusion

Treatment sequences using nivolumab in the second-line setting are less costly compared with sequential use of targeted agents. In addition to efficacy and safety data, cost considerations may be taken into account when considering treatment algorithms for patients with advanced RCC.

Outcomes and costs of remote patient monitoring among patients with implanted cardiac defibrillators: An economic model based on the PREDICT RM database

Background

Remote monitoring of implantable cardioverter‐defibrillators has been associated with reduced rates of all‐cause rehospitalizations and mortality among device recipients, but long‐term economic benefits have not been studied.

Methods and Results

An economic model was developed using the PREDICT RM database comparing outcomes with and without remote monitoring. The database included patients ages 65 to 89 who received a Boston Scientific device from 2006 to 2010. Parametric survival equations were derived for rehospitalization and mortality to predict outcomes over a maximum time horizon of 25 years. The analysis assessed rehospitalization, mortality, and the cost‐effectiveness (expressed as the incremental cost per quality‐adjusted life year) of remote monitoring versus no remote monitoring. Remote monitoring was associated with reduced mortality; average life expectancy and average quality‐adjusted life years increased by 0.77 years and 0.64, respectively (6.85 life years and 5.65 quality‐adjusted life years). When expressed per patient‐year, remote monitoring patients had fewer subsequent rehospitalizations (by 0.08 per patient‐year) and lower hospitalization costs (by $554 per patient year). With longer life expectancies, remote monitoring patients experienced an average of 0.64 additional subsequent rehospitalizations with increased average lifetime hospitalization costs of $2784. Total costs of outpatient and physician claims were higher with remote monitoring ($47 515 vs $42 792), but average per patient‐year costs were lower ($6232 vs $6244). The base‐case incremental cost‐effectiveness ratio was $10 752 per quality‐adjusted life year, making remote monitoring high‐value care.

Conclusion

Remote monitoring is a cost‐effective approach for the lifetime management of patients with implantable cardioverter‐defibrillators.

Modeling Approaches in Cost-Effectiveness Analysis of Disease-Modifying Therapies for Relapsing-Remitting Multiple Sclerosis: An Updated Systematic Review and Recommendations for Future Economic Evaluations

BACKGROUND

Numerous cost-effectiveness analyses (CEAs) of disease-modifying therapies (DMTs) for relapsing-remitting multiple sclerosis (RRMS) have been published in the last three decades. Literature reviews of the modeling methods and results from these CEAs have also been published. The last literature review that focused on modeling methods, without country or time horizon in the inclusion criteria, included studies published up to 2012. Since then, new DMTs have become available, and new models and data sources have been used to assess their cost effectiveness.

OBJECTIVE

The aim of this systematic review was to provide a detailed and comprehensive description of the relevant aspects of economic models used in CEAs of DMTs for RRMS, to understand how these models have progressed from recommendations provided in past reviews, what new approaches have been developed, what issues remain, and how they could be addressed.

METHODS

EMBASE, MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL), the National Health System (NHS) Economic Evaluations Database, the Health Technology Assessment (HTA) Database, and EconLit were searched for cost-effectiveness studies of DMTs for RRMS that used decision-analytic models, published in English between 1 January 2012 and 24 December 2017. The inclusion criteria were as follows: being a full economic evaluation, a decision-analytic model was used, the target population concerned adult patients with RRMS, and being available in full-text format. Studies were not excluded based on the methodological quality. The background information of the included studies, as well as specific information on the components of the economic models related to the areas of recommendation from previous reviews were extracted.

RESULTS

Twenty-three studies from ten countries were included. The model structure of these studies has converged over time, characterizing the course of disease progression in terms of changes in disability and the occurrence of relapses over time. Variations were found in model approach; data sources for the natural course of the disease and comparative efficacy between DMTs; number of lines of treatment modeled; long-term efficacy waning and treatment discontinuation assumptions; type of withdrawal; and criteria for selecting adverse events. Main areas for improvement include using long-term time horizons and societal perspective; reporting relevant health outcomes; conducting scenario analyses using different sources of natural history and utility values; and reporting how the model was validated.

CONCLUSION

The structure of economic models used in CEAs of DMTs for RRMS has converged over time. However, variation remains in terms of model approach, inputs, and assumptions. Though some recommendations from previous reviews have been incorporated in later models, areas for improvement remain.

Smoking Cessation: A Comparison of Two Model Structures

BACKGROUND

Most economic evaluations of smoking cessation interventions have used cohort state-transition models. Discrete event simulations (DESs) have been proposed as a superior approach.

OBJECTIVE

We developed a state-transition model and a DES using the discretely integrated condition event (DICE) framework and compared the cost-effectiveness results. We performed scenario analysis using the DES to explore the impact of alternative assumptions.

METHODS

The models estimated the costs and quality-adjusted life years (QALYs) for the intervention and comparator from the perspective of the UK National Health Service and Personal Social Services over a lifetime horizon. The models considered five comorbidities: chronic obstructive pulmonary disease, myocardial infarction, coronary heart disease, stroke and lung cancer. The state-transition model used prevalence data, and the DES used incidence. The costs and utility inputs were the same between two models and consistent with those used in previous analyses for the National Institute for Health and Care Excellence.

RESULTS

In the state-transition model, the intervention produced an additional 0.16 QALYs at a cost of £540, leading to an incremental cost-effectiveness ratio (ICER) of £3438. The comparable DES scenario produced an ICER of £5577. The ICER for the DES increased to £18,354 when long-term relapse was included.

CONCLUSIONS

The model structures themselves did not influence smoking cessation cost-effectiveness results, but long-term assumptions did. When there is variation in long-term predictions between interventions, economic models need a structure that can reflect this.

Application of dynamic modeling for survival estimation in advanced renal cell carcinoma

Objective

In oncology, extrapolation of clinical outcomes beyond trial duration is traditionally achieved by parametric survival analysis using population-level outcomes. This approach may not fully capture the benefit/risk profile of immunotherapies due to their unique mechanisms of action. We evaluated an alternative approach—dynamic modeling—to predict outcomes in patients with advanced renal cell carcinoma. We compared standard parametric fitting and dynamic modeling for survival estimation of nivolumab and everolimus using data from the phase III CheckMate 025 study.

Methods

We developed two statistical approaches to predict longer-term outcomes (progression, treatment discontinuation, and survival) for nivolumab and everolimus, then compared these predictions against follow-up clinical trial data to assess their proximity to observed outcomes. For the parametric survival analyses, we selected a probability distribution based on its fit to observed population-level outcomes at 14-month minimum follow-up and used it to predict longer-term outcomes. For dynamic modeling, we used a multivariate Cox regression based on patient-level data, which included risk scores, and probability and duration of response as predictors of longer-term outcomes. Both sets of predictions were compared against trial data with 26- and 38-month minimum follow-up.

Results

Both statistical approaches led to comparable fits to observed trial data for median progression, discontinuation, and survival. However, beyond the trial duration, mean survival predictions differed substantially between methods for nivolumab (30.8 and 51.5 months), but not everolimus (27.2 and 29.8 months). Longer-term follow-up data from CheckMate 025 and phase I/II studies resembled dynamic model predictions for nivolumab.

Conclusions

Dynamic modeling can be a good alternative to parametric survival fitting for immunotherapies because it may help better capture the longer-term benefit/risk profile and support health-economic evaluations of immunotherapies.

The Economic Impact of New Therapeutic Interventions on Neuropsychiatric Inventory (NPI) Symptom Scores in Patients with Alzheimer Disease

Background/Aims

Few studies have modeled individual Neuropsychiatric Inventory (NPI) symptom scores for Alzheimer disease (AD) patients and assessed the value of therapeutic interventions that can potentially impact them. The main objective of this study was to evaluate the impact of new AD symptomatic treatments on relevant health economic outcomes via their potential effects on cognition and neuropsychiatric symptoms such as depression, irritability, anxiety, and sleep disorder.

Methods

We enhanced the previously published AHEAD model (Assessment of Health Economics in Alzheimer's Disease) by including new variables and functional relations to capture the NPI's individual neuropsychiatric symptoms in addition to the total NPI score. This update allowed us to study the longitudinal effect of improvements in specific NPI subscale scores and the downstream impact on outcomes such as psychiatric medication use, survival, and institutional placement.

Results

The model base-case results showed that a hypothetical treatment with symptomatic effects on anxiety, depression, and irritability NPI subscales was not cost-effective; however, the treatment's cost-effectiveness was improved once a direct link between NPI subscales and mortality was explored or under relatively stronger treatment effects.

Conclusion

Treatments that influence specific symptoms within the overall NPI have the potential to improve patient outcomes in a cost-effective way. This model is a useful tool for evaluating target product profiles of drugs with effect on NPI symptoms in early stages of development.

Alzheimer's disease Archimedes condition-event simulator: Development and validation

Introduction

Several advances have been made in Alzheimer's Disease (AD) modeling, however, there remains a need for a simulator that represents the full scope of disease progression and can be used to study new disease-modifying treatments for early-stage and even prodromal AD.

Methods

We developed AD Archimedes condition-event simulator, a patient-level simulator with a focus on simulating the effects of early interventions through changes in biomarkers of AD. The simulator incorporates interconnected predictive equations derived from longitudinal data sets.

Results

The results of external validations on AD Archimedes condition-event simulator showed that it provides reasonable estimates once compared to literature results on transition to dementia AD, institutionalization, and mortality. A case study comparing a disease-modifying treatment and a symptomatic treatment also showcases the benefits of early treatment.

Discussion

The AD Archimedes condition-event simulator is designed to perform economic evaluation on various interventions through close tracking of disease progression and the related clinical outcomes.

Implementing parallel spreadsheet models for health policy decisions: The impact of unintentional errors on model projections

Background

Spreadsheet software is increasingly used to implement systems science models informing health policy decisions, both in academia and in practice where technical capacity may be limited. However, spreadsheet models are prone to unintentional errors that may not always be identified using standard error-checking techniques. Our objective was to illustrate, through a methodologic case study analysis, the impact of unintentional errors on model projections by implementing parallel model versions.

Methods

We leveraged a real-world need to revise an existing spreadsheet model designed to inform HIV policy. We developed three parallel versions of a previously validated spreadsheet-based model; versions differed by the spreadsheet cell-referencing approach (named single cells; column/row references; named matrices). For each version, we implemented three model revisions (re-entry into care; guideline-concordant treatment initiation; immediate treatment initiation). After standard error-checking, we identified unintentional errors by comparing model output across the three versions. Concordant model output across all versions was considered error-free. We calculated the impact of unintentional errors as the percentage difference in model projections between model versions with and without unintentional errors, using +/-5% difference to define a material error.

Results

We identified 58 original and 4,331 propagated unintentional errors across all model versions and revisions. Over 40% (24/58) of original unintentional errors occurred in the column/row reference model version; most (23/24) were due to incorrect cell references. Overall, >20% of model spreadsheet cells had material unintentional errors. When examining error impact along the HIV care continuum, the percentage difference between versions with and without unintentional errors ranged from +3% to +16% (named single cells), +26% to +76% (column/row reference), and 0% (named matrices).

Conclusions

Standard error-checking techniques may not identify all errors in spreadsheet-based models. Comparing parallel model versions can aid in identifying unintentional errors and promoting reliable model projections, particularly when resources are limited.

Validation of a DICE Simulation Against a Discrete Event Simulation Implemented Entirely in Code

BACKGROUND

Modeling is an essential tool for health technology assessment, and various techniques for conceptualizing and implementing such models have been described. Recently, a new method has been proposed-the discretely integrated condition event or DICE simulation-that enables frequently employed approaches to be specified using a common, simple structure that can be entirely contained and executed within widely available spreadsheet software. To assess if a DICE simulation provides equivalent results to an existing discrete event simulation, a comparison was undertaken.

METHODS

A model of osteoporosis and its management programmed entirely in Visual Basic for Applications and made public by the National Institute for Health and Care Excellence (NICE) Decision Support Unit was downloaded and used to guide construction of its DICE version in Microsoft Excel^®. The DICE model was then run using the same inputs and settings, and the results were compared.

RESULTS

The DICE version produced results that are nearly identical to the original ones, with differences that would not affect the decision direction of the incremental cost-effectiveness ratios (<1% discrepancy), despite the stochastic nature of the models.

LIMITATION

The main limitation of the simple DICE version is its slow execution speed.

CONCLUSIONS

DICE simulation did not alter the results and, thus, should provide a valid way to design and implement decision-analytic models without requiring specialized software or custom programming. Additional efforts need to be made to speed up execution.

OP11 Structural Uncertainty In Economic Modelling For Smoking Cessation

INTRODUCTION:

Guidance for developing economic models recommend that model structure is carefully considered, and assumptions varied in sensitivity analysis (1). Models in smoking cessation have typically used cohort-level approaches, although recently discrete event simulations (DESs) have been developed (2). DESs allow additional flexibility such as modelling changing risk over time, and recurrent events. Our aim was to explore the impact of varying model structure and assumptions on the cost-effectiveness of smoking cessation programs.

METHODS:

We built a cohort state-transition model which related mortality to smoking status and considered the prevalence (based on smoking status) of five comorbidities associated with smoking, each of which has an associated cost and quality of life decrement. We additionally built a patient-level DES, using the Discretely Integrated Condition Event framework (3). The DES used the same data as the cohort model, except considering incidence for comorbidities rather than prevalence. We considered a population of smokers aged 16 years old and an intervention costing GBP827 on which 27 percent of people quit, compared with no treatment. We produced results using the two models for comparable scenarios, and ran additional scenarios considering different assumptions.

RESULTS:

In the cohort model, the incremental cost-effectiveness ratio (ICER) for intervention versus no treatment was GBP4,000/quality-adjusted life year (QALY). In the DES, modelling mortality linked to smoker status produced an ICER of GBP1,000/QALY and modelling mortality linked to comorbidities produced an ICER of GBP6,000/QALY. In the DES with mortality linked to comorbidities, varying the relative risk of comorbidities with time since quitting gave an ICER of GBP3,000/QALY. Including relapse increased the ICER to GBP21,000/QALY.

CONCLUSIONS:

The ICER for the smoking cessation program changes when model assumptions are varied, although the choice of DES versus cohort model appears to make a relatively small difference. Inclusion of relapse substantially changes the ICER, demonstrating the importance of long-term effects in economic models.