Our latest research, all in one place. Browse our collection of journal articles, reports and conference proceedings to see how we’re contributing to HEOR research. Remember to:
BACKGROUND: In levodopa-treated individuals with Parkinson’s disease (PD) and end-of-dose motor fluctuations, the BIPARK-I randomized controlled trial (RCT) demonstrated that opicapone is noninferior to entacapone in reducing OFF-time. Furthermore, the BIPARK-II RCT demonstrated that opicapone is well tolerated and significantly reduces OFF-time compared with placebo. This study developed a cost-effectiveness model (CEM) of opicapone compared with entacapone from the perspective of the English National Health Service (NHS) and personal social services (PSS).
METHODS: The CEM used a Markov model with three health states, including “<25% OFF-time,” “≥25% OFF-time,” and “dead,” as individuals spending less than 25% of their awake time experiencing OFF-time have previously been shown to have a significantly improved health-related quality of life and to accumulate fewer healthcare costs. The CEM had a 25-year time horizon, expressed costs as 2021/22 Great British Pounds (GBPs), and health outcomes as quality-adjusted life years (QALYs). Both costs and health outcomes were discounted at 3.5% annually, and a cost-effectiveness threshold of £20,000 per QALY was used. Probabilistic sensitivity analysis (PSA) considered parameter uncertainty. RESULTS: The deterministic base case indicates that an individual treated with opicapone accrues fewer costs and more QALYs compared with each entacapone comparator and, therefore, is considered cost-effective. The PSA indicates that the probability that opicapone is cost-effective ranges from 87.2% to 98.0%, depending on the choice of entacapone comparator. CONCLUSIONS: Opicapone is cost-effective when compared with entacapone for levodopa-treated PD patients experiencing end-of-dose motor fluctuations.
Health economics is an important field not just in the UK but globally. Decision-makers can use it to prioritise resources used in healthcare systems to maximise population health whilst ensuring value for money. Since 1999, the National Institute for Health and Care Excellence (NICE) has been responsible for these decisions in the UK. There is a need to provide these decision-makers with the best possible evidence for new treatments to support well-informed decisions.
The vast majority of health economic modelling is still conducted using Markov models to capture disease progression. However, these models rely on restrictive assumptions that may not accurately capture real-world disease progression. This talk will discuss two examples from a recent publication
where regression modelling was used to improve the health economic model. The first example will cover using a multinomial logistic regression to predict the changes in haemoglobin health states over time in patients with chronic kidney disease. Traditional Markov models assume a fixed disease progression rate which fails to account for patient subgroups and time-varying rates. Regression models enable dynamic transitions based on patient characteristics like age, sex, and baseline haemoglobin levels. This leads to better modelling of disease progression and subsequent decision-making.
The second example will cover using a generalised linear mixed model (GLMM) to estimate drug doses for each haemoglobin health state whilst adjusting for subgroup characteristics. This allows for a more granular approach to cost-effectiveness analysis. Unlike Markov models, which often require simplifying assumptions to capture drug doses, a GLMM better reflects the uncertainty in the target-population.
As real-world evidence and individual patient data (IPD) become more accessible, alongside increasing computational power, statisticians must apply and share advanced modelling approaches. By doing so, we can enhance healthcare decision-making, ultimately leading to better policies and improved population health.
OBJECTIVES: Probabilistic sensitivity analysis (PSA) is used to characterise uncertainty in cost-effectiveness models. A model was developed using R and Shiny to explore the impact of different parameter correlation structures on PSA outputs.
METHODS: A Markov model was built in R to compare a hypothetical treatment and comparator. Three options were built into the model: no correlation (inputs varied independently); part correlation (correlation within but not between costs, utilities and transition matrices); and full correlation (correlation between all inputs). A Shiny interface allowed users to explore the impact of the correlation options with different model parameters. Features of the Shiny model included preloading base case results, running the Markov model with an arbitrary number of health states and costs, and displaying results for different subsets of correlation options. A scenario analysis was included in the Shiny model to determine the circumstances in which correlation had the largest impact by varying the treatment cost as a proxy for the ICER.
RESULTS: While the ICER was comparable across all correlation options, the likelihood of cost-effectiveness differed substantially from 61% to 93%. In all scenarios, the 'no correlation' option displayed the most certain likelihood (closest to either 0 or 1) of cost-effectiveness, while the least certain was produced by the full correlation option. Counterintuitively, correlating inputs increased uncertainty because it allowed for a greater number of 'extreme' scenarios to be generated, whereas allowing independent generation of large numbers of inputs tends to lead to a 'cancelling out' effect. This effect was most pronounced when the ICER is moderately close to the willingness-to-pay threshold.
CONCLUSION: This analysis demonstrates that input correlation can have a substantial impact on the level of certainty in model outputs, and by ignoring this, the model may be over- or under-stating the true level of confidence.
OBJECTIVES: Health Technology Assessment (HTA) guidelines in the UK mention the importance of uncertainty in cost-effectiveness analysis (CEA). The presence of missing data within data sets used to provide inputs for CEA can be a source of uncertainty. One setting in which missing data may be prevalent is for utility scores and these are often used in CEA. The intention of this analysis is to formalise recommendations for dealing with missingness and outline a minimum accepted reporting standard for missing utility data.
METHODS: A simulated patient level dataset (SIPD) was created in an oncology setting. The SIPD included utility scores for hypothetical individuals and treatments. Key prognostic variables were also simulated. Missing data was generated for utility for 10% and 30% of the observations, for missing completely at random (MCAR), at random (MAR) and not at random (MNAR) missingness mechanisms. Four methods for addressing missing data were analysed, complete case analysis (CCA), mean score estimation (MSE), multiple imputation via chained equations (MICE) and linear mixed modelling (LMM). The outcome of interest was the mean difference between the true utility value and the estimated utility value from each method.
RESULTS: MICE tended to result in the lowest mean difference across levels and mechanism. For the 10% level and MCAR mechanism, CCA resulted in negligible differences from the true utility value. MNAR data tended to result in substantial differences from the true utility value regardless of the level of missingness and method used. LMMs suppressed the standard deviation of utility.
CONCLUSION: Differing results across levels and mechanisms for each method used to deal with missingness were observed here. This suggest that the level and mechanism of missingness should be investigated and reported as a minimum standard for all data sets that inform economic models for HTA in the UK.
BACKGROUND AND OBJECTIVE: Health technology assessment (HTA) of haemato-oncology therapies typically requires extrapolation of long-term survival beyond a trial’s follow-up. Health technology assessment agencies must balance caution around uncertainty in early follow-up trial data whilst aiming to provide timely access. This study qualitatively and quantitatively assessed how eight HTA agencies considered maturing data and external evidence.
METHODS: The eight HTA appraisals were based on ZUMA-7, a phase III trial for axicabtagene ciloleucel (axi-cel) for second-line diffuse large B-cell lymphoma. ZUMA-7 survival data were submitted with either a 25-month (‘Interim’) or 47-month (‘Primary’) follow-up. To inform axi-cel Interim survival extrapolations, external evidence was available from a prior mature single-arm trial for third-line or later diffuse large B-cell lymphoma (ZUMA-1). A qualitative assessment of eight different submissions to HTA agencies was undertaken to determine key discussion points. The value and cost of waiting for evidence to mature between Interim and Primary analyses were quantified using value of information methods to evaluate the impact of waiting for further evidence collection on population health.
RESULTS: Agencies used varied approaches to account for uncertainty in survival extrapolations in both Interim and Primary analyses. No agency considered external evidence fully during Interim submissions; one used it partially to inform clinical plausibility; four did not consider it. Health technology assessment agencies that did not consider the relevance of ZUMA-1 were more inclined to wait for more mature evidence to mitigate uncertainty. When ZUMA-1 aided in determining a plausible range for Interim extrapolations, the less valuable more mature evidence became, with the cost of waiting for Primary analysis results exceeding the value conferred.
CONCLUSIONS: There was limited consideration of external evidence during the included HTA submissions. In the future, it is recommended that external evidence should be considered to a greater degree by both manufacturers and HTA agencies when extrapolating survival to ensure appropriate and timely HTA decisions that minimise the undue burden on healthcare systems.
BACKGROUND: The SECURE trial (NCT02596126) demonstrated the efficacy of the cardiovascular polypill ("CV-Polypill" - acetyl salicylic acid, atorvastatin and ramipril) in reducing the risk of recurrent major cardiovascular events compared with standard care when initiated within six months of a myocardial infarction. This analysis aimed to estimate the cost-effectiveness of the CV-Polypill from the Spanish healthcare perspective using SECURE trial data.
METHODS: A decision analytic Markov modelling approach was conducted to compare the CV-Polypill with standard care over a lifetime time horizon. Six parametric distributions were fitted to SECURE trial data on time to reinfarction, stroke or death (cardiovascular or non-cardiovascular). Cost and utility data were sourced from literature. Respective model outputs were discounted at 3%. The model captured direct medical costs associated with treatment acquisition and acute/ongoing cardiovascular events. Probabilistic sensitivity analyses (PSA) and scenario analyses were conducted.
FINDINGS: The CV-Polypill is dominant (improves health outcomes and reduces costs) in 84·8% of PSA iterations (848/1000 iterations), and cost effective in 89·3% of PSA iterations (893/1000 iterations) at a €30,000 threshold. Secondary prevention with the CV-Polypill reduces the recurrence of cardiovascular events and costs over the time horizon, from the Spanish healthcare perspective. A range of scenario analyses were conducted, demonstrating the robustness of the results when different inputs and assumptions were varied.
INTERPRETATION: The CV-Polypill is a dominant strategy in secondary cardiovascular prevention, compared with standard care, from the Spanish healthcare perspective. The CV-Polypill should be considered as a secondary prevention for Spanish patients, like those enrolled in SECURE, at hospital discharge.
BACKGROUND & OBJECTIVES: Patients surviving a non-ST-elevation myocardial infarction (NSTEMI) have an elevated risk of future major adverse cardiovascular events (MACE), which can be mitigated through long-term cardiac arrhythmia monitoring. The present study evaluated the cost-effectiveness of continuous remote arrhythmia monitoring using an insertable cardiac monitor (ICM) combined with standard of care (SoC) compared with SoC alone.
METHODS: A cost-effectiveness analysis using a lifetime partitioned survival model was developed for high-risk NSTEMI patients from a UK National Health Service (NHS) perspective. Survival analysis was used to determine the transition of patients from the pre-MACE health state (where patients could experience arrhythmia, major bleeding, or systemic embolism) to the MACE health state (worsening heart failure, stroke, and acute coronary syndrome events). The survival analysis and arrhythmia diagnosis rates were informed by the BIO|GUARD-MI trial. The model captured direct costs associated with each MACE and implantation and removal of the ICM device and treatment costs following arrhythmia detection. The model captured the health implications for an ICM with SoC, compared with SoC alone, in terms of the total quality-adjusted life years (QALYs). Deterministic and probabilistic sensitivity analyses were undertaken to explore the impact of parameter uncertainty on the model results.
RESULTS: The use of ICMs plus SoC for daily remote cardiac arrhythmia monitoring is cost effective, when compared with SoC alone, in high-risk NSTEMI patients over a lifetime horizon, with an incremental cost-effectiveness ratio of £7766 per QALY gained. The ICM was associated with an additional 0.184 QALYs per patient for an additional cost of £1430. The ICM remained cost effective during the deterministic and probabilistic sensitivity analyses.
CONCLUSION: The addition of an ICM to SoC in high-risk NSTEMI patients is cost effective from the perspective of the UK NHS and would, therefore, be a further option for the management of such patients in clinical practice.
