PharmacoEconomics最新文献

Background: Individual-level microsimulation models are essential for evaluating colorectal cancer (CRC) screening programmes to capture the heterogeneity in disease progression. To ensure regional relevance, such models require detailed natural history structures and robust calibration to population-specific data. This study presents the development of the first CRC natural history microsimulation model tailored to Northern Ireland (NI) for evaluating the NI Bowel Cancer Screening Programme (NI BCSP).

Method: The model simulates individual trajectories from adenoma onset to CRC diagnosis. Eight natural history parameters were calibrated to sex-specific CRC incidence data, initially using empirical (frequentist) calibration and Approximate Bayesian Computation (ABC) rejection, followed by the ABC-Markov Chain Monte Carlo (ABC-MCMC) algorithm. Other parameters were informed by NI-specific data sources.

Results: The frequentist and ABC rejection calibration approach's posterior distributions informed the prior distribution for the ABC-MCMC approach. ABC-MCMC was informative, yielding 55 parameter sets, but results were constrained by limited calibration targets and parameter identifiability.

Conclusion: This is the first NI-specific CRC microsimulation model, providing a regionally tailored platform for evaluating screening strategies and supporting policy. Calibration was feasible in a data-limited context, but further refinement and additional targets are needed to improve parameter estimation.

Microsimulation models have become increasingly common in the field of decision modeling for health. Because microsimulation models are computationally more demanding than traditional Markov cohort models, the use of computer programming languages in their development has become more common. C++ is a programming language that has gained widespread recognition in computationally intensive fields, including systems modeling and performance-critical applications. It offers powerful tools for building high-performance microsimulation models, outpacing many traditional modeling software solutions, such as native R, in terms of speed and control over memory management. However, there is limited accessible guidance for implementing microsimulation models in C++. This tutorial offers a step-by-step approach to constructing microsimulation models in C++ and demonstrates its application through simplified but adaptable example decision models. We walk the reader through essential steps and provide generic C++ code that is flexible and suitable for adapting to a range of models. Finally, we present the standalone C++ models and their Rcpp counterparts run within R, and compare their performance to equivalent R implementations in terms of speed and memory efficiency.

Introduction: When developing health economic simulation models, individual-level and cohort state-transition model types are commonly used. However, heterogeneity and the extent to which it is taken into account is thought to affect simulation outcomes differently in individual-level and cohort simulations, even when model structures are identical.

Objective: This study aimed to investigate the conditions under which the use of different model types may lead to different outcomes and therefore potentially different policy decisions.

Methods: A microsimulation model was used to reflect an individual-level simulation, simulating patient characteristics and, artificially, a cohort-level simulation of identical patients, using the exact same model structure. Four scenarios were analyzed: heterogeneity in age (scenario 1) influencing progression and recovery probabilities when on treatment, heterogeneity in sex (scenario 2) influencing progression and recovery probabilities when on treatment, combined heterogeneity in age and sex (scenario 3) influencing progression and recovery probabilities when on treatment, and heterogeneity in age when including age-dependent all-cause mortality (scenario 4). In every scenario, heterogeneity impact was varied, and health state occupancy, incremental costs, incremental effects, and the net monetary benefit of treatment versus no treatment were compared between the individual-level and cohort simulations.

Results: When introducing heterogeneity in age, sex, and age and sex combined, all scenarios showed differences between outcomes of individual-level and cohort simulations. However, these differences did not change the cost-effectiveness conclusions. When age influenced only age-dependent mortality, there were differences between the outcomes for the individual-level and cohort simulations when heterogeneity in age was introduced.

Conclusion: Patient heterogeneity can affect the outcomes of individual and cohort simulations differently, but reflecting more heterogeneity does not necessarily increase differences in simulation outcomes. However, age-dependent mortality affected analytic outcomes differently, suggesting a need for caution when developing cohort models if age is heterogeneous.

Background and objective: In cost-effectiveness analysis, treatment decisions are analysed at the population level. Combinations of treatment strategies that account for the heterogeneity of costs and effects across patients can be more cost-effective than a "one size fits all" approach. Individualized treatment rules (ITRs) assign a specific treatment to every patient based on their relevant characteristics, such that overall cost-effectiveness is optimized, but do not include feasibility or ethical considerations. We propose an approach for the design of ITRs based on simulated patient data from microsimulation models using statistical learning techniques.

Methods: We mathematically define the optimal ITR and how to measure the value of an ITR in a cost-effectiveness context. We explore least absolute shrinkage and selection operator (LASSO) regression, classification trees, and policy trees to illustrate how standard statistical learning techniques can be used to derive ITRs. We compare the strengths and limitations of these three approaches in terms of three criteria: the incremental value of the ITRs compared to optimal treatment assignment in terms of net monetary benefit (NMB), computational speed, and the interpretability of the ITRs. We propose methods to describe the impact of parameter uncertainty on the ITRs. We also explore how stochastic uncertainty can impact the ITR incremental value. We illustrate the methods by applying them to a microsimulation model for haemophilia B comparing four treatment strategies as a case study. The relevant patient characteristics in this model are the annualized bleeding rate, age, and sex.

Results: In our case study, a simple two-layer-deep classification tree is best suited based on the three criteria. This classification tree allocates treatments depending on whether the annualized bleeding rate of a patient is above or below 30 and whether their age is above or below 51. The optimal threshold values are uncertain based on the 95% credible ranges from the probabilistic analysis: 21-46 for annualized bleeding rate and 42-56 for age. Scenarios show that stochastic uncertainty has an impact on the incremental value of the ITR.

Discussion: Based on methodological considerations and the empirical findings in our case study, we expect the superiority of classification trees for the derivation of ITRs to be generalizable to other microsimulation models. This finding needs to be confirmed in future applications. Stochastic uncertainty has significant impacts on the ITRs, such that accurate representations of individual patient pathways are particularly crucial when designing ITRs. Future research could explore further empirical models and analytical approaches for ITRs or consider the translation of ITRs into the real-world decision-making context.

Background and objective: A primary elective total knee replacement is routinely used for patients with advanced osteoarthritis. Knee implants differ in characteristics (constraint, fixation, mobility), costs, need for revisions and other health outcomes, and so models evaluating their relative cost effectiveness are required to optimise decision making. Economic modelling approaches differ in complexity, the simplest in use being discrete time Markov models (DTMMs). Continuous-time Markov models (CTMMs) can capture transition timing in finer detail, and can more flexibly relax the constant hazard assumption. Multistate microsimulation can more easily capture patient history and time dependence. This paper aims to explore how the choice of modelling approach influences the cost effectiveness of various implant types for a total knee replacement. Based on the frequency of implant use in the National Joint Registry, 12 commonly used implants were included in the analysis.

Methods: We compared four different models of increasing complexity for male and female individuals in five age categories undergoing a total knee replacement. The DTMM and constant hazard CTMM assumed fixed revision probabilities over time. The individual-level CTMM with splines were semi-Markov, allowing time-varying rates of first revision surgery. The multistate microsimulation incorporated time-dependent splines for all revision rates but also dependence on time spent in previous health states. All revision rates were estimated using data from the National Joint Registry. The models were implemented using the hesim package in R.

Results: Under the constant hazard assumption, DTMM and CTMM yielded similar results, identifying the most commonly used implant as the most cost effective. However, using the spline-based hazard CTMM and patient history informed multistate microsimulation, other implants were identified as the most cost-effective options. The increased model complexity required high-performance computing facilities for CTMMs and multistate microsimulation.

Conclusions: This study shows that the choice of model can impact cost-effectiveness results. The multistate microsimulation model, which incorporates time-dependent transitions, provides a realistic representation of patient pathways over time, but is computationally complex and may be preferable only when time-varying risks are a key factor. The CTMM or DTMM models may be more efficient when data are limited or computational resources are constrained. Improving the accuracy and applicability of economic models can improve healthcare decision making. Future research should extend these methodologies to other disease areas, refine continuous-time models and explore their impact across diverse healthcare contexts.

Simulation models inform health policy decisions by integrating data from multiple sources and forecasting outcomes when there is a lack of comprehensive evidence from empirical studies. Such models have long supported health policy for cancer, the first or second leading cause of death in over 100 countries. Discrete-event simulation (DES) and Bayesian calibration have gained traction in the field of decision science because they enable flexible modeling of complex health conditions and produce estimates of model parameters that reflect real-world disease epidemiology and data uncertainty given model constraints. This uncertainty is then propagated to model-generated outputs, enabling decision-makers to assess confidence in recommendations and estimate the value of collecting additional information. However, there is limited end-to-end guidance on structuring a DES model for cancer progression, estimating its parameters using Bayesian calibration, and applying the calibration outputs to policy evaluation. To fill this gap, we introduce the DES Modeling Framework for Cancer Interventions and Population Health in R (DESCIPHR), an open-source codebase integrating a flexible DES model for the natural history of cancer, Bayesian calibration for parameter estimation, and an example application of screening strategy evaluation. To illustrate the framework, we apply DESCIPHR to calibrate bladder and colorectal cancer models to real-world cancer registry targets. We also introduce an automated method for generating data-informed parameter prior distributions and increase the functionality of a neural network emulator-based Bayesian calibration algorithm. We anticipate that the adaptable DESCIPHR modeling template will facilitate the construction of future decision models evaluating the risks and benefits of health interventions.

Objective: To develop a patient-level simulation model of type 1 diabetes (T1D) covering both childhood and adulthood. The goal is to identify and evaluate the cost-effectiveness of optimal screening for pre-symptomatic T1D.

Methods: We developed a Python-based simulation model to track 100,000 participants screened in childhood, capturing a subset of those at risk and transitioning to T1D, to estimate the incremental cost-effectiveness per life year gained of screening versus no screening. Our multi-objective optimisation approach sought to minimise three objectives: incremental cost effectiveness ratio, diabetic ketoacidosis (DKA) events at onset and the maximum number of screening tests a child can have with the healthcare system. The NSGA-II algorithm is used to explore the set of possible screening strategies from combinations of genetic risk score (GRS) and islet autoantibody (IA) measurements at different ages and frequencies during the first 15 years of life. Data for transition probabilities include large scale screening studies such as The Environmental Determinants of Diabetes in the Young, TrialNet, published risk functions, clinical trials and epidemiologic studies.

Results: We illustrate the use of multi-objective optimisation in patient-level simulations by estimating an optimal subset of T1D screening strategies in the USA. We identify four screening strategies with incremental cost-effectiveness ratios that meet commonly cited cost-effectiveness thresholds, which require, respectively, a maximum of 1, 2 3 and 4 islet autoantibody (IA) tests.

Conclusions: This article and corresponding model code can be used as a reference for implementing a multi-objective optimisation pipeline in patient-level simulation models.

This illustration uses the Scottish Cardiovascular Disease (CVD) Policy Model as a case study to provide a comprehensive, step-by-step guide to building a discrete event simulation (DES) model in R. It is specifically designed for practitioners who are familiar with constructing Markov models in R and wish to transition their theoretical knowledge of DES into practical implementation. The Scottish CVD Policy Model was originally developed as an Excel-based Markov model with a sophisticated structure: a primary Markov model for first events and nested sub-Markov models for subsequent events. Later replicated in R by Xin, Yiqiao et al., the model's source code was made publicly available on GitHub, underscoring its potential as a teaching tool. The intricate structure of this model presents several challenges in health economic modeling, making it an ideal candidate for demonstrating how DES techniques can address such complexities effectively. In this illustration, we deliberately avoid using R packages developed specifically for DES to enhance transparency. Instead, we rely on base R functions, and the tidyverse package for tidy data wrangling. This approach ensures that every step of the DES implementation is clear and reproducible. In addition to covering fundamental topics such as how to simulate a time to event according to an assumed distribution, and continuous discounting, the illustration also provides solutions to more advanced modeling challenges, such as handling piecewise-modeled cost and utility. By discussing both general principles and complex scenarios, this paper equips readers with the practical tools needed to transition from Markov to DES frameworks, enhancing the accuracy and flexibility of health economic evaluations.

Background and objective: The cost-utility of a panel-based pre-emptive pharmacogenomic (PPGx) test has not been evaluated in a multi-ethnic Asian population. Prior studies have largely focused on reactive, single drug-gene tests. This study assessed the cost-utility of a PPGx panel test and identified key drivers influencing its economic value.

Methods: We developed a prioritization framework integrating clinical and economic criteria to select drug-gene pairs for economic analysis. Cost-utility analysis was conducted using Discretely Integrated Condition Event (DICE) simulation, which allowed simultaneous analysis of multiple diseases and treatments of varying duration. The analysis focused on a hypothetical cohort of healthy 40-year-old Singaporeans and assessed the lifetime impact of a one-time panel test on outcomes such as disease occurrence and serious adverse drug events (ADE). Costs were evaluated from a healthcare payer's perspective and reported in 2024 Singapore dollars (S$). Both costs and health outcomes were discounted at 3% annually. Deterministic, probabilistic, and scenario analyses were performed to address uncertainty.

Results:  Four drug-gene pairs were selected: clopidogrel-CYP2C19, capecitabine-DPYD, allopurinol-HLA-B*58:01, and simvastatin-SLCO1B1. In the base case, panel testing was dominant, resulting in savings of S$37,600 and gain of 9.32 quality-adjusted life years (QALYs) per 1000 individuals compared with no PGx testing. Results were sensitive to drug costs, ADE-related costs, and the age for panel administration. Ideal drug-gene pairs for panel inclusion involve commonly prescribed drugs with variants associated with severe ADEs, where genotype-guided alternatives (e.g., dose adjustment or switching therapy) have costs comparable to standard care.

Conclusions: Pre-emptive PGx panel testing is economically viable when panel design, variant prevalence, drug costs, and local prescribing patterns are carefully considered. As more data become available, the model can be tailored to evaluate additional drug-gene pairs and their downstream consequences.