Individual-level Simulation Model Research Articles

A Fast Nonparametric Sampling (NPS) Method for Time-to-Event in Individual-Level Simulation Models.

Individual-level simulation models often require sampling times to events, however efficient parametric distributions for many processes may often not exist. For example, time to death from life tables cannot be accurately sampled from existing parametric distributions. We propose an efficient nonparametric method to sample times to events that does not require any parametric assumption on the hazards. We developed a nonparametric sampling (NPS) approach that simultaneously draws multiple time-to-event samples from a categorical distribution. This approach can be applied to univariate and multivariate processes. We discretize the entire period into equal-length time intervals and then derived the interval-specific probabilities. The times to events can then be used directly in individual-level simulation models. We compared the accuracy of our approach in sampling time-to-events from common parametric distributions, including exponential, gamma, and Gompertz. In addition, we evaluated the method's performance in sampling age to death from US life tables and sampling times to events from parametric baseline hazards with time-dependent covariates. The NPS method estimated similar expected times to events from 1 million draws for the three parametric distributions, 100,000 draws for the homogenous cohort, 200,000 draws from the heterogeneous cohort, and 1 million draws for the parametric distributions with time-varying covariates, all in less than a second. Our method produces accurate and computationally efficient samples for time-to-events from hazards without requiring parametric assumptions.

Economic Assessment in Resource-Constrained Systems: Individual-Level Simulation Model in Wet Age-Related Macular Degeneration and Diabetic Macular Oedema.

Cost-effectiveness analyses typically ignore healthcare system resource constraints. Ophthalmology is affected by resource constraints because of increasing disease prevalence and the use of resource-intensive treatments. This study evaluated the impact of resource constraints on the cost-effectiveness of faricimab 6mg, compared with aflibercept 2mg and ranibizumab biosimilar 0.5mg, for treating wet age-related macular degeneration (wAMD) or diabetic macular oedema (DMO) over a 5-year horizon. A microsimulation model estimated the impact of resource constraints on patients visits, delays, costs and quality-adjusted life-year (QALY) losses due to treatment delays at a typical UK National Health Service eye hospital treating 1500 patients with wAMD and 500 patients with DMO. Patient characteristics, treatment regimens and treatment intervals were informed using published literature and expert opinion. Resource constraint was represented by limiting the number of available intravitreal injection appointments per week, with growing demand caused by rising disease prevalence. The model compared outcomes across three scenarios; each scenario involved treating all patients with one of the three treatments. Over 5years, in a resource-constrained hospital, compared with aflibercept, faricimab use resulted in the avoidance of 12,596 delays, saved GBP/£15,108,609 in cost and avoided the loss of 60.06 QALYs. Compared with ranibizumab biosimilar, faricimab use resulted in the avoidance of 18,910 delays, incurred £2,069,088 extra cost and avoided the loss of 105.70 QALYs, resulting in an incremental cost-effectiveness ratio of £19,574/QALY. Accounting for resource constraints in health economic evaluation is crucial. Emerging therapies that are more durable and require less frequent clinic visits can reduce treatment delays, leading to improved QALY outcomes and reduced burden on healthcare systems. Faricimab reduced the number of delayed injections, leading to improved QALY outcomes for patients in a healthcare system with resource constraints. Faricimab is cost-saving when compared with aflibercept and cost-effective when compared with ranibizumab biosimilar.

Emulator-Based Bayesian Calibration of the CISNET Colorectal Cancer Models.

To calibrate Cancer Intervention and Surveillance Modeling Network (CISNET)'s SimCRC, MISCAN-Colon, and CRC-SPIN simulation models of the natural history colorectal cancer (CRC) with an emulator-based Bayesian algorithm and internally validate the model-predicted outcomes to calibration targets. We used Latin hypercube sampling to sample up to 50,000 parameter sets for each CISNET-CRC model and generated the corresponding outputs. We trained multilayer perceptron artificial neural networks (ANNs) as emulators using the input and output samples for each CISNET-CRC model. We selected ANN structures with corresponding hyperparameters (i.e., number of hidden layers, nodes, activation functions, epochs, and optimizer) that minimize the predicted mean square error on the validation sample. We implemented the ANN emulators in a probabilistic programming language and calibrated the input parameters with Hamiltonian Monte Carlo-based algorithms to obtain the joint posterior distributions of the CISNET-CRC models' parameters. We internally validated each calibrated emulator by comparing the model-predicted posterior outputs against the calibration targets. The optimal ANN for SimCRC had 4 hidden layers and 360 hidden nodes, MISCAN-Colon had 4 hidden layers and 114 hidden nodes, and CRC-SPIN had 1 hidden layer and 140 hidden nodes. The total time for training and calibrating the emulators was 7.3, 4.0, and 0.66 h for SimCRC, MISCAN-Colon, and CRC-SPIN, respectively. The mean of the model-predicted outputs fell within the 95% confidence intervals of the calibration targets in 98 of 110 for SimCRC, 65 of 93 for MISCAN, and 31 of 41 targets for CRC-SPIN. Using ANN emulators is a practical solution to reduce the computational burden and complexity for Bayesian calibration of individual-level simulation models used for policy analysis, such as the CISNET CRC models. In this work, we present a step-by-step guide to constructing emulators for calibrating 3 realistic CRC individual-level models using a Bayesian approach. We use artificial neural networks (ANNs) to build emulators that surrogate complex individual-based models to reduce the computational burden in the Bayesian calibration process.ANNs showed good performance in emulating the CISNET-CRC microsimulation models, despite having many input parameters and outputs.Using ANN emulators is a practical solution to reduce the computational burden and complexity for Bayesian calibration of individual-level simulation models used for policy analysis.This work aims to support health decision scientists who want to quantify the uncertainty of calibrated parameters of computationally intensive simulation models under a Bayesian framework.

A Canadian Simulation Model for Major Depressive Disorder: Study Protocol.

Major depressive disorder (MDD) is a common, often recurrent condition and a significant driver of healthcare costs. People with MDD often receive pharmacological therapy as the first-line treatment, but the majority of people require more than one medication trial to find one that relieves symptoms without causing intolerable side effects. There is an acute need for more effective interventions to improve patients' remission and quality of life and reduce the condition's economic burden on the healthcare system. Pharmacogenomic (PGx) testing could deliver these objectives, using genomic information to guide prescribing decisions. With an already complex and multifaceted care pathway for MDD, future evaluations of new treatment options require a flexible analytic infrastructure encompassing the entire care pathway. Individual-level simulation models are ideally suited for this purpose. We sought to develop an economic simulation model to assess the effectiveness and cost effectiveness of PGx testing for individuals with major depression. Additionally, the model serves as an analytic infrastructure, simulating the entire patient pathway for those with MDD. Key stakeholders, including patient partners, clinical experts, researchers, and modelers, designed and developed a discrete-time microsimulation model of the clinical pathways of adults with MDD in British Columbia (BC), including all publicly-funded treatment options and multiple treatment steps. The Simulation Model of Major Depression (SiMMDep) was coded with a modular approach to enhance flexibility. The model was populated using multiple original data analyses conducted with BC administrative data, a systematic review, and an expert panel. The model accommodates newly diagnosed and prevalent adult patients with MDD in BC, with and without PGx-guided treatment. SiMMDep comprises over 1500 parameters in eight modules: entry cohort, demographics, disease progression, treatment, adverse events, hospitalization, costs and quality-adjusted life-years (payoff), and mortality. The model predicts health outcomes and estimates costs from a health system perspective. In addition, the model can incorporate interactive decision nodes to address different implementation strategies for PGx testing (or other interventions) along the clinical pathway. We conducted various forms of model validation (face, internal, and cross-validity) to ensure the correct functioning and expected results of SiMMDep. SiMMDep is Canada's first medication-specific, discrete-time microsimulation model for the treatment of MDD. With patientpartner collaboration guiding its development, it incorporates realistic care journeys. SiMMDep synthesizes existing information and incorporates provincially-specific data to predict the benefits and costs associated with PGx testing. These predictions estimate the effectiveness, cost-effectiveness, resource utilization, and health gains of PGx testing compared with the current standard of care. However, the flexible analytic infrastructure can be adapted to support other policy questions and facilitate the rapid synthesis of new data for a broader search for efficiency improvements in the clinical field of depression.

Using Age-Specific Rates for Parametric Survival Function Estimation in Simulation Models.

To describe a procedure for incorporating parametric functions into individual-level simulation models to sample time to event when age-specific rates are available but not the individual data. Using age-specific event rates, regression analysis was used to parametrize parametric survival distributions (Weibull, Gompertz, log-normal, and log-logistic), select the best fit using the R2 statistic, and apply the corresponding formula to assign random times to events in simulation models. We used stroke rates in the Spanish population to illustrate our procedure. The 3 selected survival functions (Gompertz, Weibull, and log-normal) had a good fit to the data up to 85 y of age. We selected Gompertz distribution as the best-fitting distribution due to its goodness of fit. Our work provides a simple procedure for incorporating parametric risk functions into simulation models without individual-level data. We describe the procedure for sampling times to event for individual-level simulation models as a function of age from parametric survival functions when age-specific rates are available but not the individual dataWe used linear regression to estimate age-specific hazard functions, obtaining estimates of parameter uncertainty.Our approach allows incorporating parameter (second-order) uncertainty in individual-level simulation models needed for probabilistic sensitivity analysis in the absence of individual-level survival data.

Would initiating colorectal cancer screening from age of 45 be cost-effective in Germany? An individual-level simulation analysis.

Colorectal cancer (CRC) screening has been shown to be effective and cost-saving. However, the trend of rising incidence of early-onset CRC challenges the current national screening program solely for people ≥50 years in Germany, where extending the screening to those 45-49 years might be justified. This study aims to evaluate the cost-effectiveness of CRC screening strategies starting at 45 years in Germany. DECAS, an individual-level simulation model accounting for both adenoma and serrated pathways of CRC development and validated with German CRC epidemiology and screening effects, was used for the cost-effectiveness analysis. Four CRC screening strategies starting at age 45, including 10-yearly colonoscopy (COL), annual/biennial fecal immunochemical test (FIT), or the combination of the two, were compared with the current screening offer starting at age 50 years in Germany. Three adherence scenarios were considered: perfect adherence, current adherence, and high screening adherence. For each strategy, a cohort of 100,000 individuals with average CRC risk was simulated from age 20 until 90 or death. Outcomes included CRC cases averted, prevented death, quality-adjusted life-years gained (QALYG), and total incremental costs considering both CRC treatment and screening costs. A 3% discount rate was applied and costs were in 2023 Euro. Initiating 10-yearly colonoscopy-only or combined FIT + COL strategies at age 45 resulted in incremental gains of 7-28 QALYs with incremental costs of €28,360-€71,759 per 1,000 individuals, compared to the current strategy. The ICER varied from €1,029 to €9,763 per QALYG, and the additional number needed for colonoscopy ranged from 129 to 885 per 1,000 individuals. Among the alternatives, a three times colonoscopy strategy starting at 45 years of age proves to be the most effective, while the FIT-only strategy was dominated by the currently implemented strategy. The findings remained consistent across probabilistic sensitivity analyses. The cost-effectiveness findings support initiating CRC screening at age 45 with either colonoscopy alone or combined with FIT, demonstrating substantial gains in quality-adjusted life-years with a modest increase in costs. Our findings emphasize the importance of implementing CRC screening 5 years earlier than the current practice to achieve more significant health and economic benefits.

Cost Effectiveness of Denosumab for Secondary Prevention of Osteoporotic Fractures Among Postmenopausal Women in China: An Individual-Level Simulation Analysis.

This study evaluated the cost effectiveness of denosumab versus alendronate for secondary prevention of osteoporotic fractures among post-menopausal women in China. A validated individual-level simulation model of osteoporotic fractures in the Chinese setting was adapted. Allowing both treatment discontinuation and waning effects, the analysis aimed to evaluate the incremental cost-effectiveness ratio of denosumab compared to alendronate by simulating a cohort of previously fractured individuals over the residual lifetime from the healthcare system perspective. Hip, vertebral, and wrist/humeral fractures were tracked along with the associated medical costs and quality-adjusted life-years. Age-related health state utility values, health state utility values of fractures, costs, fracture incidence, and mortality risks for Chinese were used whenever available. Comparative effectiveness data were obtained from a published network meta-analysis. One-way and probabilistic sensitivity analyses were conducted. In the base case, denosumab was dominated by alendronate with incremental costs of CN¥2743 (US$425) and incremental health outcomes of -0.20 quality-adjusted life-years at its current price in mainland China. It remained dominated in all one-way sensitivity analysis robustness checks. However, denosumab was cost effective if both drugs did not carry any waning effects. In the probabilistic sensitivity analysis, denosumab remained dominated in all replications. Denosumab is not cost effective for preventing secondary fractures among overall postmenopausal women in China. It is advisable to identify alternative denosumab regimens for high-risk subgroups among previously fractured postmenopausal women.

Process mining for healthcare decision analytics with micro-costing estimations

Managing constrained healthcare resources is an important and inescapable role of healthcare decision makers. Allocative decisions are based on downstream consequences of changes to care processes: judging whether the costs involved are offset by the magnitude of the consequences, and therefore whether the change represents value for money. Process mining techniques can inform such decisions by quantitatively discovering, comparing and detailing care processes using recorded data, however the scope of techniques typically excludes anything ‘after-the-process’ i.e., their accumulated costs and resulting consequences. Cost considerations are increasingly incorporated into process mining techniques, but the majority of healthcare costs for service and overhead components are commonly apportioned and recorded at the patient (trace) level, hiding event level detail. Within decision-analysis, event-driven and individual-level simulation models are sometimes used to forecast the expected downstream consequences of process changes, but are expensive to manually operationalise. In this paper, we address both of these gaps within and between process mining and decision analytics, by better linking them together. In particular, we introduce a new type of process model containing trace data that can be used in individual-level or cohort-level decision-analytical model building. Furthermore, we enhance these models with process-based micro-costing estimations. The approach was evaluated with health economics and decision modelling experts, with discussion centred on how the outputs could be used, and how similar information would otherwise be compiled.

Modeling the impact of COVID-19 pandemic-driven increases in alcohol consumption on health outcomes and hospitalization costs in the United States.

Background and aimsAlcohol consumption increased in the early phases of the COVID‐19 pandemic in the United States. Alcohol use disorder (AUD) and risky drinking are linked to harmful health effects. This paper aimed to project future health and cost impacts of shifts in alcohol consumption during the COVID‐19 pandemic.DesignAn individual‐level simulation model of the long‐term drinking patterns for people with life‐time AUD was used to simulate 10 000 individuals and project model outcomes to the estimated 25.9 million current drinkers with life‐time AUD in the United States. The model considered three scenarios: (1) no change (counterfactual for comparison); (2) increased drinking levels persist for 1 year (‘increase‐1’) and (3) increased drinking levels persist for 5 years (‘increase‐5’).SettingUnited States.ParticipantsCurrent drinkers with life‐time AUD.MeasurementsLife expectancy [life‐years (LYs)], quality‐adjusted life‐years (QALYs), alcohol‐related hospitalizations and associated hospitalization costs and alcohol‐related deaths, during a 5‐year period.FindingsShort‐term increases in alcohol consumption (increase‐1 scenario) resulted in a loss of 79 000 [95% uncertainty interval (UI]) 26 000–201 000] LYs, a loss of 332 000 (104 000–604 000) QALYs and 295 000 (82 000–501 000) more alcohol‐related hospitalizations, costing an additional $5.4 billion ($1.5–9.3 billion) over 5 years. Hospitalizations for cirrhosis of the liver accounted for approximately $3.0 billion ($0.9–4.8 billion) in hospitalization costs, more than half the increase across all alcohol‐related conditions. Health and cost impacts were more pronounced for older age groups (51+), women and non‐Hispanic black individuals. Increasing the duration of pandemic‐driven increases in alcohol consumption in the increase‐5 scenario resulted in larger impacts.ConclusionsSimulations show that if the increase in alcohol consumption observed in the United States in the first year of the pandemic continues, alcohol‐related mortality, morbidity and associated costs will increase substantially over the next 5 years.

P1704: PROJECTED LIFETIME ECONOMIC BURDEN OF SEVERE SICKLE CELL DISEASE IN THE UNITED STATES

Background: Sickle cell disease (SCD) is a rare hereditary blood disorder affecting approximately 100,000 individuals in the United States (US). SCD is associated with severe symptom burden, frequent hospitalizations, organ damage and increased risk of mortality. Vaso-occlusive crises (VOCs) are a hallmark clinical manifestation of SCD and often require emergency room visits and inpatient admissions. While there is some evidence describing the lifetime direct healthcare costs for SCD, the economic burden for patients with severe SCD, defined as experiencing recurrent VOCs, is currently unknown. Aims: To estimate the per-patient lifetime direct healthcare cost of severe SCD from a US payer perspective using an economic model developed based on published model frameworks. Methods: An individual-level simulation model was used to estimate the lifetime direct healthcare cost for patients with severe SCD in the US. The model followed patients with severe SCD, with an average of 7 VOCs per year, from birth until death. Similar to previously published economic models, mortality was estimated based on the occurrence of VOCs and other acute and chronic complications of SCD. Cost inputs were derived from published literature. Wholesale acquisition costs were utilized for all drug cost inputs. Scenario analyses were conducted to determine the impact of disease severity, defined by the number of VOCs per year, on the estimates of lifetime cost. The cost and impact of recently approved chronic, non-curative SCD therapies were not included in analyses. Results: The model followed patients with severe SCD for an average of 35 years, over which time each patient experienced an average of 246 VOCs. The total undiscounted cost of SCD over the lifetime was estimated to be $5.2 million. The majority of costs (77%, $4.0 million) were attributable to the care associated with acute complications, which includes VOCs, acute chest syndrome and stroke. The lifetime cost of chronic complications, including pulmonary hypertension, chronic kidney disease, retinopathy, and neurocognitive impairment, was estimated as $760,000. Scenario analyses demonstrated the lifetime costs of SCD are highly sensitive to the annual number of VOCs; when varying the average number of VOCs per year from 5 to 9, the lifetime costs ranged from $4.2 million to $6.2 million. Patients with SCD who have a significantly elevated number of VOCs (14 annually) were projected to have $8.6 million in lifetime direct healthcare costs. Summary/Conclusion: Model projections show significant lifetime economic burden with severe SCD in the US, with per-patient cost estimates substantially higher than those previously published for the overall SCD population. Treatments that can eliminate the occurrence of VOCs in this population would significantly lower this economic burden.

The Challenges of Parameterizing Direct Effects in Individual-Level Simulation Models.

Individual-level simulation models are used to assess the effects of health interventions in complex settings. However, estimating valid causal effects using these models requires correct parametrization of the relationships between time-varying treatments, outcomes, and other variables in the causal structure. To parameterize these relationships, individual-level simulation models typically need estimates of the direct effects of treatment. However, direct effects of treatment are often not well- defined and therefore cannot be validly estimated from any data. In this paper, we explain the causal meaning of the parameters of individual-level simulation models as direct effects, describe why direct effects may be difficult to define unambiguously in some settings, and conclude with some suggestions for the design of individual-level simulation models in those settings.

Quantifying the impact of adherence to screening on colorectal cancer incidence and mortality

Abstract Background Current recommendations of The US Preventive Services Task Force (USPSTF) on colorectal cancer (CRC) screening strategies are based on models that assume 100% adherence to screening. Since adherence can largely affect the outcomes of a screening modality, we aimed to assess the comparative effectiveness of CRC screening strategies under published rates of actual adherence. Methods We developed an individual-level simulation model and validated it against landmark trials and USPSTF models. Then we assessed the effectiveness of colonoscopy (COL), flexible sigmoidoscopy (FS), high-sensitivity guaiac fecal occult blood test (HS-gFOBT), fecal immunochemical test (FIT), multitarget stool DNA test (FIT-DNA), computed tomography colonography (CTC), and methylated SEPT9 DNA test (SEPT9) in reducing CRC incidence and mortality. For each strategy, we also estimated the incremental life-years gained, number of colonoscopies, and adverse events. Results Assuming 100% adherence, FIT-DNA, FIT, HS-gFOBT, and SEPT9 averted 58 to 59 CRC cases and 28 CRC deaths; COL and CTC strategies 56 cases and 27 deaths, while FS averted 39 cases and 19 deaths per 1,000 individuals. Life-years gained were similar across FIT-DNA, FIT, HS-gFOBT, SEPT9, CTC, and COL strategies. The total number of colonoscopies was highest with COL (3,567), followed by SEPT9 (3,231), HS-gFOBT (2,584), FIT-DNA (2,079), FIT (2,067), CTC (1,691) and FS (1,538) strategies. Assuming actual adherence, SEPT9 averted 54 CRC cases and 26 CRC deaths, followed by COL with 49 cases and 24 deaths, and FIT-DNA, FIT, CTC and HS-gFOBT with approximately 36 to 41 cases and 18 to 21 deaths averted per 1000 individuals screened. Life-years gained reflected the effectiveness of each strategy in reducing CRC cases and deaths. Conclusions Adherence is a key factor in determining the effectiveness of CRC screening and strategies with higher expected adherence rates have the potential to reduce cancer incidence and mortality. Key messages Adherence has a substantial impact on screening outcomes, such as cancer incidence and mortality, and may influence selection of optimal screening strategies. Strategies with higher expected adherence rates can lead to clinically meaningful benefits compared to strategies that may have better one-time sensitivity and/or specificity.

The economic case for prevention of population vitamin D deficiency: a modelling study using data from England and Wales.

Vitamin D deficiency (VDD) affects the health and wellbeing of millions worldwide. In high latitude countries such as the United Kingdom (UK), severe complications disproportionally affect ethnic minority groups. To develop a decision-analytic model to estimate the cost effectiveness of population strategies to prevent VDD. An individual-level simulation model was used to compare: (I) wheat flour fortification; (II) supplementation of at-risk groups; and (III) combined flour fortification and supplementation; with (IV) a 'no additional intervention' scenario, reflecting the current Vitamin D policy in the UK. We simulated the whole population over 90 years. Data from national nutrition surveys were used to estimate the risk of deficiency under the alternative scenarios. Costs incurred by the health care sector, the government, local authorities, and the general public were considered. Results were expressed as total cost and effect of each strategy, and as the cost per 'prevented case of VDD' and the 'cost per Quality Adjusted Life Year (QALY)'. Wheat flour fortification was cost saving as its costs were more than offset by the cost savings from preventing VDD. The combination of supplementation and fortification was cost effective (£9.5 per QALY gained). The model estimated that wheat flour fortification alone would result in 25% fewer cases of VDD, while the combined strategy would reduce the number of cases by a further 8%. There is a strong economic case for fortifying wheat flour with Vitamin D, alone or in combination with targeted vitamin D3 supplementation.

A protocol for an economic evaluation of a polypill in patients with established or at high risk of cardiovascular disease in a UK NHS setting: RUPEE (NHS) study

IntroductionThe ‘Use of a Multi-drug Pill in Reducing cardiovascular Events’ (UMPIRE) trial was a randomised controlled clinical trial evaluating the impact of a polypill strategy on adherence to indicated medication...

Using Observational Data to Calibrate Simulation Models.

Individual-level simulation models are valuable tools for comparing the impact of clinical or public health interventions on population health and cost outcomes over time. However, a key challenge is ensuring that outcome estimates correctly reflect real-world impacts. Calibration to targets obtained from randomized trials may be insufficient if trials do not exist for populations, time periods, or interventions of interest. Observational data can provide a wider range of calibration targets but requires methods to adjust for treatment-confounder feedback. We propose the use of the parametric g-formula to estimate calibration targets and present a case-study to demonstrate its application. We used the parametric g-formula applied to data from the HIV-CAUSAL Collaboration to estimate calibration targets for 7-y risks of AIDS and/or death (AIDS/death), as defined by the Center for Disease Control and Prevention under 3 treatment initiation strategies. We compared these targets to projections from the Cost-effectiveness of Preventing AIDS Complications (CEPAC) model for treatment-naïve individuals presenting to care in the following year ranges: 1996 to 1999, 2000 to 2002, or 2003 onwards. The parametric g-formula estimated a decreased risk of AIDS/death over time and with earlier treatment. The uncalibrated CEPAC model successfully reproduced targets obtained via the g-formula for baseline 1996 to 1999, but over-estimated calibration targets in contemporary populations and failed to reproduce time trends in AIDS/death risk. Calibration to g-formula targets improved CEPAC model fit for contemporary populations. Individual-level simulation models are developed based on best available information about disease processes in one or more populations of interest, but these processes can change over time or between populations. The parametric g-formula provides a method for using observational data to obtain valid calibration targets and enables updating of simulation model inputs when randomized trials are not available.

The Influence of Spatial Configuration of Residential Area and Vector Populations on Dengue Incidence Patterns in an Individual-Level Transmission Model.

Dengue is a mosquito-borne infectious disease that is endemic in tropical and subtropical countries. Many individual-level simulation models have been developed to test hypotheses about dengue virus transmission. Often these efforts assume that human host and mosquito vector populations are randomly or uniformly distributed in the environment. Although, the movement of mosquitoes is affected by spatial configuration of buildings and mosquito populations are highly clustered in key buildings, little research has focused on the influence of the local built environment in dengue transmission models. We developed an agent-based model of dengue transmission in a village setting to test the importance of using realistic environments in individual-level models of dengue transmission. The results from one-way ANOVA analysis of simulations indicated that the differences between scenarios in terms of infection rates as well as serotype-specific dominance are statistically significant. Specifically, the infection rates in scenarios of a realistic environment are more variable than those of a synthetic spatial configuration. With respect to dengue serotype-specific cases, we found that a single dengue serotype is more often dominant in realistic environments than in synthetic environments. An agent-based approach allows a fine-scaled analysis of simulated dengue incidence patterns. The results provide a better understanding of the influence of spatial heterogeneity on dengue transmission at a local scale.

Forecasting the burden of type 2 diabetes in Singapore using a demographic epidemiological model of Singapore

ObjectiveSingapore is a microcosm of Asia as a whole, and its rapidly ageing, increasingly sedentary population heralds the chronic health problems other Asian countries are starting to face and will...

Keeping the noise down: common random numbers for disease simulation modeling

Disease simulation models are used to conduct decision analyses of the comparative benefits and risks associated with preventive and treatment strategies. To address increasing model complexity and computational intensity, modelers use variance reduction techniques to reduce stochastic noise and improve computational efficiency. One technique, common random numbers, further allows modelers to conduct counterfactual-like analyses with direct computation of statistics at the individual level. This technique uses synchronized random numbers across model runs to induce correlation in model output thereby making differences easier to distinguish as well as simulating identical individuals across model runs. We provide a tutorial introduction and demonstrate the application of common random numbers in an individual-level simulation model of the epidemiology of breast cancer.

Monte Carlo probabilistic sensitivity analysis for patient level simulation models: efficient estimation of mean and variance using ANOVA

Probabilistic sensitivity analysis (PSA) is required to account for uncertainty in cost-effectiveness calculations arising from health economic models. The simplest way to perform PSA in practice is by Monte Carlo methods, which involves running the model many times using randomly sampled values of the model inputs. However, this can be impractical when the economic model takes appreciable amounts of time to run. This situation arises, in particular, for patient-level simulation models (also known as micro-simulation or individual-level simulation models), where a single run of the model simulates the health care of many thousands of individual patients. The large number of patients required in each run to achieve accurate estimation of cost-effectiveness means that only a relatively small number of runs is possible. For this reason, it is often said that PSA is not practical for patient-level models. We develop a way to reduce the computational burden of Monte Carlo PSA for patient-level models, based on the algebra of analysis of variance. Methods are presented to estimate the mean and variance of the model output, with formulae for determining optimal sample sizes. The methods are simple to apply and will typically reduce the computational demand very substantially.

Simulation Modeling in Political Science

Simulation research has made some notable contributions in political science. This article describes a variety of simulation projects and points out the strengths and weaknesses of simulation in contrast with other methods of research. The strength of simulation research is that it is naturally suited to modeling projects that include a large number of autonomous, interacting agents. Statistical and formal methods of analysis have made contributions in these areas, but there is good reason to believe that simulation can go further. This point is explored in several contexts, including social choice theory, individual-level simulation models, international relations, the prisoner's dilemma game, and more general agent-based models of multiperson interaction.