Emulator-based Bayesian calibration of the CISNET colorectal cancer models.

Purpose: 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. Methods: 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 (ANN) 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. Results: The optimal ANN for SimCRC had four hidden layers and 360 hidden nodes, MISCAN-Colon had 4 hidden layers and 114 hidden nodes, and CRC-SPIN had one hidden layer and 140 hidden nodes. The total time for training and calibrating the emulators was 7.3, 4.0, and 0.66 hours 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. Conclusions: 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, like the CISNET CRC models. In this work, we present a step-by-step guide to constructing emulators for calibrating three realistic CRC individual-level models using a Bayesian approach.

Approaches to developing de novo cancer population models to examine questions about cancer and race in bladder, gastric, and endometrial cancer and multiple myeloma: the Cancer Intervention and Surveillance Modeling Network incubator program.

BACKGROUND: We are developing 10 de novo population-level mathematical models in 4 malignancies (multiple myeloma and bladder, gastric, and uterine cancers). Each of these sites has documented disparities in outcome that are believed to be downstream effects of systemic racism. METHODS: Ten models are being independently developed as part of the Cancer Intervention and Surveillance Modeling Network incubator program. These models simulate trends in cancer incidence, early diagnosis, treatment, and mortality for the general population and are stratified by racial subgroup. Model inputs are based on large population datasets, clinical trials, and observational studies. Some core parameters are shared, and other parameters are model specific. All models are microsimulation models that use self-reported race to stratify model inputs. They can simulate the distribution of relevant risk factors (eg, smoking, obesity) and insurance status (for multiple myeloma and uterine cancer) in US birth cohorts and population. DISCUSSION: The models aim to refine approaches in prevention, detection, and management of 4 cancers given uncertainties and constraints. They will help explore whether the observed racial disparities are explainable by inequities, assess the effects of existing and potential cancer prevention and control policies on health equity and disparities, and identify policies that balance efficiency and fairness in decreasing cancer mortality.

Characterization and Valuation of the Uncertainty of Calibrated Parameters in Microsimulation Decision Models.

Background: We evaluated the implications of different approaches to characterize the uncertainty of calibrated parameters of microsimulation decision models (DMs) and quantified the value of such uncertainty in decision making. Methods: We calibrated the natural history model of CRC to simulated epidemiological data with different degrees of uncertainty and obtained the joint posterior distribution of the parameters using a Bayesian approach. We conducted a probabilistic sensitivity analysis (PSA) on all the model parameters with different characterizations of the uncertainty of the calibrated parameters. We estimated the value of uncertainty of the various characterizations with a value of information analysis. We conducted all analyses using high-performance computing resources running the Extreme-scale Model Exploration with Swift (EMEWS) framework. Results: The posterior distribution had a high correlation among some parameters. The parameters of the Weibull hazard function for the age of onset of adenomas had the highest posterior correlation of -0.958. When comparing full posterior distributions and the maximum-a-posteriori estimate of the calibrated parameters, there is little difference in the spread of the distribution of the CEA outcomes with a similar expected value of perfect information (EVPI) of $653 and $685, respectively, at a willingness-to-pay (WTP) threshold of $66,000 per quality-adjusted life year (QALY). Ignoring correlation on the calibrated parameters' posterior distribution produced the broadest distribution of CEA outcomes and the highest EVPI of $809 at the same WTP threshold. Conclusion: Different characterizations of the uncertainty of calibrated parameters affect the expected value of eliminating parametric uncertainty on the CEA. Ignoring inherent correlation among calibrated parameters on a PSA overestimates the value of uncertainty.

Methods for Communicating the Impact of Parameter Uncertainty in a Multiple-Strategies Cost-Effectiveness Comparison.

PURPOSE: Analyzing and communicating uncertainty is essential in medical decision making. To judge whether risks are acceptable, policy makers require information on the expected outcomes but also on the uncertainty and potential losses related to the chosen strategy. We aimed to compare methods used to represent the impact of uncertainty in decision problems involving many strategies, enhance existing methods, and provide an open-source and easy-to-use tool. METHODS: We conducted a systematic literature search to identify methods used to represent the impact of uncertainty in cost-effectiveness analyses comparing multiple strategies. We applied the identified methods to probabilistic sensitivity analysis outputs of 3 published decision-analytic models comparing multiple strategies. Subsequently, we compared the following characteristics: type of information conveyed, use of a fixed or flexible willingness-to-pay threshold, output interpretability, and the graphical discriminatory ability. We further proposed adjustments and integration of methods to overcome identified limitations of existing methods. RESULTS: The literature search resulted in the selection of 9 methods. The 3 methods with the most favorable characteristics to compare many strategies were 1) the cost-effectiveness acceptability curve (CEAC) and cost-effectiveness acceptability frontier (CEAF), 2) the expected loss curve (ELC), and 3) the incremental benefit curve (IBC). The information required to assess confidence in a decision often includes the average loss and the probability of cost-effectiveness associated with each strategy. Therefore, we proposed the integration of information presented in an ELC and CEAC into a single heat map. CONCLUSIONS: This article presents an overview of methods presenting uncertainty in multiple-strategy cost-effectiveness analyses, with their strengths and shortcomings. We proposed a heat map as an alternative method that integrates all relevant information required for health policy and medical decision making. HIGHLIGHTS: To assess confidence in a chosen course of action, decision makers require information on both the probability and the consequences of making a wrong decision.This article contains an overview of methods for presenting uncertainty in multiple-strategy cost-effectiveness analyses.We propose a heat map that combines the probability of cost-effectiveness from the cost-effectiveness acceptability curve (CEAC) with the consequences of a wrong decision from the expected loss curve.Collapsing of the CEAC can be reduced by relaxing the CEAC, as proposed in this article.Code in Microsoft Excel and R is provided to easily analyze data using the methods discussed in this article.

CDX2 Biomarker Testing and Adjuvant Therapy for Stage II Colon Cancer: An Exploratory Cost-Effectiveness Analysis.

OBJECTIVES: Adjuvant chemotherapy is not recommended for patients with average-risk stage II (T3N0) colon cancer. Nevertheless, a subgroup of these patients who are CDX2-negative might benefit from adjuvant chemotherapy. We evaluated the cost-effectiveness of testing for the absence of CDX2 expression followed by adjuvant chemotherapy (fluorouracil combined with oxaliplatin [FOLFOX]) for patients with stage II colon cancer. METHODS: We developed a decision model to simulate a hypothetical cohort of 65-year-old patients with average-risk stage II colon cancer with 7.2% of these patients being CDX2-negative under 2 different interventions: (1) test for the absence of CDX2 expression followed by adjuvant chemotherapy for CDX2-negative patients and (2) no CDX2 testing and no adjuvant chemotherapy for any patient. We derived disease progression parameters, adjuvant chemotherapy effectiveness and utilities from published analyses, and cancer care costs from the Surveillance, Epidemiology, and End Results (SEER)-Medicare data. Sensitivity analyses were conducted. RESULTS: Testing for CDX2 followed by FOLFOX for CDX2-negative patients had an incremental cost-effectiveness ratio of $5500/quality-adjusted life-years (QALYs) compared with no CDX2 testing and no FOLFOX (6.874 vs 6.838 discounted QALYs and $89 991 vs $89 797 discounted US dollar lifetime costs). In sensitivity analyses, considering a cost-effectiveness threshold of $100 000/QALY, testing for CDX2 followed by FOLFOX on CDX2-negative patients remains cost-effective for hazard ratios of <0.975 of the effectiveness of FOLFOX in CDX2-negative patients in reducing the rate of developing a metastatic recurrence. CONCLUSIONS: Testing tumors of patients with stage II colon cancer for CDX2 and administration of adjuvant treatment to the subgroup found CDX2-negative is a cost-effective and high-value management strategy across a broad range of plausible assumptions.

Age-specific rates of onset of cannabis use in Mexico.

BACKGROUND: Over the previous two decades, the lifetime prevalence of cannabis use has risen among Mexico's population. AIMS: Estimate the sex- and age-specific rates of onset of cannabis use over time. DESIGN: Five nationally representative cross-sectional surveys, the Mexican National Surveys of Addictions (1998, 2002, 2008, 2012) and the Mexican National Survey on Drugs, Alcohol, and Tobacco Consumption (2016). SETTING: Mexico. PARTICIPANTS: Pooled sample of 141,342 respondents aged between 12 and 65 years of which 43.6%(n = 61,658) are male and 56.4% (n = 79,684) are female. MEASUREMENTS: We estimated the age-specific rates of onset of cannabis as the conditional rate of consuming cannabis for the first time at a specific age. METHODS: Time-to-event flexible-parametric models with spline specifications of the hazard function. Stratified analysis by sex and control for temporal trends by year of data collection or decennial birth cohort. FINDINGS: Age-specific rates of onset of cannabis use per 1,000 individuals increased over time for females and males. Peak rates of onset of cannabis use per 1,000 ranged from 0.935 (95%CI = [0.772, 1.148]) in 1998, to 5.391 (95%CI = [4.924, 5.971]) in 2016 for females; and from 7.513 (95%CI = [6.732, 10.063]) in 1998, to 26.107 (95%CI = [25.918,30.654]) in 2016 for males. Across decennial birth-cohorts, peak rates of onset of cannabis use per 1,000 individuals for females ranged from 0.234 (95%CI = [0.078, 0.768]) for those born in the 1930s, to 14.611 (95%CI = [13.243, 16.102]) for those born in the 1990s; and for males, from 4.086 (95%CI = [4.022, 7.857]) for those born in the 1930s, to 38.693 (95%CI = [24.847, 48.670]) for those born in the 1990s. CONCLUSION: Rates of onset of cannabis increased over the previous two decades for both females and males but remained higher for males. Across recent cohorts, the rates of onset have increased at a faster rate among females than males.

BayCANN: Streamlining Bayesian Calibration With Artificial Neural Network Metamodeling.

Purpose: Bayesian calibration is generally superior to standard direct-search algorithms in that it estimates the full joint posterior distribution of the calibrated parameters. However, there are many barriers to using Bayesian calibration in health decision sciences stemming from the need to program complex models in probabilistic programming languages and the associated computational burden of applying Bayesian calibration. In this paper, we propose to use artificial neural networks (ANN) as one practical solution to these challenges. Methods: Bayesian Calibration using Artificial Neural Networks (BayCANN) involves (1) training an ANN metamodel on a sample of model inputs and outputs, and (2) then calibrating the trained ANN metamodel instead of the full model in a probabilistic programming language to obtain the posterior joint distribution of the calibrated parameters. We illustrate BayCANN using a colorectal cancer natural history model. We conduct a confirmatory simulation analysis by first obtaining parameter estimates from the literature and then using them to generate adenoma prevalence and cancer incidence targets. We compare the performance of BayCANN in recovering these "true" parameter values against performing a Bayesian calibration directly on the simulation model using an incremental mixture importance sampling (IMIS) algorithm. Results: We were able to apply BayCANN using only a dataset of the model inputs and outputs and minor modification of BayCANN's code. In this example, BayCANN was slightly more accurate in recovering the true posterior parameter estimates compared to IMIS. Obtaining the dataset of samples, and running BayCANN took 15 min compared to the IMIS which took 80 min. In applications involving computationally more expensive simulations (e.g., microsimulations), BayCANN may offer higher relative speed gains. Conclusions: BayCANN only uses a dataset of model inputs and outputs to obtain the calibrated joint parameter distributions. Thus, it can be adapted to models of various levels of complexity with minor or no change to its structure. In addition, BayCANN's efficiency can be especially useful in computationally expensive models. To facilitate BayCANN's wider adoption, we provide BayCANN's open-source implementation in R and Stan.

Comparing the Cost-Effectiveness of Innovative Colorectal Cancer Screening Tests.

BACKGROUND: Colorectal cancer (CRC) screening with colonoscopy and the fecal immunochemical test (FIT) is underused. Innovative tests could increase screening acceptance. This study determined which of the available alternatives is most promising from a cost-effectiveness perspective. METHODS: The previously validated Microsimulation Screening Analysis-Colon model was used to evaluate the cost-effectiveness of screening with capsule endoscopy every 5 or 10 years, computed tomographic colonography every 5 years, the multi-target stool DNA test every 1 or 3 years, and the methylated SEPT9 DNA plasma assay (mSEPT9) every 1 or 2 years. We also compared these strategies with annual FIT screening and colonoscopy screening every 10 years. Quality-adjusted life-years gained (QALYG), number of colonoscopies, and incremental cost-effectiveness ratios were projected. We assumed a willingness-to-pay threshold of $100 000 per QALYG. RESULTS: Among the alternative tests, computed tomographic colonography every 5 years, annual mSEPT9, and annual multi-target stool DNA screening had incremental cost-effectiveness ratios of $1092, $63 253, and $214 974 per QALYG, respectively. Other screening strategies were more costly and less effective than (a combination of) these 3. Under the assumption of perfect adherence, annual mSEPT9 screening resulted in more QALYG, CRC cases averted, and CRC deaths averted than annual FIT screening but led to a high rate of colonoscopy referral (51% after 3 years, 69% after 5 years). The alternative tests were not cost-effective compared with FIT and colonoscopy. CONCLUSIONS: This study suggests that for individuals not willing to participate in FIT or colonoscopy screening, mSEPT9 is the test of choice if the high colonoscopy referral rate is acceptable to them.

Estimating Population-Based Recurrence Rates of Colorectal Cancer over Time in the United States.

BACKGROUND: Population-based metastatic recurrence rates for patients diagnosed with nonmetastatic colorectal cancer cannot be estimated directly from population-based cancer registries because recurrence information is not reported. We derived population-based colorectal cancer recurrence rates using disease-specific survival data based on our understanding of the colorectal cancer recurrence-death process. METHODS: We used a statistical continuous-time multistate survival model to derive population-based annual colorectal cancer recurrence rates from 6 months to 10 years after colorectal cancer diagnosis using relative survival data from the Surveillance, Epidemiology, and End Results Program. The model was based on the assumption that, after 6 months of diagnosis, all colorectal cancer-related deaths occur only in patients who experience a metastatic recurrence first, and that the annual colorectal cancer-specific death rate among patients with recurrence was the same as in those diagnosed with de novo metastatic disease. We allowed recurrence rates to vary by post-diagnosis time, age, stage, and location for two diagnostic time periods. RESULTS: In patients diagnosed in 1975-1984, annual recurrence rates 6 months to 5 years after diagnosis ranged from 0.054 to 0.060 in stage II colon cancer, 0.094 to 0.105 in stage II rectal cancer, and 0.146 to 0.177 in stage III colorectal cancer, depending on age. We found a statistically significant decrease in colorectal cancer recurrence among patients diagnosed in 1994-2003 compared with those diagnosed in 1975-1984 for 6 months to 5 years after diagnosis (hazard ratios between 0.43 and 0.70). CONCLUSIONS: We derived population-based annual recurrence rates for up to 10 years after diagnosis using relative survival data. IMPACT: Our estimates can be used in decision-analytic models to facilitate analyses of colorectal cancer interventions that are more generalizable.

Discussing Cervical Cancer Screening Options: Outcomes to Guide Conversations Between Patients and Providers.

Purpose. In 2018, the US Preventive Services Task Force (USPSTF) endorsed three strategies for cervical cancer screening in women ages 30 to 65: cytology every 3 years, testing for high-risk types of human papillomavirus (hrHPV) every 5 years, and cytology plus hrHPV testing (co-testing) every 5 years. It further recommended that women discuss with health care providers which testing strategy is best for them. To inform such discussions, we used decision analysis to estimate outcomes of screening strategies recommended for women at age 30. Methods. We constructed a Markov decision model using estimates of the natural history of HPV and cervical neoplasia. We evaluated the three USPSTF-endorsed strategies, hrHPV testing every 3 years and no screening. Outcomes included colposcopies with biopsy, false-positive testing (a colposcopy in which no cervical intraepithelial neoplasia grade 2 or worse was found), treatments, cancers, and cancer mortality expressed per 10,000 women over a shorter-than-lifetime horizon (15-year). Results. All strategies resulted in substantially lower cancer and cancer death rates compared with no screening. Strategies with the lowest likelihood of cancer and cancer death generally had higher likelihood of colposcopy and false-positive testing. Conclusions. The screening strategies we evaluated involved tradeoffs in terms of benefits and harms. Because individual women may place different weights on these projected outcomes, the optimal choice for each woman may best be discerned through shared decision making.

Estimating the Natural History of Cervical Carcinogenesis Using Simulation Models: A CISNET Comparative Analysis.

BACKGROUND: The natural history of human papillomavirus (HPV)-induced cervical cancer (CC) is not directly observable, yet the age of HPV acquisition and duration of preclinical disease (dwell time) influences the effectiveness of alternative preventive policies. We performed a Cancer Intervention and Surveillance Modeling Network (CISNET) comparative modeling analysis to characterize the age of acquisition of cancer-causing HPV infections and implied dwell times for distinct phases of cervical carcinogenesis. METHODS: Using four CISNET-cervical models with varying underlying structures but fit to common US epidemiological data, we estimated the age of acquisition of causal HPV infections and dwell times associated with three phases of cancer development: HPV, high-grade precancer, and cancer sojourn time. We stratified these estimates by HPV genotype under both natural history and CC screening scenarios, because screening prevents cancer development that affects the mix of detected cancers. RESULTS: The median time from HPV acquisition to cancer detection ranged from 17.5 to 26.0 years across the four models. Three models projected that 50% of unscreened women acquired their causal HPV infection between ages 19 and 23 years, whereas one model projected these infections occurred later (age 34 years). In the context of imperfect compliance with US screening guidelines, the median age of causal infection was 4.4-15.9 years later compared with model projections in the absence of screening. CONCLUSIONS: These validated CISNET-CC models, which reflect some uncertainty in the development of CC, elucidate important drivers of HPV vaccination and CC screening policies and emphasize the value of comparative modeling when evaluating public health policies.

A Value of Information Analysis of Research on the 21-Gene Assay for Breast Cancer Management.

OBJECTIVES: The 21-gene assay Oncotype DX (21-GA) shows promise as a guide in deciding when to initiate adjuvant chemotherapy in women with hormone receptor-positive early-stage breast cancer. Nevertheless, its routine use remains controversial, owing to insufficient evidence of its clinical utility and cost-effectiveness. Accordingly, we aim to quantify the value of conducting further research to reduce decision uncertainty in the use of the 21-GA. METHODS: Using value of information methods, we first generated probability distributions of survival and costs for decision making with and without the 21-GA alongside traditional risk prediction. These served as the input to a comparison of 3 alternative study designs: a retrospective observational study to update risk classification from the 21-GA, a prospective observational study to estimate prevalence of chemotherapy use, and a randomized controlled trial (RCT) of the 21-GA predictive value. RESULTS: We found that current evidence strongly supports the use of the 21-GA in intermediate- and high-risk women. Further research should focus on low-risk women, among whom the cost-effectiveness findings remained equivocal. For this population, we identified a high value of reducing uncertainty in the 21-GA use for all proposed research studies. The RCT had the greatest potential to efficiently reduce the likelihood of choosing a suboptimal strategy, providing a value between $162 million and $1.1 billion at willingness-to-pay thresholds of $150 000 to $200 000/quality-adjusted life years. CONCLUSION: Future research to inform 21-GA decision making is of high value. The RCT of the 21-GA predictive value has the greatest potential to efficiently reduce decision uncertainty around 21-GA use in women with low-risk early-stage breast cancer.

A Cost-effectiveness Analysis of Systemic Therapy for Metastatic Hormone-sensitive Prostate Cancer.

BACKGROUND: Following the recent publication of results from randomized trials that have demonstrated a survival benefit for the addition of docetaxel or abiraterone acetate to androgen deprivation therapy (ADT) in the treatment of metastatic prostate cancer, it is important to assess whether the benefits of treatment with these agents outweigh their costs. OBJECTIVE: To perform a cost-effectiveness analysis of immediate docetaxel or abiraterone acetate treatment in addition to ADT in men with metastatic hormone-sensitive prostate cancer (PC). DESIGN, SETTING, AND PARTICIPANTS: We developed a state-transition model to simulate the natural progression of metastatic PC. Model parameters were derived from the published literature and through calibration to observed epidemiological data. Following diagnosis, a hypothetical cohort of men with metastatic hormone-sensitive PC could be treated with docetaxel+ADT, abiraterone+ADT, or ADT alone. Once disease progresses to castration-resistant PC, treatment with one of the approved therapies in this setting was initiated. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: The outcomes measured were quality-adjusted life years (QALYs) and costs from a US private payer, health sector perspective. RESULTS AND LIMITATIONS: Compared to treatment with ADT alone, docetaxel and abiraterone resulted in a discounted quality-adjusted survival gain of 3.6 and 22.0mo, respectively. Using standard cost-effectiveness metrics, treatment with docetaxel and ADT provides high value for money (ie, is cost effective) with an incremental cost-effectiveness ratio (ICER) of $34723, compared to an ICER of $295212 for abiraterone. The monthly cost of abiraterone would have to be less than $3114 for it to be cost effective. CONCLUSIONS: Docetaxel+ADT is likely the most cost-effective treatment option for men with metastatic hormone-sensitive PC. Although potentially more effective than docetaxel, the costs of abiraterone would have to be considerably lower to match the cost effectiveness of docetaxel+ADT. PATIENT SUMMARY: This study evaluated the balance of costs and benefits for treatment of metastatic hormone-sensitive prostate cancer with docetaxel plus androgen deprivation therapy (ADT), abiraterone plus ADT, or ADT alone. We found that treatment with docetaxel plus ADT likely represents the most cost-effective option in this setting.

Estimated Quality of Life and Economic Outcomes Associated With 12 Cervical Cancer Screening Strategies: A Cost-effectiveness Analysis.

Importance: Many cervical cancer screening strategies are now recommended in the United States, but the benefits, harms, and costs of each option are unclear. Objective: To estimate the cost-effectiveness of 12 cervical cancer screening strategies. Design, Setting, and Participants: The cross-sectional portion of this study enrolled a convenience sample of 451 English-speaking or Spanish-speaking women aged 21 to 65 years from September 22, 2014, to June 16, 2016, identified at women's health clinics in San Francisco. In this group, utilities (preferences) were measured for 23 cervical cancer screening-associated health states and were applied to a decision model of type-specific high-risk human papillomavirus (hrHPV)-induced cervical carcinogenesis. Test accuracy estimates were abstracted from systematic reviews. The evaluated strategies were cytologic testing every 3 years for women aged 21 to 65 years with either repeat cytologic testing in 1 year or immediate hrHPV triage for atypical squamous cells of undetermined significance (ASC-US), cytologic testing every 3 years for women age 21 to 29 years followed by cytologic testing plus hrHPV testing (cotesting), or primary hrHPV testing alone for women aged 30 to 65 years. Screening frequency, abnormal test result management, and the age to switch from cytologic testing to hrHPV testing (25 or 30 years) were varied. Analyses were conducted from both the societal and health care sector perspectives. Main Outcomes and Measures: Utilities for 23 cervical cancer screening-associated health states (cross-sectional study) and quality-adjusted life-years (QALYs) and total costs for each strategy. Results: Utilities were measured in a sociodemographically diverse group of 451 women (mean [SD] age, 38.2 [10.7] years; 258 nonwhite [57.2%]). Cytologic testing every 3 years with repeat cytologic testing for ASC-US yielded the most lifetime QALYs and conferred more QALYs at higher costs ($2166 per QALY) than the lowest-cost strategy (cytologic testing every 3 years with hrHPV triage of ASC-US). All cytologic testing plus hrHPV testing (cotesting) and primary hrHPV testing strategies provided fewer QALYs at higher costs. Adding indirect costs did not change the conclusions. In sensitivity analyses, hrHPV testing every 5 years with genotyping triage beginning at age 30 years was the lowest-cost strategy when hrHPV test sensitivity was markedly higher than cytologic test sensitivity or when hrHPV test cost was equated to the lowest reported cytologic test cost ($14). Conclusions and Relevance: Cytologic testing every 3 years for women aged 21 to 29 years with either continued cytologic testing every 3 years or switching to a low-cost hrHPV test every 5 years confers a reasonable balance of benefits, harms, and costs. Comparative modeling is needed to confirm the association of these novel utilities with cost-effectiveness.

Cost-effectiveness Analysis of Active Surveillance Strategies for Men with Low-risk Prostate Cancer.

BACKGROUND: Active surveillance (AS) has become the recommended management strategy for men with low-risk prostate cancer. However, there is considerable uncertainty about the optimal follow-up schedule in terms of the tests to perform and their frequency. OBJECTIVE: To assess the costs and benefits of different AS follow-up strategies compared to watchful waiting (WW) or immediate treatment. DESIGN, SETTING, AND PARTICIPANTS: A state-transition Markov model was developed to simulate the natural history (ie, no testing or intervention) of prostate cancer for a hypothetical cohort of 50-yr-old men newly diagnosed with low-risk prostate cancer. Following diagnosis, men were hypothetically managed with immediate treatment, watchful waiting, or one of several AS strategies. AS follow-up was performed either with transrectal ultrasound-guided biopsy or magnetic resonance imaging (MRI) which was scheduled annually, biennially, every 3yrs, according to the PRIAS protocol (yrs 1, 4, 7, and 10, and then every 5yr) or every 5yr. Diagnosis of higher-grade or -stage disease while on AS resulted in curative treatment. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: We measured discounted quality-adjusted life years (QALYs), discounted lifetime medical costs (2017 US$), and incremental cost-effectiveness ratios (ICERs). RESULTS AND LIMITATIONS: Compared to WW, MRI-based surveillance performed every 5yr improved quality-adjusted survival by 4.47 quality-adjusted months and represented high-value health care at the Medicare reimbursement rate using standard cost-effectiveness metrics. Biopsy-based strategies were less effective and less costly than the corresponding MRI-based strategies for each testing interval. MRI-based surveillance at more frequent intervals had ICERs greater than $800000 per QALY and would not be considered cost-effective according to standard metrics. Our results were sensitive to the diagnostic accuracy and costs of both biopsy modes in detecting clinically significant cancer. CONCLUSIONS: Incorporation of MRI into surveillance protocols at Medicare reimbursement rates and decreasing the intensity of repeat testing may be cost-effective options for men opting for conservative management of low-risk prostate cancer. PATIENT SUMMARY: Our study modeled outcomes for men with low-risk prostate cancer undergoing watchful waiting, immediate treatment, or active surveillance with different follow-up schedules. We found that conservative management of low-risk disease optimizes health outcomes and costs. Furthermore, we showed that decreasing the intensity of active surveillance follow-up and incorporating magnetic resonance imaging (MRI) into surveillance protocols can be cost-effective, depending on the MRI costs.

Nonidentifiability in Model Calibration and Implications for Medical Decision Making.

BACKGROUND: Calibration is the process of estimating parameters of a mathematical model by matching model outputs to calibration targets. In the presence of nonidentifiability, multiple parameter sets solve the calibration problem, which may have important implications for decision making. We evaluate the implications of nonidentifiability on the optimal strategy and provide methods to check for nonidentifiability. METHODS: We illustrate nonidentifiability by calibrating a 3-state Markov model of cancer relative survival (RS). We performed 2 different calibration exercises: 1) only including RS as a calibration target and 2) adding the ratio between the 2 nondeath states over time as an additional target. We used the Nelder-Mead (NM) algorithm to identify parameter sets that best matched the calibration targets. We used collinearity and likelihood profile analyses to check for nonidentifiability. We then estimated the benefit of a hypothetical treatment in terms of life expectancy gains using different, but equally good-fitting, parameter sets. We also applied collinearity analysis to a realistic model of the natural history of colorectal cancer. RESULTS: When only RS is used as the calibration target, 2 different parameter sets yield similar maximum likelihood values. The high collinearity index and the bimodal likelihood profile on both parameters demonstrated the presence of nonidentifiability. These different, equally good-fitting parameter sets produce different estimates of the treatment effectiveness (0.67 v. 0.31 years), which could influence the optimal decision. By incorporating the additional target, the model becomes identifiable with a collinearity index of 3.5 and a unimodal likelihood profile. CONCLUSIONS: In the presence of nonidentifiability, equally likely parameter estimates might yield different conclusions. Checking for the existence of nonidentifiability and its implications should be incorporated into standard model calibration procedures.

Incorporating Biomarkers into the Primary Prostate Biopsy Setting: A Cost-Effectiveness Analysis.

PURPOSE: We performed a cost-effectiveness analysis using the PHI (Prostate Health Index), 4Kscore®, SelectMDx™ and the EPI (ExoDx™ Prostate [IntelliScore]) in men with elevated prostate specific antigen to determine the need for biopsy. MATERIALS AND METHODS: We developed a decision analytical model in men with elevated prostate specific antigen (3 ng/ml or greater) in which 1 biomarker test was used to determine which hypothetical individuals required biopsy. In the current standard of care strategy all individuals underwent biopsy. Model parameters were derived from a comprehensive review of the literature. Costs were calculated from a health sector perspective and converted into 2017 United States dollars. RESULTS: The cost and QALYs (quality adjusted life-years) of the current standard of care, which was transrectal ultrasound guided biopsy, was $3,863 and 18.085, respectively. Applying any of the 3 biomarkers improved quality adjusted survival compared to the current standard of care. The cost of SelectMDx, the PHI and the EPI was lower than performing prostate biopsy in all patients. However, the PHI was more costly and less effective than the SelectMDx strategy. The EPI provided the highest QALY with an incremental cost-effectiveness ratio of $58,404 per QALY. The use of biomarkers could reduce the number of unnecessary biopsies by 24% to 34% compared to the current standard of care. CONCLUSIONS: Applying biomarkers in men with elevated prostate specific antigen to determine the need for biopsy improved quality adjusted survival by decreasing the number of biopsies performed and the treatment of indolent disease. Using SelectMDx or the EPI following elevated prostate specific antigen but before proceeding to biopsy is a cost-effective strategy in this setting.

A Gaussian Approximation Approach for Value of Information Analysis.

Most decisions are associated with uncertainty. Value of information (VOI) analysis quantifies the opportunity loss associated with choosing a suboptimal intervention based on current imperfect information. VOI can inform the value of collecting additional information, resource allocation, research prioritization, and future research designs. However, in practice, VOI remains underused due to many conceptual and computational challenges associated with its application. Expected value of sample information (EVSI) is rooted in Bayesian statistical decision theory and measures the value of information from a finite sample. The past few years have witnessed a dramatic growth in computationally efficient methods to calculate EVSI, including metamodeling. However, little research has been done to simplify the experimental data collection step inherent to all EVSI computations, especially for correlated model parameters. This article proposes a general Gaussian approximation (GA) of the traditional Bayesian updating approach based on the original work by Raiffa and Schlaifer to compute EVSI. The proposed approach uses a single probabilistic sensitivity analysis (PSA) data set and involves 2 steps: 1) a linear metamodel step to compute the EVSI on the preposterior distributions and 2) a GA step to compute the preposterior distribution of the parameters of interest. The proposed approach is efficient and can be applied for a wide range of data collection designs involving multiple non-Gaussian parameters and unbalanced study designs. Our approach is particularly useful when the parameters of an economic evaluation are correlated or interact.

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

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

Trade-offs Between Efficacy and Cardiac Toxicity of Adjuvant Chemotherapy in Early-Stage Breast Cancer Patients: Do Competing Risks Matter?

Evidence about treatment efficacy and long-term toxicities for adjuvant chemotherapy in patients with early-stage breast cancer is often presented in different formats and studies. This leads to challenges for patients and their physicians to adequately weigh the trade-offs between effectiveness and long-term cardiac toxicity when making decisions about adjuvant chemotherapy. We used a decision-analytic framework to quantify these trade-offs by combining the available evidence into a single, comparable metric. We developed a Markov model to simulate a hypothetical cohort of newly diagnosed breast cancer patients under three scenarios: no treatment, anthracycline (AC)-based adjuvant chemotherapy (more effective but also more cardiotoxic), and non-AC-based adjuvant chemotherapy. We derived the model parameters from medical literature (e.g., clinical trials). Our primary outcome is 10-year mortality, and other metrics such as cause of death; life years (LYs) and quality-adjusted LYs over 10 years were evaluated in sensitivity analysis. For 55-year-old women with a 10-year risk of metastatic recurrence <12.5% no chemotherapy resulted in the preferred strategy. In general, non-AC-based adjuvant chemotherapy resulted in lower 10-year mortality than AC-based chemotherapy. Patients with low risk of metastatic recurrence are better off without adjuvant chemotherapy regardless of the outcome considered (i.e., the risks of cardiac toxicity from chemotherapy outweighed the benefits). Trade-offs between effectiveness and induced cardiac toxicity impact health outcomes. The choice of adjuvant treatment must consider the patient's risk of distant recurrence and the quality of life associated with different health outcomes.