The Optimal Age to Stop Endoscopic Surveillance of Patients With Barrett's Esophagus Based on Sex and Comorbidity: A Comparative Cost-Effectiveness Analysis.

BACKGROUND AND AIMS: Current guidelines recommend surveillance for patients with nondysplastic Barrett's esophagus (NDBE) but do not include a recommended age for discontinuing surveillance. This study aimed to determine the optimal age for last surveillance of NDBE patients stratified by sex and level of comorbidity. METHODS: We used 3 independently developed models to simulate patients diagnosed with NDBE, varying in age, sex, and comorbidity level (no, mild, moderate, and severe). All patients had received regular surveillance until their current age. We calculated incremental costs and quality-adjusted life-years (QALYs) gained from 1 additional endoscopic surveillance at the current age versus not performing surveillance at that age. We determined the optimal age to end surveillance as the age at which incremental cost-effectiveness ratio of 1 more surveillance was just less than the willingness-to-pay threshold of $100,000/QALY. RESULTS: The benefit of having 1 more surveillance endoscopy strongly depended on age, sex, and comorbidity. For men with NDBE and severe comorbidity, 1 additional surveillance at age 80 years provided 4 more QALYs per 1000 patients with BE at an additional cost of $1.2 million, whereas for women with severe comorbidity the benefit at that age was 7 QALYs at a cost of $1.3 million. For men with no, mild, moderate, and severe comorbidity, the optimal ages of last surveillance were 81, 80, 77, and 73 years, respectively. For women, these ages were younger: 75, 73, 73, and 69 years, respectively. CONCLUSIONS: Our comparative modeling analysis illustrates the importance of considering comorbidity status and sex when deciding on the age to discontinue surveillance in patients with NDBE.

Genetic testing to guide screening for pancreatic ductal adenocarcinoma: Results of a microsimulation model.

BACKGROUND: First-degree relatives (FDRs) of patients with pancreatic ductal adenocarcinoma (PDAC) have elevated PDAC risk, partially due to germline genetic variants. We evaluated the potential effectiveness of genetic testing to target MRI-based screening among FDRs. METHODS: We used a microsimulation model of PDAC, calibrated to Surveillance, Epidemiology, and End Results (SEER) data, to estimate the potential life expectancy (LE) gain of screening for each of the following groups of FDRs: individuals who test positive for each of eight variants associated with elevated PDAC risk (e.g., BRCA2, CDKN2A); individuals who test negative; and individuals who do not test. Screening was assumed to take place if LE gains were achievable. We simulated multiple screening approaches, defined by starting age and frequency. Sensitivity analysis evaluated changes in results given varying model assumptions. RESULTS: For women, 92% of mutation carriers had projected LE gains from screening for PDAC, if screening strategies (start age, frequency) were optimized. Among carriers, LE gains ranged from 0.1 days (ATM+ women screened once at age 70) to 510 days (STK11+ women screened annually from age 40). For men, LE gains were projected for all mutation carriers, ranging from 0.2 days (BRCA1+ men screened once at age 70) to 620 days (STK11+ men screened annually from age 40). For men and women who did not undergo genetic testing, or for whom testing showed no variant, screening yielded small LE benefit (0-2.1 days). CONCLUSIONS: Genetic testing of FDRs can inform targeted PDAC screening by identifying which FDRs may benefit.

Colorectal Cancer Screening: An Updated Modeling Study for the US Preventive Services Task Force.

Importance: The US Preventive Services Task Force (USPSTF) is updating its 2016 colorectal cancer screening recommendations. Objective: To provide updated model-based estimates of the benefits, burden, and harms of colorectal cancer screening strategies and to identify strategies that may provide an efficient balance of life-years gained (LYG) from screening and colonoscopy burden to inform the USPSTF. Design, Setting, and Participants: Comparative modeling study using 3 microsimulation models of colorectal cancer screening in a hypothetical cohort of 40-year-old US individuals at average risk of colorectal cancer. Exposures: Screening from ages 45, 50, or 55 years to ages 70, 75, 80, or 85 years with fecal immunochemical testing (FIT), multitarget stool DNA testing, flexible sigmoidoscopy alone or with FIT, computed tomography colonography, or colonoscopy. All persons with an abnormal noncolonoscopy screening test result were assumed to undergo follow-up colonoscopy. Screening intervals varied by test. Full adherence with all procedures was assumed. Main Outcome and Measures: Estimated LYG relative to no screening (benefit), lifetime number of colonoscopies (burden), number of complications from screening (harms), and balance of incremental burden and benefit (efficiency ratios). Efficient strategies were those estimated to require fewer additional colonoscopies per additional LYG relative to other strategies. Results: Estimated LYG from screening strategies ranged from 171 to 381 per 1000 40-year-olds. Lifetime colonoscopy burden ranged from 624 to 6817 per 1000 individuals, and screening complications ranged from 5 to 22 per 1000 individuals. Among the 49 strategies that were efficient options with all 3 models, 41 specified screening beginning at age 45. No single age to end screening was predominant among the efficient strategies, although the additional LYG from continuing screening after age 75 were generally small. With the exception of a 5-year interval for computed tomography colonography, no screening interval predominated among the efficient strategies for each modality. Among the strategies highlighted in the 2016 USPSTF recommendation, lowering the age to begin screening from 50 to 45 years was estimated to result in 22 to 27 additional LYG, 161 to 784 additional colonoscopies, and 0.1 to 2 additional complications per 1000 persons (ranges are across screening strategies, based on mean estimates across models). Assuming full adherence, screening outcomes and efficient strategies were similar by sex and race and across 3 scenarios for population risk of colorectal cancer. Conclusions and Relevance: This microsimulation modeling analysis suggests that screening for colorectal cancer with stool tests, endoscopic tests, or computed tomography colonography starting at age 45 years provides an efficient balance of colonoscopy burden and life-years gained.

Cost-Effectiveness Analysis of Screening for Pancreatic Cancer Among High-Risk Populations.

PURPOSE: We evaluated the potential cost-effectiveness of combined magnetic resonance imaging (MRI) and endoscopic ultrasound (EUS) screening for pancreatic ductal adenocarcinoma (PDAC) among populations at high risk for the disease. METHODS: We used a microsimulation model of the natural history of PDAC to estimate the lifetime health benefits, costs, and cost-effectiveness of PDAC screening among populations with specific genetic risk factors for PDAC, including BRCA1 and BRCA2, PALB2, ATM, Lynch syndrome, TP53, CDKN2A, and STK11. For each high-risk population, we simulated 29 screening strategies, defined by starting age and frequency. Screening included MRI with follow-up EUS in a subset of patients. Costs of tests were based on Medicare reimbursement for MRI, EUS, fine-needle aspiration biopsy, and pancreatectomy. Cancer-related cost by stage of disease and phase of treatment was based on the literature. For each high-risk population, we performed an incremental cost-effectiveness analysis, assuming a willingness-to-pay (WTP) threshold of $100,000 US dollars (USD) per quality-adjusted life year (QALY) gained. RESULTS: For men with relative risk (RR) 12.33 (CDKN2A) and RR 28 (STK11), annual screening was cost-effective, starting at age 55 and 40 years, respectively. For women, screening was only cost-effective for those with RR 28 (STK11), with annual screening starting at age 45 years. CONCLUSION: Combined MRI/EUS screening may be a cost-effective approach for the highest-risk populations (among mutations considered, those with RR >12). However, for those with moderate risk (RR, 5-12), screening would only be cost-effective at higher WTP thresholds (eg, $200K USD/QALY) or with once-only screening.

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.

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 (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. RESULTS: 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. 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, 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. HIGHLIGHTS: 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.

Ultrasound Surveillance of Small, Incidentally Detected Gallbladder Polyps: Projected Benefits by Sex, Age, and Comorbidity Level.

OBJECTIVE: Incidentally detected gallbladder polyps are commonly encountered when performing upper abdominal ultrasound. Our purpose was to estimate the life expectancy (LE) benefit of ultrasound-based gallbladder surveillance in patients with small (6-7 to <10 mm), incidentally detected gallbladder polyps, accounting for patient sex, age, and comorbidity level. METHODS: We developed a decision-analytic Markov model to evaluate hypothetical cohorts of women and men with small gallbladder polyps, with varying age (66-80 years) and comorbidity level (none, mild, moderate, severe). Drawing from current evidence, in the base case, we assumed no increased risk of gallbladder cancer in patients with small gallbladder polyps. To estimate maximal possible LE gains from surveillance, we assumed perfect cancer control consequent to 5 years of surveillance. We varied key assumptions including cancer risk and test performance characteristics in sensitivity analysis. RESULTS: Projected LE gains from surveillance were <3 days across most cohorts and scenarios evaluated. For 66- and 80-year-olds with no comorbidities, LE gains were 1.46 and 1.45 days, respectively, for women, and 0.67 and 0.75 days for men. With 10 years of surveillance, LE gains increased to 2.94 days for 66-year-old women with no comorbidities (men: 1.35 days). If we assumed a 10% increase in gallbladder cancer risk among individuals with polyps, LE gains increased slightly to 1.60 days for 66-year-old women with no comorbidities (men: 0.74 days). Results were sensitive to test performance and surgical mortality. DISCUSSION: Even under unrealistic, optimistic assumptions of cancer control, ultrasound surveillance of incidentally detected small gallbladder polyps provided limited benefit.

Surveillance in Older Women With Incidental Ovarian Cysts: Maximal Projected Benefits by Age and Comorbidity Level.

PURPOSE: The aim of this study was to estimate effects on life expectancy (LE) of imaging-based ovarian surveillance after detection of incidental postmenopausal ovarian cysts, under different assumptions of patient age, comorbidity level, and cancer risk and detection. METHODS: A decision-analytic Markov model was developed to estimate LE benefits. Hypothetical cohorts of postmenopausal women with simple ovarian cysts were evaluated, with varied age (66-80 years) and comorbidity level (none, mild, moderate, severe). For each cohort, imaging "follow-up" (2 years) and "no-follow-up" strategies were compared. Consistent with current evidence, increased cancer risk in patients with cysts was not assumed; however, incident ovarian cancers could be detected during follow-up. To estimate theoretical maximal LE gains from follow-up, perfect ovarian cancer detection and treatment during follow-up were assumed. This and other key assumptions were varied in sensitivity analysis. RESULTS: Projected LE gains from follow-up were limited. For 66-, 70-, 75-, and 80-year-old women with no comorbidities, LE gains were 5.1, 5.1, 4.5, and 3.7 days; with severe comorbidities, they were 3.5, 3.2, 2.7, and 2.1 days. With sensitivity of 50% for cancer detection, they were 3.7 days for 66-year-old women with no comorbidities and 1.3 days for 80-year-old women with severe comorbidities. When cancer risk for women with cysts was assumed to be elevated (1.1 times average risk), LE gains increased only modestly (5.6 and 2.3 days for analogous cohorts). CONCLUSIONS: Even in the circumstance of perfect ovarian cancer detection and treatment, surveillance of postmenopausal women (≥66 years of age) with simple cysts affords limited benefits, particularly in women with advanced age and comorbidities.

Identifying Barriers to Building a Diverse Physician Workforce: A National Survey of the ACR Membership.

PURPOSE: The aim of this study was to identify potential barriers to building a diverse workforce in radiology and radiation oncology by conducting a national survey of physicians in these fields and studying their reported career experiences. METHODS: An electronic survey of ACR members (February 27, 2018, to April 26, 2018) was conducted in which physicians' attitudes about their work environment, relationships, and culture were queried. The aim was to determine if responses differed by gender or race/ethnicity. In total, 900 invitations were issued; women were oversampled with the goal of equal representation. Descriptive summaries (proportions of yes or no responses) were calculated per item, per subgroup of interest. Logistic regression analysis was used to identify significant associations between gender- and item-specific responses; it was not used in the race/ethnicity analysis because of the small sizes of many subgroups. RESULTS: The response rate was 51.2% (461 of 900). In total, 51.0% of respondents identified as women (235 of 461); the 9.5% (44 of 461) who identified as black or African American, Hispanic, or American Indian or Alaska Native were considered underrepresented minorities. Respondents' mean age was 40.2 ± 10.4 years. Subgroups varied most in their reporting of unfair or disrespectful treatment. Women were significantly more likely than men to report such treatment attributable to gender (50.6% versus 5.4%; odds ratio, 18.00; 95% confidence interval, 9.29-34.86; P < .001), and 27.9% of underrepresented minorities compared with 2.6% of white non-Hispanic respondents reported such treatment attributable to race/ethnicity. CONCLUSIONS: Women and underrepresented minorities disproportionately experience unfair or disrespectful treatment in the workplace. Addressing this problem is likely to be critically important for improving workforce diversity.