Methods for Inclusive Underwriting of Breast Cancer Risk with Machine Learning and Innovative Algorithms.

INTRODUCTION: -Due to early detection and improved therapies, the prevalence of long-term breast cancer survivors is increasing. This has increased the need for more inclusive underwriting in individuals with a history of breast cancer. Herein, we developed a method using algorithm aiming facilitating the underwriting of multiple parameters in breast cancer survivors. METHODS: -Variables and data were extracted from the SEER database and analyzed using 4 different machine learning based algorithms (Logistic Regression, GA2M, Random Forest, and XGBoost) that were compared with Kaplan Meier survival estimates. The performances of these algorithms have been compared with multiple metrics (Log Loss, AUC, and SMR). In situ (non-invasive) and metastatic breast cancer were excluded from this analysis. RESULTS: -Parameters included the pathological subtype, pTNM staging (T: tumor size, N; number of nodes; M presence or absence of metastases), Scarff-Bloom-Richardson grading, the expression of estrogen and progesterone hormone receptors were selected to predict the individual outcome at any time point from diagnosis. While all models had identical performance in terms of statistical metrics (AUC, Log Loss, and SMR), the logistic regression was the one and only model that respects all business constraints and was intelligible for medical and underwriting users. CONCLUSION: -This study provides insight to develop algorithms to set underwriter-friendly calculators for more accurate risk estimations that can be used to rationalize insurance pricing for breast cancer survivors. This study supports the development of a more inclusive underwriting based on models that can encompass the heterogeneity of several malignancies such as breast cancer.

Simulation exercises and after action reviews - analysis of outputs during 2016-2019 to strengthen global health emergency preparedness and response.

BACKGROUND: Under the International Health Regulations (2005) [IHR (2005)] Monitoring and Evaluation Framework, after action reviews (AAR) and simulation exercises (SimEx) are two critical components which measure the functionality of a country's health emergency preparedness and response under a "real-life" event or simulated situation. The objective of this study was to describe the AAR and SimEx supported by the World Health Organization (WHO) globally in 2016-2019. METHODS: In 2016-2019, WHO supported 63 AAR and 117 SimEx, of which 42 (66.7%) AAR reports and 56 (47.9%) SimEx reports were available. We extracted key information from these reports and created two central databases for AAR and SimEx, respectively. We conducted descriptive analysis and linked the findings according to the 13 IHR (2005) core capacities. RESULTS: Among the 42 AAR and 56 SimEx available reports, AAR and SimEx were most commonly conducted in the WHO African Region (AAR: n = 32, 76.2%; SimEx: n = 32, 52.5%). The most common public health events reviewed or tested in AAR and SimEx, respectively, were epidemics and pandemics (AAR: n = 38, 90.5%; SimEx: n = 46, 82.1%). For AAR, 10 (76.9%) of the 13 IHR core capacities were reviewed at least once, with no AAR conducted for food safety, chemical events, and radiation emergencies, among the reports available. For SimEx, all 13 (100.0%) IHR capacities were tested at least once. For AAR, the most commonly reviewed IHR core capacities were health services provision (n = 41, 97.6%), risk communication (n = 39, 92.9%), national health emergency framework (n = 39, 92.9%), surveillance (n = 37, 88.1%) and laboratory (n = 35, 83.3%). For SimEx, the most commonly tested IHR core capacity were national health emergency framework (n = 56, 91.1%), followed by risk communication (n = 48, 85.7%), IHR coordination and national IHR focal point functions (n = 45, 80.4%), surveillance (n = 31, 55.4%), and health service provision (n = 29, 51.8%). For AAR, the median timeframe between the end of the event and AAR was 125 days (range = 25-399 days). CONCLUSIONS: WHO has recently published guidance for the planning, execution, and follow-up of AAR and SimEx. Through the guidance and the simplified reporting format provided, we hope to see more countries conduct AAR and SimEx and standardization in their methodology, practice, reporting and follow-up.

The relationship between mortality and daily number of steps in type 2 diabetes.

BACKGROUND: Physical activity (PA) is an established modifiable factor for the prevention of cardiovascular disease. Our objective was to assess the association of PA with mortality rates in a national sample of patients with diabetes. METHODS: We analyzed a nationally representative sample from The National Health and Nutrition Examination Survey (NHANES, periods 2003-2004 and 2005-2006) that used PA Monitors. Individuals were matched for BMI, number of steps/per day and age. Three groups were created: subjects with less than 5000/steps per day (low), 5000-7500/steps per day (moderate) and more than 7500/steps per day (high levels of physical activity). All-cause mortality was ascertained through December 2015. RESULTS: A sample of 3072 individuals (1018 with diabetes) was analyzed. Patients with diabetes had 30% increased risk of mortality of all causes (RR: 1.298, 95% CI [1.162-1.451], P<0.001), higher levels of PA (>7500 steps/day) provided similar relative risk for subjects with diabetes compared to their controls (RR:1.256 [95% CI 0.910-1.732]). In a Poisson model adjusted for sex, history of previous cardiovascular event or cancer, ethnicity, Hb1ac, SBP, and total cholesterol to HDL ratio, patients with diabetes and moderate or high PA had an associated 44% to 80% lower risk of all-cause mortality compared to those with low PA. CONCLUSIONS: The subgroup of patients with diabetes and high PA had no excess of mortality compared to the general population. PA can reduce the gap for all-cause mortality, used as an index of cardiovascular fitness and a clinical tool for the assessment of mortality risk.

Causal Inference in the Presence of Interference in Sponsored Search Advertising.

In classical causal inference, inferring cause-effect relations from data relies on the assumption that units are independent and identically distributed. This assumption is violated in settings where units are related through a network of dependencies. An example of such a setting is ad placement in sponsored search advertising, where the likelihood of a user clicking on a particular ad is potentially influenced by where it is placed and where other ads are placed on the search result page. In such scenarios, confounding arises due to not only the individual ad-level covariates but also the placements and covariates of other ads in the system. In this paper, we leverage the language of causal inference in the presence of interference to model interactions among the ads. Quantification of such interactions allows us to better understand the click behavior of users, which in turn impacts the revenue of the host search engine and enhances user satisfaction. We illustrate the utility of our formalization through experiments carried out on the ad placement system of the Bing search engine.

Lessons Learned From a Large Cross-Border Field Simulation Exercise to Strengthen Emergency Preparedness in East Africa, 2019.

Field simulation exercises (FSXs) require substantial time, resources, and organizational experience to plan and implement and are less commonly undertaken than drills or tabletop exercises. Despite this, FSXs provide an opportunity to test the full scope of operational capacities, including coordination across sectors. From June 11 to 14, 2019, the East African Community Secretariat conducted a cross-border FSX at the Namanga One Stop Border Post between the Republic of Kenya and the United Republic of Tanzania. The World Health Organization Department of Health Security Preparedness was the technical lead responsible for developing and coordinating the exercise. The purpose of the FSX was to assess and further enhance multisectoral outbreak preparedness and response in the East Africa Region, using a One Health approach. Participants included staff from the transport, police and customs, public health, animal health, and food inspection sectors. This was the first FSX of this scale, magnitude, and complexity to be conducted in East Africa for the purpose of strengthening emergency preparedness capacities. The FSX provided an opportunity for individual learning and national capacity strengthening in emergency management and response coordination. In this article, we describe lessons learned and propose recommendations relevant to FSX design, management, and organization to inform future field exercises.