Description: Cover a variety of topics centered around the unsupervised learning technique called 'clustering'. The presenters will take you through a wide array of topics that range from 'What is Clustering?' to 'How can I use this technique to solve problems I didn't even know existed?' In the health care industry, everyone is looking for an edge in reducing total spend-cluster analysis finds the unambiguous in an ambiguous framework of results.

Description: Actuaries historically modeled new launches in pharmaceuticals by anticipating the expected change to the market basket. This became more complicated with the recent launches of gene and cell therapies as they address rare diseases. Rare conditions have low total member volume to study and some do not currently have a diagnosis code to identify them within claims data. Addressing this modeling challenge has been growing more important in the recent years with the number of therapies launching. What has the health care industry, and more specifically health care, done to enhance modeling and how should this be altered for the future? Additionally, historic modeling has sometimes incorporated a level of direct medical treatment cost offsets where studies have shown lower treatment costs are achievable. Can gene and cell financial modeling include cost offsets as well? And if so, what are suggested precautions or limitations to this effort?

Description: How do we take a Minimum Viable Products (MVP) approach to launch new insurance products? Join this moderated discussion and focus on three key development aspects: Distribution, Systems, and the design and pricing of the Insurance product itself. Come discuss reasons to choose an MVP approach and how it solves many of the product launch challenges.

Description: The presenters give a history lesson on the origin of CMS Medicare star ratings, including why this demonstration was originally created and how the program has worked over the years. The discussion will cover both the evolution of the underlying measures and the financial impact to Medicare Advantage stakeholders. By attending the session, you will have a stronger understanding of how member experience and outcome-based measures can be factored into the funding within a health care system to better align incentives among various stakeholders.

Description: Review a recently completed research project for a multi-stakeholder data consortium on social determinants of health (SDOH). In this project, claims, external SDOH data, geospatial analysis, and machine learning models were used to examine several questions: How to identify SDOH vulnerabilities? How sensitive are the findings with respect to provider coding pattern variations? How much variability in cost can be explained by SDOH vs. morbidity? What are the key risk drivers revealed by machine learning models? The session will review the methodology and main findings from this project.

Description: With the mandated health plan transparency files, health plans and providers are now able to substantially enhance their provider contracting efforts. Although the files are large and cumbersome they are market disruptors and provide new insights for provider contracting efforts. Explore a demonstration on the type of information now available and how it can be used to enhance this from both a health plan and health care provider perspective.

Description: This webcast will walk through several statistical methods for imputing race and ethnicity from U.S. datasets, highlighting strengths and weaknesses and evaluating accuracy.

Description: "A crucial importance in forecasting is to be able to distinguish cases when temporary slowdowns in health care utilization are followed by a “rebound” period of higher utilization. Come explore an analysis of a new mathematical model describing the return-of-care process. By means of simulation and case studies using the R programming language, the relationship between the visibility of these effects and the endogenous variance of the utilization time series is explored. After attending this session, you will be able to: Describe the parameters and mathematical form of the model Apply the model to real-world utilization data"

Description: You will be reminded how to spot and avoid statistical biases and fallacies that are common to health care actuarial data and problems. We plan to explain various statistical biases and fallacies and for each show one or two examples of how they might pop up in healthcare data and what strategies can be used to avoid/account for them. Likely topics include simpson's paradox, regression to the mean, base rate fallacy, sampling bias, survivor bias, Berkson's fallacy, and others.

Description: CMS has required hospitals to publish their chargemaster rates, cash discounts and negotiated rates as of January 1, 2021. Our panel is comprised of a data strategist, an executive from a large hospital system and an actuary.