How Do Actuaries Model Healthcare Costs?

This video provides an in-depth exploration of how actuaries model healthcare costs, specifically focusing on projecting an employer-sponsored health plan's spend. Dr. Eric Bricker, the speaker, demystifies the actuarial process, which often appears opaque to non-actuaries, by breaking it down into eight distinct steps. His approach is to explain the underlying logic and data requirements, emphasizing the importance of historical data analysis and forward-looking adjustments to arrive at a projected Per Member Per Month (PMPM) cost. The video aims to educate a broad audience, including HR, CFOs, and professionals in the pharma and med device sectors, on the intricate methodology behind these critical financial projections.

The process begins with the foundational step of gathering comprehensive historical data, typically 24 to 36 months of both paid and pending claims, alongside detailed monthly plan member enrollment data. This granular data is crucial for accurately calculating the historical PMPM, which involves dividing total claims expense by total member months, accounting for members joining or leaving the plan throughout the year. Dr. Bricker highlights the breakdown of claims into categories like inpatient, outpatient, physician, and prescriptions, as these categories are subject to different medical cost trends. He then explains how to apply these trends, which incorporate factors such as medical inflation (e.g., 5-6% for inpatient/outpatient/physician, 7.5-8% for RX), changes in utilization (like those seen during COVID-19), the impact of new treatments and drugs (specifically mentioning GLP-1s), and regulatory changes (such as those introduced by the Affordable Care Act).

Further adjustments are made for plan design changes, where actuaries use proprietary tables and historical data to quantify the impact of increased deductibles, higher cost-sharing, narrow networks, or benefit enhancements like IVF coverage. A critical step involves excluding high-cost claimants who are non-recurring, such as those who have passed away, left the plan, or undergone successful, non-chronic treatments. Demographic changes within the covered population, such as an influx of younger or older employees due to early retirement programs or high turnover, also necessitate adjustments to the risk profile. Finally, a risk margin (typically 2-5%) is added to account for projection uncertainties, and the projected PMPM is benchmarked against similar groups to provide a range of estimates: high, mid, and low. Dr. Bricker underscores that the value of actuaries often lies in their access to extensive proprietary historical data and actuarial tables, which allow them to quantify these complex adjustments.

Key Takeaways:

• Foundational Data Requirements: Accurate healthcare cost modeling begins with 24-36 months of historical claims data (both paid and pending) and detailed monthly member enrollment data to understand plan utilization and membership fluctuations.
• Per Member Per Month (PMPM) Calculation: The core metric for comparison and projection is the PMPM, calculated by dividing total claims expense by total member months, which normalizes costs across varying group sizes and enrollment periods.
• Categorization of Claims: Breaking down claims into categories like inpatient, outpatient, physician, and prescriptions is essential because each category can have different medical cost inflation rates and utilization patterns.
• Applying Medical Cost Trend: Projections must incorporate medical inflation (e.g., 5-6% for most services, 7.5-8% for RX), changes in utilization (e.g., post-COVID rebound), the introduction of new treatments/drugs (e.g., GLP-1s), and regulatory mandates (e.g., ACA benefit changes).
• Impact of New Treatments and Drugs: Pharmaceutical companies and medical device manufacturers should note that new drugs like GLP-1s are explicitly factored into actuarial trend calculations, influencing overall healthcare spend projections.
• Adjusting for Plan Design Changes: Changes in plan design, such as increasing deductibles or co-insurance, implementing narrow networks, or adding new benefits (e.g., IVF), have quantifiable impacts on claims costs, which actuaries model using historical data and proprietary tables.
• Excluding Non-Recurring High-Cost Claimants: It is crucial to identify and exclude high-cost claimants who are unlikely to recur (e.g., deceased members, those who left the plan, or those with successful, one-time treatments) to avoid overstating future costs.
• Demographic Adjustments: Significant demographic shifts within an employer's population (e.g., early retirement programs leading to an older workforce or high turnover resulting in a younger workforce) necessitate adjustments to the overall risk profile and projected costs.
• Risk Margin for Uncertainty: Actuarial projections are educated guesses, requiring the addition of a 2-5% risk margin to account for potential errors or unforeseen circumstances, providing a buffer in the final cost estimate.
• Benchmarking for Validation: Final PMPM estimates are benchmarked against similar historical groups to validate the projections and typically presented as a range (high, mid, low) to reflect inherent uncertainties.
• Value of Actuarial Tables and Data: A significant value proposition of actuaries lies in their access to extensive proprietary historical data sets and actuarial tables, which enable them to quantify complex adjustments that would otherwise be impossible for individual employers.

Key Concepts:

• Per Member Per Month (PMPM): A standardized metric used in healthcare finance to express the average cost of healthcare services for each plan member per month, allowing for comparison across different groups and time periods.
• Medical Cost Trend: The projected rate of increase in healthcare costs over time, incorporating factors like inflation, changes in utilization, new technologies, and regulatory impacts.
• Actuarial Tables: Proprietary databases and models developed by actuaries that contain historical data and statistical relationships used to quantify the financial impact of various factors, such as plan design changes or demographic shifts.
• Specific Stop-Loss Level: A threshold (e.g., $100,000 per claimant per year) above which an employer's self-funded health plan is reimbursed by a stop-loss insurance carrier, protecting against catastrophic claims.

Examples/Case Studies:

• COVID-19 Impact on Utilization: The pandemic dramatically reduced healthcare utilization initially, followed by an increase in subsequent years as deferred care was sought, illustrating how utilization changes significantly affect cost trends.
• GLP-1 Medications: Mentioned as a specific example of new drugs that actuaries must factor into RX spend trends due to their significant cost and growing adoption.
• Affordable Care Act (ACA): Cited as an example of regulatory changes that mandated coverage for certain services, thereby increasing healthcare costs for plans that previously did not cover them.
• Deductible Increase Example: An increase in deductible from $500 to $1,000 is used to illustrate how actuaries can quantify a 4% decrease in employer claims costs based on their proprietary tables.
• Benefit Enhancement (IVF): Offering invitro fertilization coverage is given as an example of a benefit enhancement that would increase healthcare costs, with actuaries able to project the specific financial impact.
• Non-Recurring Claimants: Examples include a deceased plan member, a dialysis patient who left the plan (saving $60k-$120k annually), or a successful cardiac surgery patient (e.g., stent or CABG) who may incur few subsequent costs.
• Demographic Shift (Early Retirement): An employer offering early retirement could see a shift in their workforce's age profile, impacting the overall risk and cost of the health plan.