Health actuaries have been estimating incurred but not paid (IBNP) claim liabilities for decades. As claim payments moved from manual to electronic methods, payments processing has become quicker. This has caused IBNP to decrease as a percentage of total incurred claims. However, it is still critical to estimate total incurred claims before all claims are paid in order to evaluate profitability and set future premiums.
Typically, the IBNP estimate is based on completion factors developed from historical payment patterns. This methodology, however, can be volatile for the most recent months, with minimal runout. For recent months, a per unit trend analysis is typically used (such as monthly cost per member, per employee, per hospital day, etc.).
In order to improve accuracy of IBNP claim estimates, additional information can be used to inform the trend analysis for the recent months. This includes:
• Working days. Working days in a month varies based on when weekends and holidays fall, impacting the availability of care as well as the claim processing capability of carriers. The impact on incurred services varies by setting. Hospitals empty out over holiday periods and specialists may work fewer hours on holidays and weekends. Prescription drugs have a pattern as well that varies by weekday, weekend days, and holidays. However, Medicaid nursing homes may be paid on a monthly basis regardless of the number of days or holidays in the month.
• Claim payment pattern indicators. Traditionally, these include claim inventory, preauthorization, real-time reports from hospitals on inpatient days and/or admissions, and reported high-dollar claims.
• Tracking the flu season. Wall Street investors follow the flu season to see the impact on health insurers. Higher-than-typical numbers of flu cases will lead to increased claims and vice versa. Some organizations track the flu at more granular levels, such as regional flu counts for high-risk members.
• Weather. Snow days, flooding, and other weather-related events should be measured. These events can have an effect on incurred claims similar to additional holidays, and also can produce a pent-up demand impact later.
• One-time events. Items that may impact claim processing such as system conversion or the transition to ICD-10 have the potential to impact claim payment patterns. System conversions often create significant claim backlogs and the potential for overpayments until corrections work through the system edits.
• Seasonality. The seasonal curve for typical commercial business keeps getting steeper, with larger deductibles and maximum out-of-pocket expenses. This impacts the timing of incurred claims and the resulting IBNP significantly throughout the year, but typically it’s most exaggerated at year-end and the first quarter. Some specialty blocks or employer groups exhibit clear seasonal patterns. Dental plans, Medicare Supplement, school groups, and other cohorts present very different claim costs by month.
• Benefit changes and risk scores. January 1 can cause major changes for Medicare, business related to the Patient Protection and Affordable Care Act (ACA), and some large plans. Benefit changes may have significant impact on cost-sharing and seasonality as noted above. A review of the change in risk score or monthly premium can help determine if the risk profile of membership may also be materially different from the prior month. These items could influence incurred claims and resulting IBNP estimates in the first quarter.
• Comparison of prior completion factors. More sophisticated organizations will look not only at averages but at percentiles. Completion factors falling into higher/lower percentiles relative to prior months should be reviewed with more scrutiny.
• Staffing changes. Looking at the number of claim payers working per claim submitted may be a useful metric for smaller plans that adjudicate their own claims.
• Large claimants. It is often useful to estimate hospital costs on a contract-by-contract basis, taking into consideration stop-loss provisions and other hospital-specific items. Some plans perform analytics to determine if there is backlog in some key facilities, increased claim denials that are due to new policies put into place, and any changes in the claim adjudication process that can throw off prior relationships between inventory and actual claim payment.
Reserving can sometimes feel more like art than science. Many of these factors can improve your estimates, but what is good for some blocks of business does not always improve the estimates for others. A successful valuation team will also incorporate frequent monitoring, communication between pricing and forecasting personnel, input from claim payment personnel, and evaluations of the reasons for deviations from the expected.