The Impact of COVID-19 on Health Care Spending in Three States: A Health Care Cost Driver Analysis 

Focus Area:
Sustainable Health Care Costs
Peterson-Milbank Program for Sustainable Health Care Costs

Key Takeaway

  • Analysis of commercial insurer data in all-payer claims databases finds that increases in use of medical services and drug price increases in 2021 drove higher spending

In recent years, eight states have adopted statewide spending targets to assess and curb escalating health care costs. These programs typically gather aggregate spending data from participating payers, enabling states to track annual changes in spending relative to the target and identify contributions from individual spending categories. 

State all-payer claims databases (APCDs) are being used to complement these efforts by supporting secondary analyses of cost drivers at the market level and within specific service categories. While APCDs lack most commercial self-insured spending in a state (in contrast to cost growth target measurement), they yield analytic insights that are otherwise unavailable to states and the public. 

APCD-based cost driver analysis helps states understand why spending is growing. It primarily employs three measures: “service units,” “units per 1000 members,” and “payment per service unit.” This type of analysis typically excludes non-claims spending data and drug manufacturer rebates, which are often not found in APCDs and thus require supplemental data sources.  

Connecticut, Oregon, and Rhode Island are three states with cost growth targets and operational APCDs supporting cost driver analyses. In the spring of 2023, all three reported substantial increases in health care spending in 2021, causing them to exceed their cost growth targets for the commercial market.   

We examined state analyses of claims-level commercial market data from the APCDs within these three states to understand why their spending growth was higher than the target. We isolated the contributions from changes in utilization and payment per service unit and sought to understand their relative impact on service category trends from 2020 to 2021. We found a large increase in medical service utilization in 2021, resulting in a significant increase in spending across states. We can reasonably speculate that this reflects the reduced use of medical services in 2020 due to the COVID-19 pandemic. We also saw increases in retail pharmacy spending, driven primarily by increases in payment per service unit.  

Spending Across All Services Rose but Cost Drivers Varied by State 

The commercial markets in each of the three states observed large increases in claims spending from 2020 to 2021. While the spending growth rates varied by state, spending across all services, including medical services and retail pharmacy, rose over 10% at the per member per month (PMPM) level in each state. Connecticut experienced the largest increase in PMPM spending.  

When stratifying spending by medical services and retail pharmacy, similar trends were observed in all three states. Notably, PMPM-level spending on medical services exhibited larger increases than that of retail pharmacy (Figure 1). Spending on medical services increased by 22.3%, 17.6%, and 11.2% in Connecticut, Oregon, and Rhode Island, respectively. Although the growth rate of retail pharmacy spending was less dramatic, it exceeded 5% in all states and 10% in Connecticut. 

Though the trends in medical services and retail pharmacy spending shared the same directionality, different factors contributed. When evaluating the role of utilization versus payment per service unit in the change in medical services spending, utilization accounted for over 80% of the increase in all three states (Figure 2). In preceding years, payment per service unit was the driving factor in annual per capita spending growth. The greater role of utilization is not a surprising finding, however, in light of the significant decline in medical service utilization that occurred in 2020. 

In contrast, the increases in retail pharmacy spending were primarily driven by increases in payment per service unit and not by significant shifts in utilization. (While Oregon measured 30-day equivalent prescription utilization somewhat differently than did Connecticut and Rhode Island, we did not find this to have had a material impact on the finding.)  

Where the three states differed in terms of retail pharmacy spending growth was in the role of utilization. Utilization was a significant cost driver in Connecticut and Oregon, but not in Rhode Island. 

Professional and Outpatient Services Drove Trends 

An analysis of a broader list of service categories by state also revealed similar service category trends across these three states. Commercial PMPM spending in every service category saw an increase from 2020 to 2021. Although all states use somewhat different service categories and have slightly different definitions for shared categories, the largest contributors to the rise in claims spending from 2020 to 2021 were professional services and hospital outpatient services across all three states. 

In each state, professional services contributed to over 30% of the change in trend, while hospital outpatient services accounted for about 20% to over 30% of the year-over-year increase in spending. Taken together, these two service categories made up over 50% of the increase in overall PMPM spending in Oregon, and more than 60% of the increase in both Connecticut and Rhode Island. 

Contributions from all other categories were relatively modest but largely consistent across states. Retail pharmacy emerged as the next largest contributor to the increase in claims spending in Connecticut and Rhode Island, accounting for over 10% of the observed trend in overall spending. Hospital inpatient services were the third-largest contributor in Oregon and accounted for over 10% of overall increase in claims spending in both Connecticut and Oregon.  

An analysis assessing the contributions of utilization and payment per service unit across all service categories indicated that nearly every medical service category in each state was primarily driven by changes in utilization rather than changes in payment per service unit. This was especially evident in hospital outpatient and professional services, where utilization was the key driver of spending for these two service categories in all three states (Figure 3). Interestingly, while utilization drove hospital inpatient spending upwards in both Rhode Island and Connecticut, an increase in payment per service unit fueled the rise in hospital inpatient spending in Oregon. 

Collectively, these results indicate that an increase in utilization of medical services, particularly hospital outpatient and professional services, largely contributed to the increase in PMPM spending for all three cost growth target states.  

Looking Ahead 

States have invested significantly in new tools. allowing them to leverage their APCDs and explain where health care spending is growing, and why. Some states are even making the tools available for public use

This utility will be essential to analyze and understand the problem of high spending and spending growth, as well as identifying what spending needs to be tackled to slow spending growth. Increasingly sophisticated analyses will improve our collective understanding and remove reliance upon conjecture. While understanding of cost drivers is critical, they must lead to effective policy solutions if health care cost growth in the commercial market is to grow at an affordable pace.  Neither spending targets nor well-publicized analysis of cost drivers will be sufficient to achieve this ultimate aim. 

State staff from Connecticut, Oregon, and Rhode Island provided data for this analysis and reviewed this post prior to its publication.