Healthcare cost trend is a central assumption in health insurance pricing and forecasting. It is commonly defined as the expected annual increase in the cost of healthcare services and may be estimated using a combination of historical experience, market benchmarks and forward-looking adjustments. In practice, however, trend is often treated as a single number, which hides the different drivers of cost growth.
A common source of confusion is that the term trend is used to describe both what has already happened and what is expected to happen. Observed, or experience, trend reflects the change in costs between periods and includes the combined impact of all underlying factors. In practice, experience trend may include not only underlying cost growth, but also the effects of high-cost claimants, changes in population morbidity, benefit and network changes and other factors that could be adjusted or addressed separately in the pricing process.
In contrast, pricing trend refers to the forward-looking estimate of cost growth used in rate development. Pricing trend reflects underlying cost growth along with expectations about how those drivers will evolve over time, while excluding or separating one-time and temporary effects.
Treating trend as a single number hides important differences between the drivers of cost growth. In reality, costs change for different reasons. Some factors compound year after year, while others reflect changes in utilization, one-time shifts in cost levels or temporary effects that reverse.
Viewing trend through this lens leads to a different way of thinking about cost growth:
Healthcare cost trend is not a single rate of change, but the combined effect of multiple components that affect costs in different ways over time.
To make these distinctions explicit, this article introduces a framework for decomposing observed or experience trends into components that inform the development of pricing trend:
- Continuous drivers: persistent, compounding forces that generate the underlying rate of cost growth
- Adoption drivers: factors that increase costs as utilization expands across a population and whose contribution to trend diminishes once adoption stabilizes
- Level adjustments: discrete, nonrecurring changes that shift the overall cost level without affecting ongoing growth
- Temporary deviations: short-term departures from trend that subsequently reverse and do not affect long-term cost levels or growth
Each category exhibits a distinct pattern in how per member per month (PMPM) costs change over time. Modeling these components in isolation and in combination is intended to help actuaries interpret observed experience, improve pricing assumptions and communicate the sources of cost growth.
In practice, observed or experience trend can reflect a mix of underlying cost growth, discrete changes and temporary fluctuations that actuaries may consider separating and normalizing for pricing.
Continuous Drivers
Continuous drivers are the forces that drive ongoing healthcare cost growth. These factors operate year after year, increasing costs and compounding over time. In the absence of benefit changes, population shifts or other one-time effects, continuous drivers represent what actuaries typically think of as “trend.”
The most common examples are familiar from pricing work. Provider reimbursement increases, particularly fee-for-service escalators embedded in contracts, produce recurring annual cost growth. Baseline utilization also tends to increase gradually over time, and intensity of care evolves as clinical practice changes. In the prescription drug space, price increases within existing therapies contribute to ongoing growth. These effects persist and accumulate over time.
The defining characteristic of continuous drivers is that they affect the rate of cost growth rather than the level of costs. Their effect can be illustrated with a simple example. If costs increase each year due to ongoing factors such as provider reimbursement and utilization, PMPM evolves as shown in Table 1.
Table 1
Effect of Continuous Drivers on Rate of Cost Growth
PMPM (With Impact) is calculated as the product of PMPM (No Impact) and the applicable factor. Trend reflects the year‑over‑year change in PMPM (With Impact). The continuous driver factor represents the cumulative effect of cost growth, with each year’s increase compounding on that of the prior year.
In many pricing applications, these drivers form an important input to the pricing trend assumption. They represent the portion of cost growth that is expected to persist over time and are therefore the primary component used to project future costs.
Adoption Drivers
Adoption drivers reflect changes in cost that occur as a therapy, service or pattern of care is used by a growing portion of the population. Unlike continuous drivers, which produce steady and ongoing growth, adoption drivers increase costs only while utilization is expanding in this framework. Once adoption stabilizes and other factors are held constant, the adoption component’s contribution to trend goes to zero.
A simple way to understand this is through an example. Consider a new therapy that takes several years to reach full adoption within a population. At full adoption, the therapy increases overall costs by 4.5%. The impact on trend does not occur all at once; instead, it builds gradually as more members begin using the therapy each year.
A current example often discussed in this context is the use of GLP‑1 medications. Although overall GLP‑1 cost growth may reflect multiple factors, including changes in unit costs, rebates, coverage criteria or indication mix, this example isolates the effect of increasing utilization as adoption expands across a population. Under this framework, the contribution to trend is driven by the pace of adoption and diminishes as adoption stabilizes.
During the adoption period, costs increase as utilization expands; the impact on trend is driven by the rate of adoption; and once adoption stabilizes, the contribution to trend goes to zero. The key distinction is that the therapy permanently increases the level of cost but only affects the rate of growth while adoption is occurring. Table 2 illustrates how this relationship translates into observed cost and trend over time.
Table 2
Effects of Adoption Drivers on Cost and Trend Over Time
In this example, the adoption driver factor represents the cumulative increase toward a total cost impact of approximately 4.5% at full adoption. Trend reflects the year‑over‑year change in PMPM (With Impact). It is driven by changes in that factor from one year to the next, and once adoption stabilizes, the contribution to trend goes to zero.
In practice, adoption drivers are most frequently associated with new medications and medical technologies. Examples include the uptake of new specialty drugs, the expansion of behavioral health services, the increased use of telehealth and changes in clinical guidelines that expand treatment eligibility.
Level Adjustments
Level adjustments are one-time changes in cost that shift the overall cost level without affecting the ongoing rate of growth. Unlike continuous drivers, which compound every year, and adoption drivers, which build gradually over time, level adjustments take effect once and are then incorporated into the baseline.
A simple way to understand this is through a repricing example. Consider a contract renegotiation that increases provider reimbursement above prior expectations. If underlying cost growth is 3%, and the renegotiation adds an additional increase, the observed change in that year reflects both the underlying growth and the one-time adjustment. The portion above the expected trend represents a level adjustment.
The defining characteristic of a level adjustment is that it affects the level of costs, not the rate of growth. After the change is implemented, costs continue to grow at the underlying rate, but the adjustment itself does not recur.
This behavior can be illustrated with a simple example, shown in Table 3, in which a one-time change reduces costs by 1% in a given year.
Table 3
Effects of Level Adjustment on Cost
In this example, the level adjustment produces a one-time change in cost. It appears as a change in trend in the period in which it occurs but does not contribute to trend in subsequent periods. Growth resumes from the new level based on underlying drivers. Trend reflects the year‑over‑year change in PMPM (With Impact).
Level adjustments can increase or decrease costs and arise in a variety of common pricing situations. Provider contract changes are a primary example. When reimbursement differs from prior expectations, the difference represents a level adjustment. Similarly, network changes or competitive realignment may result in lower unit costs, producing a downward adjustment. On the pharmacy side, the introduction of generics often produces a discrete reduction in cost. Benefit changes, such as higher deductibles or changes in cost sharing, also shift the level of costs without affecting underlying trend.
In practice, observed changes in cost often reflect a combination of underlying trend and level adjustments. For example, an observed increase may reflect both ongoing cost growth and a one-time change in reimbursement. Separating these effects is critical in pricing, as only the underlying component is expected to persist.
Temporary Deviations
Temporary deviations are short-term changes in cost that move experience away from the underlying trend and then reverse. Unlike level adjustments, which create lasting changes in cost level, and adoption drivers, which build and then stabilize, temporary deviations do not persist. They affect observed results in a given period but do not change the long-term level or growth rate of costs.
A common example is a surge or decline in utilization driven by external factors. Seasonal flu activity, severe weather events or disruptions in access to care can cause costs to increase or decrease in a given year. These effects are typically followed by an offset in subsequent periods, as deferred or accelerated care returns to expected levels. High-cost claimants are another common example, as they occur every year but can cause costs to fluctuate from one year to the next. Because of this, they are typically treated as temporary deviations and adjusted for in pricing. If a consistent pattern of higher costs emerges, it may instead reflect a sustained change in morbidity, severity or treatment patterns and should be reflected in other components of the framework.
The COVID‑19 pandemic provides a clear illustration. For many services, utilization declined materially in 2020 and subsequently rebounded. These shifts created significant deviations from expected costs across multiple periods. While the magnitude and persistence of pandemic-related effects varied by service category and population, the experience illustrates how temporary deviations can materially affect observed trend even when the underlying cost trajectory remains unchanged.
Temporary deviations affect when costs are incurred, not how costs grow over time. Table 4 illustrates how a temporary deviation that increases costs in one year is offset in the following year.
Table 4
Effects of Temporary Deviations on Trend
In this example, the temporary deviation produces a short-term increase in cost followed by a reversal. Over time, the net impact is zero, and costs return to their original path. Trend reflects the year‑over‑year change in PMPM (With Impact).
In practice, temporary deviations are often visible in experience data and can significantly influence observed trend from one period to the next. However, they are typically adjusted for in pricing, as they do not represent ongoing cost pressures.
Putting It All Together
The preceding sections describe each component of cost change in isolation. In practice, however, healthcare cost trend reflects the combined impact of continuous drivers, adoption drivers, level adjustments and temporary deviations operating at the same time.
Observed cost and trend in any given year reflect a combination of these components. Some drive persistent growth, others temporarily increase or decrease costs, and some represent one-time structural changes. Understanding how these components interact is critical, as their combined effect can produce patterns that are not immediately intuitive when viewed as a single trend.
To illustrate this interaction, consider a combined example using the same patterns introduced in the earlier sections:
- Continuous drivers produce ongoing cost growth at approximately 3% per year.
- Adoption drivers reflect a gradual increase in utilization, reaching a total impact of 4.5% at full adoption.
- A one-time level adjustment reduces costs by 1% in a single year.
- A temporary deviation increases costs by 2% in one year and reverses in the following year.
To maintain consistency, each component follows the same pattern shown previously. For purposes of this illustration, the factors are applied multiplicatively in Table 5, reflecting how changes in cost could accumulate over time.
Table 5
Combined Effect of All Factors on Trend
This example highlights how different components influence observed trend across time, even though the underlying system changes in a predictable way. PMPM (With Impact) reflects the combined effect of all factors applied multiplicatively. Trend reflects the year‑over‑year change in PMPM (With Impact).
- In Year 2, observed trend is elevated at 6.6%, driven by the combined effect of continuous growth, early-stage adoption and a temporary deviation. Viewed in isolation, this could be interpreted as a structural increase in trend.
- In Year 3, the temporary deviation reverses, reducing observed trend to 2.5%, even though both continuous and adoption drivers continue to increase costs.
- In Year 4, the level adjustment reduces costs by 1%, partially offsetting the ongoing effects of continuous growth and adoption. As a result, observed trend is further suppressed to 2.0%.
- In Year 5, the effect of the level adjustment has been fully incorporated into the cost baseline, and adoption is nearing completion. Observed trend increases to 3.5%, reflecting continued underlying growth plus residual adoption effects.
- By Year 6, adoption has stabilized, and the temporary deviation has been fully reversed. Observed trend returns to approximately 3.0%, reflecting the underlying continuous drivers.
This illustrates how observed trend can vary from year to year even when underlying cost growth remains stable.
Implications for Pricing
In practice, pricing begins with observed experience, but observed trend reflects a mix of components that may not all persist into the future. Under this framework, an actuary can separate these components and identify the portion of cost growth that is expected to continue.
A useful way to think about this is as follows:
- Continuous drivers form the foundation of trend and are expected to persist.
- Adoption drivers contribute to trend only while utilization is expanding, and their contribution goes to zero once adoption stabilizes.
- Level adjustments represent one-time changes in cost and are handled separately.
- Temporary deviations affect observed results but are normalized, as they reverse over time.
As a result, pricing trend may not be equal to observed trend, but rather a refined view that potentially reflects underlying cost growth. In practice, the following steps could be taken:
- Removing temporary fluctuations from experience
- Separating discrete changes in cost level
- Evaluating how much adoption-related growth remains
- Identifying the underlying continuous trend
Failure to distinguish these components can lead to common errors, such as overstating long-term trend due to temporary deviations or embedding one-time changes into ongoing assumptions.
Pricing implies moving from observed experience to a structured view of cost growth, focusing on the components that are expected to persist.
Conclusion
Healthcare cost trend is often treated as a single number, but in practice it often reflects the combined effect of components that behave very differently over time. Using the framework described here with continuous drivers, adoption drivers, level adjustments and temporary deviations allows for the explicit contribution of each component to observed cost changes in distinct ways and potentially decreasing the risk of misinterpreting experience or inaccurately estimating components of future cost growth.
By separating these components, actuaries can move from observed experience to a clearer view of underlying cost growth. Continuous drivers form the foundation of pricing trend, adoption drivers contribute while utilization is expanding, and level adjustments and temporary deviations are addressed separately as they do not represent ongoing growth.
Viewing trend through this framework can provide a more structured approach to pricing, improving both the accuracy of assumptions and the ability to clearly explain the drivers of cost change.
This article is provided for informational and educational purposes only. Neither the Society of Actuaries nor the respective authors’ employers make any endorsement, representation or guarantee with regard to any content, and disclaim any liability in connection with the use or misuse of any information provided herein. This article should not be construed as professional or financial advice. Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries or the respective authors’ employers.
Joe Slater, FSA, MAAA, is an area vice president of managed care and payer relations. He can be reached at jpslater1511@gmail.com.