cover story one

Tiered Cost-Sharing Health Insurance

Is this the Holy Grail?

BY BRYAN DOWD, PhD, TYLER BOESE, AND TIM MCDONALD

At the same time, numerous studies have shown there is wide variation in both providers’ (hospitals’ and physicians’) prices and practice styles (quantities of prescribed services) within the same market area, including low value and wasteful care, and there is little evidence to suggest that price and quality of care are related.

Given the conjunction of affordability problems, wide variation in the cost of care and lack of correspondence between cost and quality of care, it might seem obvious that the solution would be to tell patients where to find lower cost care and share the savings with patients who choose lower cost providers. Yet that is not the approach taken by the vast majority of employers. Instead, employees are placed in high deductible health plans, often with little information to help them find lower cost care. As a result, employees tend to cut back on both high and low value services. Employees are left to shop for each individual service, with no single entity responsible for care coordination or total cost of care.

Employees who choose lower cost providers share the savings.

Health care reform initiatives

At the federal level, many health care reform initiatives focus on provider payment, including capitation, bundled payments, shared savings, fee withholds, global budgets and accountable care organizations with downside risk. These initiatives generally are invisible to the consumers, and the results thus far have been lackluster at best.

Another class of health insurance benefit designs, including reference pricing and tiered cost-sharing, take a very different approach. These initiatives simply examine the variation in providers’ historic prices and practice styles and pass that information on to employees in the form of varying out-of-pocket costs. Those approaches give employees the information they need to identify lower cost providers, and employees who choose lower cost providers share the savings. The opportunity to gain patients simultaneously gives providers an incentive to lower their costs. 

None of these approaches to health care reform are a “holy grail” that will provide a painless pathway to affordable, high quality health care services. However, our analyses have convinced us that tiered cost-sharing represents an improvement over many current reform proposals and health insurance benefit designs and that it has features both patients and physicians might find appealing. Interestingly, Minnesota is home to one of the nation’s longest, most comprehensive and most fully developed examples of tiered cost-sharing.

State Employee Group Insurance Program (SEGIP)

The State of Minnesota is the largest employer in the state and has been using a tiered cost-sharing health insurance benefit design to cover its 130,000 employees and dependents since 2002. That program is referred to as the State Employee Group Insurance Program, or SEGIP. The SEGIP system creates a different set of incentives than the health insurance benefit designs medical practices typically face in the current market.

The SEGIP system has a strong orientation towards primary care. During the November open enrollment period each year, state employees choose from one of three health plans and designate primary care clinics for themselves and their family members for the coming calendar year. Employees with dependent coverage can choose a different clinic for each family member, but all family members must choose clinics covered by the same health plan. 

Employees are required to receive most of their care through their primary care clinic or a referred provider. The primary care clinic is held responsible for the employee’s total annual risk-adjusted cost of care, but because they help direct patients to specialists and hospitals, they also can influence total cost of care. 

Based on their total annual risk-adjusted per-capita cost of care over the previous two years, each clinic is placed into one of four cost-sharing tiers. Premiums are held constant across the tiers and health plans, but the tiers vary by deductibles, copayments and the patient’s maximum annual out-of-pocket cost. Patients choosing clinics with higher risk-adjusted total cost face higher cost-sharing. The differences in cost-sharing across the tiers are substantial. For example, the family coverage deductible ranges from $500 in Tier 1 (the lowest cost tier) to $3,000 in Tier 4 (the highest cost tier). Office visit copayments range from $30 in Tier 1 to $90 in Tier 4, and maximum annual out-of-pocket spending limits for family coverage range from $3,400 to $7,200, excluding prescription drugs. 

Over 80% of SEGIP members choose clinics in the lower two cost-sharing tiers. The threat of losing patients to lower cost clinics provides a strong incentive for primary care clinics to reduce the total cost of care. Once clinics are told their initial tier assignment, they are given the opportunity to discount their fees in order to remain in their current tier or move to a more desirable tier. Approximately 25% of the clinics currently operate under negotiated price discounts, generally in the 10%-20% range.  

Tiered cost-sharing represents an improvement over many current reform proposals.

Several features of the SEGIP system would be of interest to physicians. First, patients in the SEGIP system receive a clear signal regarding each primary care clinic’s total cost of care and a strong financial incentive to choose lower cost clinics. Thus, higher cost primary care clinics need to consider the possibility that patients will switch to a lower cost clinic. That’s not true in payment systems like bundled payments and shared savings that are invisible to the patient. 

Second, the SEGIP system introduces an element of uncertainty for individual clinics. A clinic’s tier is determined by its relative position on the distribution of clinics’ total cost. Therefore, a clinic’s tier assignment can change even if it makes no changes in its prices or practice style if the prices or practice styles of other clinics change. 

Third, the SEGIP system could affect referrals from primary care clinics to specialists and hospitals. Primary care clinics are responsible for total cost of care, of which specialists’ and hospitals’ prices and practice styles are an important component. As a result, primary care clinics may take a greater interest in their referral patterns. Independent primary care clinics may have more discretion over their referral patterns than primary care clinics owned by an integrated delivery system.

Fourth, if reducing the total cost of care becomes an important part of remaining competitive in the market for health care services, then primary care clinics may need help identifying low value care and ways to reduce avoidable use of health care services. Small independent primary care practices may have fewer resources to collect and process that information than clinics that are part of larger health care systems. 

A potential advantage of tiered cost-sharing for physicians is the wide latitude offered to health care providers in setting their prices and practice styles. The SEGIP system includes virtually every hospital and clinic in the State, and there are no in-network versus out-of-network providers determined by the health plans. Currently, health care providers in the SEGIP system are subject to the medical policy and utilization management programs employed by the three participating health plans. But in theory, health care providers in a tiered cost-sharing system could set their prices and practice styles at any level they like, while recognizing that decisions resulting in higher cost might result in their patients switching to lower cost clinics.

SEGIP is working to improve the information that State employees have when selecting their primary care clinic by distributing information on clinics’ tiers and quality of care directly to State employees through emails. That information is tailored to the clinics most frequently chosen by employees based on their zip code of residence. The emails highlight those clinics that are moving up or down a tier for the coming year, and the quality measures currently include diabetes, vascular care and pediatric asthma obtained from Minnesota Community Measurement. 

We have published two analyses of the SEGIP system. The first study (Dowd, Huang, and McDonald, “Tiered Cost-sharing for Primary Care Gatekeeper Clinics,” American Journal of Health Economics, 2021, 7(3); pages 306-332) confirms that State employees are aware of clinics’ tiers and consider the clinic’s tier when choosing a clinic for the coming year. The second study (McDonald, et al., “Primary Care Clinic Responses to a Tiered Insurance Network,” American Journal of Managed Care, 2021, 27(9), pages e316-e321) summarizes the results from a small sample of interviews with SEGIP’s primary care clinics. The findings from that study indicate that clinics have two reactions to tiering. First, they are concerned about losing patients to lower cost clinics, and second, they are concerned about developing a reputation as a high cost clinic. The clinics also express interest in greater transparency in the tier assignment process and how to reduce low value care and avoidable utilization.

To summarize, SEGIP’s tiered cost-sharing system offers an improvement over ineffectual provider payment reforms that are invisible to patients. And unlike large deductible health insurance benefit designs per se, tiered cost-sharing tells patients where to find lower cost providers and shares the savings with patients who choose those providers.

Conclusion  

The health care affordability problem needs to be addressed, and the current variation in providers’ prices and practice styles suggests it is possible to do so by using data already available to change the incentives faced by both patients and providers.

We currently have a research project underway studying what barriers exist for clinics attempting to reduce low-value care and improve their referral processes. We are seeking physician input and the interests of physicians willing to provide leadership and a professional voice to the approach of sharing savings with patients for choosing better quality, lower cost providers.

Bryan Dowd, PhD, is a professor in the Division of Health Policy and Administration (HPM) in the School of Public Health at the University of Minnesota. A copy of this article with a full set of references is available from Bryan Dowd at dowdx001@umn.edu.

Tyler Boese is a PhD student at HPM.

Tim McDonald is a PhD student at the Pardee RAND Graduate School. For more information, contact Tim McDonald at tmcdonal@rand.org.