Employers and health care purchasers have turned to benefit design to manage their health care costs and to incentivize employees to act as consumers and seek higher-value care — care from health care providers that offers a good combination of cost and quality, or services from which they will particularly benefit. Together, thoughtful approaches to provider payment and benefit designs could drive higher value into health care through changing consumer behavior and shifting greater accountability for quality and cost onto providers.
Catalyst for Payment Reform (CPR) recently reviewed the state of evidence on a range of payment methods, drawing from 75 academic and self-published studies. Although our review identified mixed results on the effectiveness of payment reforms, it also highlighted confounding and contextual factors impacting the execution of payment reforms, such as health insurance benefit designs. Knowing that benefit designs may impact the willingness of a health care provider to accept a new form of payment as well as that provider’s success with it, we decided also to review the evidence for various benefit designs.
CPR’s Literature Review
To assess the effectiveness of various benefit designs, CPR searched the literature and identified about 30 articles — a combination of independent, academic studies, self-published studies, and employer case studies — on the performance of benefit design programs across the United States. Using PubMed and Google, we identified search terms for the various benefit designs (e.g., reference pricing, narrow network, etc.) to find program evaluations. We also referenced case studies of programs implemented by employer-purchasers. Our review focused on the most prevalent innovative benefit designs today, including high-deductible health plans (HDHP), value-based insurance design (V-BID), reference pricing, centers of excellence (COEs), narrow networks, and tiered networks.
High-Deductible Health Plans
CPR reviewed eight studies of large employers using high-deductible health plans in various regions in the United States. Across these employers, there were consistent reductions in spending and service utilization, including valuable primary and preventive care. For example, one employer’s introduction of HDHPs led to a reduction in spending of between 11% and 15% during 2013 — savings equivalent to approximately $82 million to $112 million — compared to what was anticipated in 2012. Furthermore, spending was reduced by 12% in 2014 compared to 2012.
This causal reduction occurred in all categories of health care spending including inpatient, outpatient, emergency room, pharmaceutical, and preventive health spending. The employer’s program saw no evidence of an increase in consumers shopping for health care services based on price, even when they had incentives to do so. However, consumers did reduce their overall use of health care services by 17.9% from 2012 to 2013. More specifically, consumers reduced their utilization of valuable preventive care by approximately 10% for 2013 and 2014, compared to 2012.
CPR conducted its own case study of the FedEx Corporation. FedEx implemented two “consumer-directed” benefit designs in 2014 — an HDHP paired with a health savings account — which saw similar savings. Aware of the tendency for HDHPs to lower care utilization, FedEx structured the plans so that primary care was not subject to the deductible. Despite the company’s efforts to educate members about this benefit, FedEx saw a 7% decrease in outpatient utilization, including for primary care — an unintended consequence. FedEx’s program also focused on educating employees about alternatives to the emergency room. As a result of its emphasis on emergency visits, those decreased by 8%, generating an estimated $8.3 million in savings. Urgent care visits also increased 28.7%. While HDHPs have demonstrated significant savings, if not paired with other thoughtful benefit designs, they sometimes come at the cost of patients not utilizing the care they need.
Value-Based Insurance Design
Value-based insurance design is built on the principle of lowering or removing financial barriers for patients to essential, high-value clinical services, including medications, that would benefit them. Across the six V-BID studies we reviewed, the results were mixed.
The Connecticut Health Enhancement Program, an effort of the State of Connecticut Employee and Retirees Organization, saw significant changes in the percentages of eligible population members receiving high-value medical services across a range of outcomes in the state, relative to comparison states. Preventive office visits and screenings increased by 13.5 percentage points in year 1 and 4.8 percentage points in year 2, relative to baseline. The program also produced a reduction in emergency room visits without a resulting inpatient stay. Across all chronic conditions there were significant increases in physician office visits, including a 1.6 percentage point increase in year 1 and 1.2 percentage point increase in year 2. Spending increased by $730 per enrollee from baseline in year 1 and $961 per enrollee from baseline in year 2, relative to the comparison group.
Across the other five programs reviewing medication adherence, V-BID programs consistently demonstrated higher medication adherence, though their impact on cost savings was also mixed.
To date, the most studied reference pricing programs are California Public Employees’ Retirement System (CalPERS) Hip and Joint Replacement Program and Safeway’s reference pricing for imaging and lab tests. Under this model, the employer or other health care purchaser, or insurer, sets a price for the service, and if the patient chooses a higher-priced provider, the patient pays the difference. Both programs led to savings for the companies due to members selecting more affordable providers. In the first year after implementation, CalPERS’ program saw a 21.2% increase in the number of its members who chose the low-price hospitals it had designated as value-based provider facilities (280 in 2011 compared to 231 in 2010). In addition, there was a 28.5% increase in volume for those facilities.
The program was associated with a 13.2% reduction in average hospital prices in 2011, or $4,597 per case. The net savings attributable to CalPERS’ reference pricing program totaled $2.1 million in 2011. The program was credited with cumulative savings of $5.5 million over the first 2 years; the average hip/knee replacement price charged declined 26%, or more than $9,000 per procedure.
Similarly, Safeway’s reference pricing program reduced the average price paid per lab test by 33% over a 3-year period from 2010–2013. The study concluded that Safeway’s reference pricing program reduced spending on 285 different lab tests by $2.56 million over 3 years compared to what would it would have spent without the program. Researchers estimate Safeway could have saved $4.08 million if it applied the program to all lab tests. While in both cases the reference pricing program led patients to select more affordable providers and prices for those services to decrease as a result, it is unclear whether providers increase prices of other services to make up for the loss in revenues.
Centers of Excellence
Across the six studies of centers of excellence programs that CPR found, the results were generally positive. COEs have successfully reduced unnecessary utilization of medical services as well as complication and readmission rates. For example, the Lowe’s Companies, Inc. COE program for major cardiac procedures saved the average Lowe’s associate utilizing a center of excellence approximately $3,300 in copayments and other fees compared to associates who received the same care not from the COE.
The Centers for Medicare and Medicaid Services’ (CMS) utilization of centers of excellence for bariatric surgery similarly generated statistically significant reductions in the 90-day complication and readmission rates, and in all adverse outcomes. The percentage of procedures performed at the CMS COE increased from 60.55% in 2008 to 73.1% in 2011. The proportion of Medicare patients receiving surgery at the COE similarly increased from 77.7% in 2008 to 88.1% in 2011.
CPR also conducted a case study of Walmart’s centers of excellence program for spine surgery. Among patients referred by physicians for spine surgery who then received an evaluation by one of Walmart’s COEs, 50% were determined to be inappropriate for surgery and received a different course of treatment. Walmart saw a high level of adherence to the program, with fewer than 2% of associates receiving the surgery despite the recommendation from a COE that surgery was unnecessary. The program generated savings through reducing medically unnecessary surgeries. Based on the program’s initial success, Walmart plans to expand the spine COE program geographically and to develop a COE program for hip and knee replacement.
CPR reviewed four studies on narrow network design, two of which focused on the narrow network design put in place by the Group Insurance Commission, Commonwealth of Massachusetts (GIC). In 2010, the GIC introduced a narrow network plan to its population. To incentivize enrollment, the GIC offered an initial 3-month period in which those who opted into the narrow network did not need to pay their share of the premium. Analysis from the National Bureau of Economic Research demonstrates GIC members who selected the narrow network plan spent 36% less overall. They increased primary care visits, and thus spending on primary care increased by 28%, while reducing visits to and spending on specialists by 45%. The program reduced the GIC’s total spending by 4.2%.
Among four studies we reviewed on tiered network benefit designs, three analyzed how these designs impact providers’ market share of patients. They found a common response: Providers in lower-performing tiers consistently earned lower market shares of new patients. For example, a review of the GIC’s tiered network design saw providers in the lowest tier experience a 12% loss of market share compared to middle-tiered providers. Patients saved an average of $60 in copayments if they visited top-tier providers and an average of $30 for seeking care from middle-tiered physicians, assuming an average of three annual visits.
While payment reforms and benefit designs have received attention independently, evidence focusing on the success of complementary payment and benefit designs is lacking. The studies we reviewed are limited in scope and in the information they provide on the details of the benefit design programs, the context of their approaches, and the impact on care delivery and costs. In addition, not much is available about the evidence behind benefit designs because while both Medicare and Medicaid tend to release more information about the impact of their health reforms than other payers, they more rarely employ innovative benefit designs.
Evaluation of benefit designs comes most often from other public purchasers, such as state employee and retiree agencies, or private employers. Existing evidence on most of these innovative benefit designs is quite promising, suggesting continued experimentation with and implementation of these approaches may help to contain health care costs for purchasers and consumers, in addition to helping to connect consumers with higher-value services and providers. With more transparent and thorough evidence on the effectiveness of various benefit designs, such as greater independent evaluation of the programs of commercial health plans, employers and health care purchasers will have greater motivation to embrace models that help connect the members of their populations with the highest-value health care.