Advocates for Evidence-Based Health Policy Speakers
Two Options to Slow Healthcare Costs
The holy grail of value-based care is to reduce costs while improving quality. Zhang, et. al. compared the success of two different insurance designs in achieving that goal for total joint replacement surgeries.
One insurance plan implemented reference-based pricing (RBP), while another insurance plan steered employees toward insurer-selected Centers of Excellence (COE). RBP capped total payments for the episode and some providers agreed to the price limits. Both RBP and COE reduced the average per capita cost of total joint replacements by approximately 30%. Both reduced total annual costs – RBP by $7.4M and COE by $14.2M.
RBP and COE both operate similarly to narrow networks by increasing the out-of-pocket cost for patients choosing non-preferred providers. RBP patients who chose outside providers were responsible for any costs above the cap. Similarly, patients choosing a non-COE provider did not receive insurance benefits for their joint replacement surgery. Narrow networks have become increasingly prevalent in ACA Marketplace and Medicare Advantage plans as a means to control healthcare costs. Patients in this study used non-preferred providers much less often for joint replacements, which contributed to the total savings reported.
However, the full explanation for savings differed between RBP and COE. RBP patients switched to RBP-preferred providers with lower prices. Interestingly, capping total price also pushed down prices for non-preferred competitors, so costs decreased across the board. However, price changes were only a one-time shift. Meanwhile, low-cost preferred providers actually increased their prices toward the cap, so savings may not be sustained.
COE savings, on the other hand, came from reducing the number of surgeries performed. COE patients avoided using non-preferred providers, but did not switch to new ones, so the preferred providers did not actually perform more surgeries. It is unclear whether this was because COE providers lacked capacity to absorb new patients or because COEs had more stringent standards for offering surgery thus reducing unnecessary care.
Both insurance designs operated like a narrow network plan, but RBP savings came from shifting surgeries to lower-cost providers, while COE savings came from reducing the number of surgeries performed. To drive down both price and utilization might require combined strategy, where COEs are designated based on their outcomes but also agree to price caps.
Abstract
Various health insurance benefit designs based on value-based purchasing have been promoted to steer patients to high-value providers, but little is known about the designs’ relative effectiveness and underlying mechanisms. We compared the impact of two designs implemented by the California Public Employees’ Retirement System on inpatient hospital total hip or knee replacement: a reference-based pricing design for preferred provider organizations (PPOs) and a centers-of-excellence design for health maintenance organizations (HMOs). Payment and utilization data for the procedures in the period 2008-13 were evaluated using pre-post and quasi-experimental designs at the system and health plan levels, adjusting for demographic characteristics, case-mix, and other confounders. We found that both designs prompted higher use of designated low-price high-quality facilities and reduced average replacement expenses per member at the plan and system levels. However, the designs used different routes: The reference-based pricing design reduced average replacement payments per case in PPOs by 26.7 percent in the first year, compared to HMOs, but did not lower PPO members’ utilization rates. In contrast, the centers-of-excellence design lowered HMO members’ utilization rates by 29.2 percent in the first year, compared to PPOs, but did not reduce HMO average replacement payments per case. The reference-based pricing design appears more suitable for reducing price variation, and the centers-of-excellence design for addressing variation in use.
PMID: 29200355
You must be logged in to post a comment.