Variations in ACOs and Narrow Networks Challenge Efforts to Track Their Impact

Published on: 

Both accountable care organizations (ACOs) and narrow networks are important components of the value-based care movement. However, while there are features common to both models, Catalyst for Payment Reform has learned that what constitutes ACOs and narrow networks is inconsistent across health plans.

This article has been co-authored by Andréa Caballero, program director at Catalyst for Payment Reform (top headshot), and Roslyn Murray, senior project and research manager at Catalyst for Payment Reform (bottom headshot).

Accountable care organizations (ACOs) and narrow networks are important components of the value-based care movement—they both intend to encourage patients to seek care from more affordable, higher-quality providers. Health insurers contract with providers to offer both ACOs and narrow networks to consumers.

ACOs are groups of contracted providers that share financial and medical responsibility for providing coordinated care to patients, with financial incentives to provide high-quality care and limit avoidable, unnecessary spending. Narrow networks are groups of contracted providers that are smaller than a health plans’ broadest network offering and provide care—sometimes better care—at a more affordable price.

While there are features common to both models, Catalyst for Payment Reform (CPR) has learned that what constitutes ACOs and narrow networks is inconsistent across health plans.

This past year, CPR set out to determine how to measure the prevalence of ACOs and narrow networks as part of our evolving Scorecards on Payment Reform. The presence of ACOs could indicate that providers are changing the way they deliver care, working toward a more coordinated and continuous experience for the patient. The presence of narrow (or limited) networks may be a good indicator that payers are signaling to healthcare providers that they will reward more affordable care with more patients. By tracking these models over time, along with measures of cost, quality, and utilization, we can attempt to examine whether they are leading to a higher-value healthcare system.

However, through interviewing a dozen health plans diverse in size and geography, we learned that it is virtually impossible to develop measures that track the prevalence of these models or their impact on the healthcare system without standard definitions.


Variation in Health Plan Definitions

ACOs can have a variety of features. ACO providers could be held accountable for spending over the target budget or they may not bear any financial risk at all. ACOs can be structured around a group of providers or a hospital or health system. In addition, not all providers alter how they deliver care when they take on the ACO moniker. We’ve all heard the quip, “If you’ve seen 1 ACO, you’ve seen 1 ACO.”

Less is known about the variation in how health plans define and design narrow networks. CPR conducted interviews with health plans in hopes of converging on a single definition. We postulated that we would be able to conclude, for example, that a narrow network is one in which health plans exclude providers whose prices are one standard deviation above the mean or who don’t meet minimum quality thresholds. In addition, we wanted to learn how health plans determine who among certain provider types (primary care providers, specialists, and hospitals) is eligible to participate.

However, CPR found no consistent formula across health plans for selecting providers below a certain price point, above a specific level of quality, or by type. Health plans primarily consider which hospital or provider group will agree to a certain price (based on a premium analysis), whether excluding others is feasible given each provider’s market power or “must have” status, and whether excluding the hospital or physician group creates access issues. It is notable that among the health plans we spoke to, none used provider quality as the primary selection criterion.

Lowest Common Denominator Definitions

Accountable care organizations

Given that ACOs can vary significantly, instead of trying to measure the quantity of ACOs, CPR opted to quantify the number of shared-risk payment arrangements health plans have with providers.

This approach won’t capture every ACO in the marketplace, especially because most start with shared savings arrangements before migrating to shared risk. However, shared-risk payment arrangements mean providers accept some financial liability for exceeding specified spending targets. Therefore, they ostensibly have care processes in place that allow them to assume financial risk for their population of patients—eg, the ability to track quality of care and healthcare spending in near-real time. Or, these providers are pursuing ways to handle such risk—eg, establishing a team-based approach to managing and overseeing patient care.

In other words, providers in shared-risk arrangements are likely delivering care in a way that is different from providers that are not. Furthermore, while only a minority of ACOs are in shared risk arrangements today, the proportion is expected to grow significantly over time, making this a good measure of trend.

CPR originally thought it would be valuable to understand the percentage of a health plan’s total contracted providers in shared-risk contracts. However, health plans warned us that counting the number of providers in these contracts would be an arduous and imprecise task. For example, some health plans count the number of individual providers under a single contractual umbrella, whereas other health plans count the number of contracts, not the number of physicians covered by the contracts. Even with clear instructions, we could receive a variety of answers leading to an inaccurate count.

Instead, health plans advised us to measure the number of shared-risk contracts, regardless of whom they are with or how many providers are covered under the contract, to track prevalence and growth.

Narrow (or limited) networks

It may be too early to identify commonality in how health plans design narrow networks; perhaps a standard formula will never materialize. As CPR learned, viable approaches depend on market-specific nuances.

CPR developed a metric that boils these concepts down to the lowest common denominator so that we can capture consistently across health plans the presence of narrow networks—what we define as a product within a health plan’s portfolio of offerings that contains fewer providers than its broadest network—and the percent of members enrolled. This metric does not specify cost or quality thresholds, nor does it specify how health plans select the network by type of provider.


Our research underscored why standard definitions or metrics quantifying the prevalence of ACOs or narrow networks did not exist prior to our efforts to create them. Without standardization, tracking their prevalence and impact on our healthcare system will remain a challenge.