Tuesday, January 26, 2016

Put The ‘Network’ In Measures Of Network Adequacy

Blog_physicians consulting

According to the Urban Institute more than 11 million people were enrolled in Affordable Care Act (ACA) marketplace health plans in 2015. While enrollment numbers tell us how many people are covered, they do not necessarily tell us whether all 11 million people have access to care.

Reports in The New York Times have identified situations where newly enrolled individuals have had difficulty accessing needed specialist care. Recent academic studies have underscored these problems, finding that many plans may not include any of a particular type of specialist or may include a relatively small portion of specialists in a given area.

Network adequacy is generally defined as sufficient number and types of providers to ensure reasonable access without delay. Defining "sufficient" and "reasonable," however, is challenging. The Medicare program and many state regulators measure network adequacy with respect to time and distance traveled by patients, but standards vary. State definitions of a reasonable distance varies considerably from a 30-minute drive or 20 miles for primary care (Delaware) to 75 miles for specialty care (Texas) reflecting local context.

In light of the growing evidence on narrow networks, out-of-date network directories, and challenging access to covered providers, the Centers for Medicare and Medicaid Services (CMS) recently called for the creation of better measures of network adequacy in its November Notice of Proposed Rulemaking (NPRM) for the 2017 Benefit and Payment Parameters for marketplace health plans. Network adequacy measures could not only be used by regulators, but also provide important information to consumers interested in signing up for or switching plans. CMS floated one potential measure of network adequacy: a proportion of providers covered in a geographic area. Measuring network adequacy, however, is inherently complex.

The tools of social network analysis may provide a useful foundation for moving forward. A. James O'Malley defined social network analysis as a field that "studies the structures of relationships linking individuals (or other social units, such as organizations) and interdependencies in behavior or attitudes related to configurations of social relations." Social network analysis reminds us that network adequacy is not only about the individual health care providers who are included in a health plan but also about the relationships between these providers.

Social network analysis has been applied to assess relationships between pairs of providers, to determine the position of a particular provider within the broader network context, and to study the network as a whole. Furthermore, social network analysis reminds us that provider relationships are dynamic — changing over time. We describe two insights from the existing work on social network analysis that may be incorporated into and used alongside of traditional measures of network adequacy.

Network Connections

Consumers may want the broadest possible access to providers. However, they also want to know that they are receiving well-coordinated care. Research at the intersection of health services research and social network analysis indicates that providers who share many patients with one another are more likely to have referral relationships and share clinical information. Seeing doctors who share clinical information may lead to patients receiving higher quality and lower cost care.

Examining not only the specific providers that are included in a health plan network but also how frequently providers in a plan network share patients with one another may be an important consideration for patients concerned about coordinated care. Importantly, such a measure does not necessarily mean a given pair of providers would coordinate care or coordinate care well, but this information would provide a sense for how frequently providers work together. Information about the number of patients providers share with one another may be measured using historical health plan claims data. These data could be supplemented with Medicare fee-for-service claims and other private plan data to provide a more complete picture of how often providers share patients.

Using this information, CMS, state regulators, and marketplace administrators could create several measures of network connections such as the proportion of provider pairs in the health plan network that share above a certain threshold of patients or the average number of patients shared between providers (where more shared patients indicates a stronger connection). Given high copays associated with seeing out-of-network providers, measuring the frequency a health plan's providers share patients with out-of-network providers may be informative for patients worried about being referred out of network for their care.

To facilitate interpretation, regulators could consider presenting health plan network connection measures relative to what is observed in the most popular state government health plan or the Medicare fee-for-service program in that geographic area. It would also be reasonable for network connection measures to focus on connections between primary care providers and specialists (rather than on connections between specialists).

Stability Over Time

Are the providers listed today going to be the same providers that I need next month? McKinsey found exchange plans included more hospitals between 2014 and 2015 in their networks. To our knowledge, changes over time in health plans' outpatient provider networks have not yet been examined — but they should be. By using multiple years of a health plan's provider network data, CMS or other regulators can assess the average turnover rate per plan — both overall and by specialty. A stable plan network should ensure consistency both for patients and for providers' ability to make in-network referrals.

Such stability can be quantified as the percent change (including providers who left as well as those who entered the health plan's network). Stability could also be specialty specific, which may be particularly important for individuals with chronic conditions.

One could imagine more complex measures which incorporate both network connections and stability to assess how often pairs of providers work together over time. With one in four US adults managing two or more chronic conditions, many will likely need to see multiple providers regularly. For the successful, long-term management of their health, these patients would benefit greatly from measures of whether one's health care team will continue to work together within the same plan.

A Critical Opportunity

Advances in data collection, statistical methods for relational data structures, and access to powerful computing servers could revolutionize how we can measure health plan network adequacy. State all-payer databases, for example, provide new, powerful, and nearly comprehensive data sources. Electronic medical record data are also a promising data source for assessing the patient sharing relationships between providers.

Moreover, network adequacy is inherently multidimensional. For consumers, it's not enough to just assess the number of providers or distance to a provider. After all, a health plan directory could be chock full of dermatologists when you actually need an oncologist. It's also not enough for a plan to cover the most doctors if those doctors rarely work together.

Not only would network adequacy measures be useful to consumers in selecting an appropriate plan, these measures could also be used by state and federal regulators to oversee health plan adequacy. Network adequacy measures can be used to set a floor with respect to what kind of limited network is too limited. These measures can also offer opportunities for health plans to differentiate themselves in a crowded field, balancing demands for broad access with the need for coordinated care and stable networks over time.

In this NPRM, we see a potentially powerful opportunity to influence how health plans design their provider networks. Health plans will focus substantial resources on what is measured and publicly reported. If CMS focuses only on plan breadth then that is what health plans will focus on, at the expense of other important factors. If, on the other hand, we deploy measures aimed at capturing and synthesizing networks' complex and dynamic nature, health plans will have an incentive to offer high-value networks and consumers will have the tools to make more informed choices.



from Health Affairs Blog http://ift.tt/1PigX39

No comments:

Post a Comment