Are Limited Networks Necessary to Reduce Health Care Costs?

Among the dirty words most hated by health care consumers–such as “capitation” and “insufficient medical necessity”–a special anxiety infuses the term “out-of-network.” Everybody harbors the fear that the world-famous specialist who can provide a miracle cure for a rare disease he or she may unexpectedly suffer from will be unavailable due to insurance limitations. So it’s worth asking whether limited networks save money, and whether they improve or degrade health care.

As I understand it, four reasons are seen for limiting the doctors covered by an insurance plan.

Increased utilization of approved facilities

This is a pure business concern, not a quality concern. If you can channel patients back into a few hospitals and clinics, you can keep your beds filled and save the money wasted when bored staff sit around posting travel photos to Instagram.

Excluding expensive facilities

Another business concern. Studies have shown no correlation between the prices charged by health care providers–including world-famous facilities and prestigious academic settings–and the quality of outcomes. So just by artificially lopping off the institutions that charge a lot, insurers can save money while still serving their recipients well.

Better coordination of care

Now we move beyond the bean-counters and enter into real issues of quality. Supposedly, institutions that know each other through frequent referrals can work more closely together, making sure that post-discharge plans are followed and patients are kept on track for improvement. There is no guarantee that such coordination will happen, but it’s a goal of health reform and underlies the Accountable Care Organization (ACO) model.

More intensive use of primary care providers

Ideally, every professional in health care would develop a holistic understanding of the patient and think long-term. In practice, the PCP is most likely to do so. (With the intensified use they’ve seen over the past several years, and consequent shortening of time spent with each patient, this valuable perspective may be less common.) If a limited network can encourage the patient to rely more on his PCP, it may keep him healthier.

I was stimulated to write this article by a recent paper by the National Bureau of Economic Research (NBER) examining the costs of health care in my state, Massachusetts. The statistical models used in this article are hard for non-specialists like me to get their heads around, but the study looks well-grounded (as one would expect from the NBER, one of the country’s leading research institutions) and the conclusions are reassuring.

People who choose limited networks pay much less, mostly because the premiums for such networks rise much more slowly than for broad networks (p. 12). The Commonwealth of Massachusetts also benefited from cost savings. Although the data provide very little on which to judge the quality of outcomes, the few statistics available on such measures as inpatient spending and emergency room visits (p. 25) indicate that quality of care for limited networks is at least as good as for broad ones.

As one might expect, the narrow plans deliberately excluded expensive health care providers (pp. 5-6).

The one variable I’m not sure the authors could control for is the possibility that healthier people were more likely to choose limited plans (p. 27), and that costs might naturally be lower for such individuals. I will have to trust that the authors took this possibility into account.

The challenge I’d like to toss into the ring is this: couldn’t a rational health care system achieve all the benefits of limited plans while allowing patients to see anyone they want? Let’s consider again the four benefits I mentioned earlier:

Increased utilization of approved facilities

A rational health care system would pay for outcome instead of utilization, encouraging hospitals and clinics to put their spending where it was needed and hold back from expensive purchases that require excess use to pay off.

Excluding expensive facilities

A rational system would collect and publicize quality measures, and place some of the financial burden on patients to encourage them to do some price-shopping. Both of these innovations are starting to be seen. Idealists among us can even hope for a standard set of fees to replace the current chaotic negotiations between provider and payers.

Better coordination of care

If patient data was stored in a standardized format–preferably by the patient herself–all providers would have access, and a fee-for-value reimbursement model would encourage them to work together for better outcomes.

More intensive use of primary care providers

A holistic approach to health–which would reach outside the individual doctor’s office to the whole community in which the patient lives–would make the PCP the natural starting point for all health issues and give PCPs the tools to maintain patient health year-round.

So even though I acknowledge the value of ACOs and other forms of limited networks so long as our current health care system is limping along, the need for that kind of trade-off could end if we lift our eyes a little higher, look farther toward the future, and make strides toward better goals.

About the author

Andy Oram

Andy Oram

Andy Oram writes and edits documents about many aspects of computing, ranging in size from blog postings to full-length books. Topics cover a wide range of computer technologies: data science and machine learning, programming languages, Web performance, Internet of Things, databases, free and open source software, and more. My editorial output at O'Reilly Media included the first books ever published commercially in the United States on Linux, the 2001 title Peer-to-Peer (frequently cited in connection with those technologies), and the 2007 title Beautiful Code. He is a regular correspondent on health IT and health policy for He also contributes to other publications about policy issues related to the Internet and about trends affecting technical innovation and its effects on society. Print publications where his work has appeared include The Economist, Communications of the ACM, Copyright World, the Journal of Information Technology & Politics, Vanguardia Dossier, and Internet Law and Business.


  • We need so much progress in the thinking on this subject. Thanks for raising it.
    A system becomes rational when participants at all levels apply their unique concepts of value to informed decisions. Top-down authoritarian decision-making contributes to irrationality. The sparsity of trustworthy and discriminatory data (by “discriminatory” I mean data that really enlighten the true effectiveness of a treatment), and the opacity of cost information make being informed nearly impossible.
    Objective quality measures in any case, important and helpful as they are when done well, will never be discriminatory enough, inclusive enough or accurate enough to be the whole answer. Without giving up on efforts to improve the objective data and disseminate it widely, I wonder if it wouldn’t be better to admit that even the smartest bureaucrat can never have enough data to get close to a true assessment of value and maybe it’s best to leave that to the consumer, the one closest to the reality and most affected by the answer, who can express her opinion through her health care choices. Allow providers to innovate and morph as they believe will improve their services, improve the performance data available to her, move cost data away from opacity and towards visibility, and leave the choices to her.
    Frankly, if we don’t do that here, I’ll bet it will happen in Costa Rica or India, and long plane trips will be required for us Americans to get excellent health care of good value.
    A recently-retired community cardiologist

  • Thanks Stephen. You bring up an issue I’m exploring now: to make fee-for-value work, what data do clinical sites have to collect and share? How do we crunch it to come up with accurate ways to reimburse doctors?

Click here to post a comment