Economics

In Defense of Narrow Networks

It wasn’t long ago that the newly established health exchanges were being celebrated. Before the ongoing website catastrophe, politicians and policymakers were lauding the low premiums in these new health insurance market places. On September 24, President Obama said, “And the premiums are significantly lower than what they were able to previously get … California — it’s about 33 percent lower. In my home state of Illinois, they just announced it’s about 25 percent lower.”

How times have changed! Even supporters of the exchanges have rightly criticized the technical problems that have prevented millions of Americans from signing up. However, many critics are also complaining about the large number of health plan offerings with “narrow networks” of physicians that enrollees can visit for medical services. The Missouri Health Advocacy Alliance expressed “major concern” when Anthem excluded BJC HealthCare from its narrow network. Seattle Children’s Hospital, which was excluded from several exchange plans, has sued the Washington State Office of Insurance for “failing to ensure adequate network coverage.”

Criticism of narrow networks is misguided and counterproductive. As we explain below, narrow networks will be of little consequence to most of the individuals who sign up for the exchanges, and the elimination of narrow networks could eliminate our single best opportunity to harness market forces to reduce costs and improve quality. Indeed, narrow networks are largely responsible for the low premiums that were being celebrated just one month ago.


We admit that narrow networks may seem like a bad idea. They limit where patients can go to receive care and threaten to interrupt the physician/patient relationship. But there are two major flaws with this line of thinking. First, patients have a choice of many different health plans in the exchanges and these plans all have different network options. A provider who is not in one plan’s narrow network is likely to be another. Patients whose providers are not in any plan’s narrow network can always choose broad network plans in exchange for paying a higher premium. They will be no worse off than they are today, and if competition in exchanges works out as planned they may even be better off.

Once they sign up for a narrow network plan, there is no guarantee that patients will receive care from in-network providers. Big medical bills may result. But we doubt this is likely to be a big concern for very long, as patients learn to navigate the new networks. Seattle Children’s Hospital is rightly worried that some enrollees in narrow network plans will end up at their doorstep. But there are other high quality providers of pediatric services in Seattle. Once patients and referring doctors get used to the new networks, the only children who show up at Seattle Children’s Hospital will be those whose networks include the hospital, or those whose parents are willing to pay for out of network care.

If more of us move into exchanges (something that the Affordable Care Act actually stifles…see our previous op-ed on this topic), we may all have to get use to narrow networks. Employers rarely offer narrow networks because it is very hard to find a single network that appeals to all (or even a large fraction) of their employees and too expensive to offer a large number of different plans. Once individuals are buying insurance for themselves, one-size-fits-all insurance will go by the wayside and people can select the plan and network that best matches their needs.


Narrow networks are not some cruel attempt to limit patient choice foisted upon us by the insurance industry. Instead, these plans may provide our best opportunity for harnessing market forces to lower prices. Even high priced providers know they stand a good chance of being in broad networks. But insurers offering narrow networks can be picky about which providers they select. Across the nation, high quality/high price sellers like Seattle Children’s Hospital will have to prove their worth.

What if insurers ignore quality? If we have learned anything about quality in the past decade, it is that insured patients making their own provider choices have done little to reward measurable high quality, instead relying on more on brand names that may or may not indicate true quality. Will insurers be any worse? While it is theoretically possible that narrow network plans will focus on low costs, quality be damned, we are unaware of any narrow networks that include only the bottom of the quality barrel. It is also hard to imagine how it would be profit maximizing for all insurers or potential entrants to the exchanges to offer only low-quality narrow network plans. Rival insurers will surely be quick to point out the shortcomings of low quality competitors.

As a nation we have reached a consensus that we must lower medical spending. While this is often presented as a choice without trade-offs, that simply is not the case. Making our lower health care cost omelet is going to require breaking some eggs. Most Americans do not place must trust in insurers, but through narrow networks, insurers can introduce some much needed cost discipline on providers. And narrow networks can even include ACOs, should they offer proof of concept.

The intensified competition from narrow networks will be messy…patients will make mistakes, and quality will sometimes go unrewarded. This is not unlike our current system, only it will be less expensive and with greater access. The only sure fired alternative way to controlled cost is centralized planning. While some have faith in the ability of bureaucrats to choose what services to cover and how much to pay for them, we are less sanguine about the role of central planning in this and other settings.

David Dranove, PhD is the Walter McNerney Distinguished Professor of Health Industry Management at Northwestern University’s Kellogg Graduate School of Management, where he is also Professor of Management and Strategy and Director of the Health Enterprise Management Program. He has published over 80 research articles and book chapters and written five books, including “The Economic Evolution of American Healthcare and Code Red.” This post first appeared at Code Red.

Craig Garthwaite, PhD is an assistant professor of management and strategy at Northwestern University’s Kellogg Graduate School of Management.

Livongo’s Post Ad Banner 728*90
Spread the love

27 replies »

  1. Yup, I had a very unhappy experience at our community hospital-wrong meds and adverse reactions. They had the nerve to bill me for it! $4,000 of wrong meds! I didn’t pay and wrote them a letter-its been six month and no followup bills. I also never got a survey even though all of their “advertising” said I would.
    Narrow networks here in CA are very narrow indeed. I wouldn’t ever consider signing up under CoveredCA because I don’t think there will be a dozen primary physicians in our small city who will be in the network. None of my doctors are and don’t plan on signing on. My understanding is that I might CoveredCA is by county-the problem is my county doesn’t have an outstanding medical facility-the hospitals (3) send everyone out of county for specialized care. All high risk premies; life threatening accidents; major heart attacks; all go 100 miles away. Not allowed under CoveredCA. I predict that the ACA will be a heartache for many, many people as they discover that health insurance does not mean healthcare.

  2. The difficulty with pts. picking the network that meets their needs is that health needs can change unexpectedly, and the pt. may not be in a good position to be trying to make those assessments when they do.

  3. The creation and designation of a provider network has traditionally been a health plan responsibility.

    This is shifting. New species of provider networks are being created by Medicare ACOs, accountable-care like initiatives between care providers and insurers, provider clinical integration initiatives, etc.

    Creation of a (narrow) provider network increasingly will be viewed as a joint health plan/provider responsibility — or even a sole “provider responsibilitiy”, e.g., an ACO contracting directly with employers.

    Thus it becomes increasingly difficult simply to attack “the health plan” for a narrow provider network.

    Recalling Pogo’s wisdom: “We have met the enemy and he is us.”

  4. Narrow networks conceptually sound great. However the reality is that they mean creating 3 tiers of service:
    One for those with jobs that have health insurance
    One for those who have to buy their own insurance
    One for Medicaid.

    You can sing the praises all day. But that is the reality. Narrow networks eliminate the high end specialty service providers. That means people get less care.

  5. Legacyflyer –

    I agree with your last comment about the challenges inherent in measuring quality among surgeons especially since the risk adjustment state of the art isn’t where it needs to be (yet).

    From a patient’s perspective, I think it would be helpful if it were easy to find out how many times a surgeon performed a given procedure both in the past year and cumulatively over the course of his or her career. At the same time, I would want to know the minimum number of procedures per year the experts think are required to keep skills sharp and up to date.

    For hospitals, I would like to see comparative data on the number of CLABSI and VAP infections per 1,000 ICU bed days various hospitals have and whether there is a trend of improvement or deterioration in those metrics. The process metrics you referred to are better than nothing but far from a complete or adequate picture.

    Regarding teachers who teach lower income students vs. those who teach in wealthy suburbs, it should be possible to take this into account by making like for like comparisons based on socioeconomic status which can be determined by looking at the percentage of students who qualify for free or reduced rate school lunches. That said, I think good teaching is something that comes under the heading of: I know it when I see it or experience it but I can’t necessarily prove it in court or reduce it to contract language.

  6. Peter1,

    I can answer that question.

    Despite having had the HSCRC in place for about 30 years, Maryland remains a high cost state (based on Medicare data)

    What has happened is that a number of high cost procedures have been “squeezed out” of hospitals into outpatient facilities which are not regulated by the HSCRC.

    Despite what a variety of proponents will tell you, the HSCRC has largely had no effect in controlling the cost of health care – its stated aim.

  7. Barry, how does Maryland’s cost of care measure against other states without the uniform pay system?

    As to lawyers charging more for the same work (not including double billing) cause they’re “experienced” does not mean they do a better job – it just means they can. Some lawyers charge a % of the selling price to close a house (in a lawyer state), that’s a scam as the only difference is the decimal places not the work.

  8. Barry,

    I practice in Maryland. Part of the higher cost of Hopkins and U of MD reflect their teaching/training/research mission. Part of the higher cost reflects payor mix because of their location.

    Getting back to how to measure quality and surgery. It has been shown that physicians and centers that do more of a particular procedure tend to have better results. I trust you went to a Cardiac Surgeon who does a lot of cases.

    More than that is hard to judge. One surgeon could take on more difficult cases than another and although he has worse results could actually be doing a better job.

    Doctors, like teachers don’t want to be judged on their results if the data has not been properly corrected for other confounding factors. For example a teacher teaching in an inner city school is not going to have the same results as a teacher teaching in a wealthy suburb. Similarly a doctor practicing in an upper middle class area is generally going to have better outcomes than one practicing in a poor area. This is despite the best efforts (or not) of the doctors and teachers.

    So to substitute for any real ability to measure quality – absent confounding factors – quality raters use various “quality indicators” such as percent of patients who get mammograms, percent of diabetics who get Hem A1C, etc. However, these “quality indicators” are necessarily incomplete and hence more of a sketch (or caricature) of quality than a real picture.

  9. “I hope for a day when all providers and hospitals in a regional area charge the same regulated fees and let the patient choose their own provider.”

    Peter1 –

    As you probably know, the state of Maryland implemented an all payer system for hospital based care in the late 1970’s. That means that every payer pays a given hospital the same price for a given service, test or procedure. However, academic medical centers like Johns Hopkins are paid more than community hospitals for similar work to reflect their inherently higher costs. In addition, hospitals in rural Western MD are paid less than similar hospitals in the Baltimore-Washington corridor due to differences in medical input costs, especially for labor and real estate.

    Moreover, I don’t think it’s appropriate for a new doctor just out of residency training to command the same fees as an established veteran with 25-30 years of experience. By the same token, senior partners in corporate law firms bill at a much higher hourly rate than young associates just out of law school which is the way it should be.

    While there are certain types of care that only sophisticated academic medical centers can handle, most care they provide is comparatively routine and can be done equally well by less expensive community hospitals. As one veteran doctor pointed out some time back, community hospitals do common things commonly every day and, for the most part, they do them well. There is no reason why patients should expect to go to an academic medical center for such care and burden payers, whether public or private, with much higher costs unnecessarily.

  10. Networks, narrow or not only serve insurance and providers – not the patient. They are designed to protect markets and restrict access.

    I hope for a day when all providers and hospitals in a regional area charge the same regulated fees and let the patient choose their own provider.

    If a doctor has a particular skill and experience that a patient wants/needs why should the patient pay more only because their insurance carrier does not have a payment contract with the doctor.

  11. Legacyflyer –

    As I think you well know, I’ve never worked in the medical field so this is not my area of expertise. However, my layman’s perception of outcomes as discussed here relates mainly to surgical procedures. In other words, how well did the surgeon do his or her job given the complexity of the case as well as the age and overall health status of the patient? If I get heart bypass surgery, which I’ve had, the surgeon could do a fine job but I could get a hospital acquired infection later. Under those circumstances, I would credit the surgeon with a good outcome but mark the hospital down on patient safety.

    Regarding your comments about radiologists, maybe they should all be considered solidly competent especially if they’re board certified. Quality measurements will always be imperfect and any system can probably be gamed to some extent and patient satisfaction scores need to be taken with several grains of salt.

    I do think we need to make a serious effort to define and measure quality though. At the very least, doing so can let hospitals and doctors see how they’re doing. Where there is room for improvement, the team should try to identify the systemic changes needed to improve performance. The most cost-effective high quality providers should be rewarded with both higher reimbursement rates or shared savings and more patients but we need data to identify who these providers are. Don’t we?

  12. Barry Carol,

    I generally find myself on the same page as you. But, explain to me the difference between “Outcomes” and “Safety”.

    “2. Outcomes – preferably risk adjusted.”
    “3. Safety – minimizing infections, preventable readmissions, etc.”

    To the extent that “Safety” really measures safety and not some politically correct abstraction, why wouldn’t it show up in “Outcomes”. In other words if one place has a higher infection rate, why wouldn’t that show up as an “Outcome” of higher morbidity and mortality? And if it doesn’t, why do we care about it?

  13. Let me give you some sense of how hard it is to develop a “valid quality metric” – at least in my field.

    I am a Radiologist. The difference between a great Radiologist and an average or poor Radiologist is probably only a couple of percent. The reason is because: most X-Rays are normal or near normal, most people aren’t that sick, most people get better (or die) anyway and/or the difference in quality of their surgeon, internist, etc. overwhelms the difference in the Radiologist anyway. To get statistically valid morbidity and mortality data attributable to Radiology would probably require good data (which we don’t have) and sample sizes of over 100,000 patients. In other words – forget it.

    So how do you get statistically valid data on Radiology quality? Basically, you don’t. Instead you substitute other “metrics” like “patient satisfaction”. And the patients generally have no idea of the quality of their films or the quality of the reading. What the patients know is; how easy it was to schedule, how nice the waiting room was, how long they waited and how nice the secretaries and techs were.

    And our techs aren’t stupid. If they are taking care of a nice old lady who calls them “dear”, they hand her a survey. A pissed off yuppie – no survey. Guess how valid that “random” survey is?

    In essence – don’t hold your breath waiting for quality metrics. On the other hand any insurance company employee (or government employee) can easily figure out the price. And a “narrow network” allows them to choose the cheapest.

    Get used to it.

  14. Good list.
    And there’s also the fun question of “at what level?” Eg, Hospital System, Hospital Site, Office, Physician.

    And with Outcomes adjusted for degree of difficulty of the patients.

    Easy 🙂

  15. Measuring care quality, especially in hospitals, is indeed a challenge to put it mildly. I’ve heard experts suggest that quality in this context has four main components each of which would have to be appropriately weighted. They are:

    1. Process – following evidence based guidelines and protocols.
    2. Outcomes – preferably risk adjusted.
    3. Safety – minimizing infections, preventable readmissions, etc.
    4. Satisfaction – which could encompass anything from the competence and responsiveness of the nurses to the quality of the food to weather the room has a flat screen TV and the hospital offers valet parking.

    If it were up to me, I would weight #1 and #3 and 20%-25% each, #2 at 40%-50%, and #4 at 10% at most.

  16. Well, at least the nation has reached a consensus about something: lower health care spending. My guess is we have a ways to go before finding the solution that actually accomplishes this. Until then, we just have to keep trying.

  17. Personally, I do not.
    But I do like to support and encourage those who are trying.

    Maybe just needs some magic, like Harry Potter and FL.

  18. Kevin,

    So you believe a “valid quality metric” will be available in several years?

    Are you holding your breath?

    “In several years” I will have a bridge to sell you or perhaps some swamp land in Florida?

  19. A valid Quality metric is a key linchpin that is lacking today (and for several more years), as you pointed out.

    Sure, lots of BS out there. But the growing prevalence of reasonable narrow networks is a net positive for industry evolution as a whole.

    The rub, we know, is that medicine is always personal, about the individual.

  20. A “narrow network” is basically a euphemism for an insurance company choosing the lowest bidder.

    And yes, I know that they are supposedly going to be choosing based on: “quality”, “patient centered metrics”, etc. etc. blah, blah, blah. But having been through “Hillary Care” in the 90’s I have seen how the “rubber meets the road”. Most insurance companies (and dare I say the Feds) have no good way to measure quality. So instead (and quite logically) they choose based on price.

    There is actually nothing wrong with this. I believe that the primary problem with American medicine is its price. By using a “narrow network” – i.e. picking providers based mostly or exclusively on price – insurance companies are able to drive down prices much more effectively than if they have to include a whole bunch of providers in their network.

    So lets cut the BS and tell it like it is. In order to save money, insurance companies, networks, etc. need the freedom to negotiate prices aggressively. In order to do so, they will (mostly) choose based on price. The claim that “you can keep your doctor”, while theoretically true, will be false for many most people. Your doctor will be the lowest bidder.

    In reality, this probably won’t make that much difference to people’s health and will save a lot of money. Just don’t believe all the BS from people trying to spin it another way.

  21. Vince — Agreed. Great points. Kaiser is one strong example.

    Does anyone know how the term “narrow networks” really got hold?
    My guesses:
    1. POSITIONING — Insurers themselves offering a lower priced product while really focused on selling higher priced products.
    2. MEDIA — Playing to consumer fears (some legit, some not) of such networks is always good for some extra page views, as is some catchy alliteration.

    Not sure how/when the industry will evolve its nomenclature to better reflect the benefits.

  22. Narrow networks, along with tiered networks and the use of reference pricing where it makes sense are the only strategies that I can think of to create countervailing power against large hospital systems and physician groups with excessive local or regional market power.

    I do think there needs to be special rules governing how much can be charged for care that must be delivered under emergency conditions as compared to care that can easily be scheduled well in advance. I also worry about what happens to patients who wind up in network hospitals but are treated by (some) non-network doctors and then receive large bills for the balance due beyond what the insurer paid. This is especially galling in the case of radiologists, pathologists, anesthesiologists and emergency room doctors, all of whom patients typically have no role in choosing.

    As for academic medical centers that specialize in treating children, they are saddled with inherently high costs. For one small example, I’ve read that they must stock eight different blood pressure cuff sizes to accommodate everyone from premature infants to hulking teenage football players. I’m not sure what the answer is here especially when such hospitals may be the only facilities capable of providing rare cancer treatment, organ transplants or complex surgery.

  23. “Sell the sizzle, not the steak”

    The very term “narrow network” creates a negative connotation and doesn’t speak to potential benefits. It only sells “steak”.

    Others are describing the same networks as “high-performance networks” or “intelligent networks”. The connotations here are much more positive.

    No need to be defensive here in deploying these networks. The value proposition of improved quality and reduced cost is strong and easily communicated. Yes, there are tradeoffs of limiting choice, but why focus on the negatives? In what other areas can consumers expect unlimited choice without parallel tradeoffs?

Leave a Reply

Your email address will not be published. Required fields are marked *