Sunday, December 20, 2015

Integrated Health Systems and Cost: The Price is the thing!

When the Affordable Care Act (ACA) was being developed, much emphasis was put on the effectiveness of integrated health systems as a way to save money but still deliver quality health care. Many studies from various research centers had looked at cost to Medicare and found that places – usually smaller cities – with large integrated health systems spent less on Medicare without noticeable decrements to quality. These systems can have a single provider of both inpatient and outpatient care (such as the Mayo Clinic) or close collaborations, including shared electronic medical records (as in Grand Junction, CO). The presumption of policy makers creating ACA was that Medicare spending, which is much easier to track, would reflect overall spending. However, a recent article from the National Bureau of Economic Research by Zack Cooper, Stuart Craig, Martin Gaynor and John Van Reenen, The Price Ain’t Right? Hospital Prices and Health Spending on the Privately Insured, demonstrates that this assumption was incorrect. Reviewing overall costs in the 306 Hospital Referral Regions (HRRs, developed by the Dartmouth Atlas of Health Care) in the US, they discovered wide discordance between Medicare costs and overall healthcare costs. Indeed, many of the places that were highly-touted for lower-Medicare-costs-but-still-high-quality, notably Grand Junction, CO (which was, for example, cited as a success story by Atul Gawande in his June, 2009 New Yorker article “The Cost Conundrum”) have far higher than average costs overall. (Dr. Gawande has just had a new piece in the New Yorker discussing the implications of this new article.)

The New York Times coverage of this study, by Kevin Quealy and Margot Sanger-Katz, The Experts Were Wrong About the Best Places for Better and Cheaper Health Care (December 15, 2015), includes a terrific feature that allows interactive access to the data collected by Cooper and his colleagues. You can put in a town (really, HRR) and find out where it ranks in terms of both Medicare and private costs. Grand Junction, for example, while ranking 3rd lowest of the 306 HRRs for per-capita Medicare spending, was the 42nd most expensive for private insurance spending. Rochester, MN, home of the Mayo Clinic, is another city lauded for its low Medicare costs (14th lowest), but its private spending is 10th highest! McAllen, TX, cited by Gawande in 2009 for being #1 in Medicare spending (and now still #4) is only 140th in private insurance spending. Tucson, AZ, on the other hand, while only in the lower middle (82nd from the bottom) in Medicare spending, is 7th lowest for overall costs. The Kansas City region, where I live, was atypically near the middle for both, 142nd lowest for Medicare and 82nd  lowest for private costs. New York City is high in both, but it is 2nd for Medicare and 34th (quite a bit lower) for private insurance. The map in the article depicts HRRs as low, middle or high for both Medicare and private insurance.

So, what’s up? Were the experts trying to fool us? No, but the flaw  was the assumption that Medicare spending reflected overall spending. The data in this article demonstrates that it does not. It also reveals something about integrated health systems, especially those that dominate their smaller cities, given that some of the “top performers” for Medicare, like Grand Junction and Rochester, MN, were so high for private insurance. The integrated nature of these plans allows them to save money on patients by a variety of methods – they can be seen in ambulatory settings rather than in hospitals or ERs, and they share electronic medical record systems, and thus the information recorded therein, saving money by not having to repeat tests, x-rays, etc. This lower utilization is good for these health systems because Medicare is a relatively low payer, and because they can’t negotiate these rates – Medicare pays what it pays (it is a single-payer system, with minor regional variations). However, the same characteristic – being the dominant player in town – allows such integrated health systems to negotiate much higher rates with private insurers. Thus the mismatch; overall cost is a multiple of price for each service times the number of services delivered. These systems decrease the number of services for people insured by Medicare, for whom they cannot control the price (whether this does or does not decrease quality is a separate question) but they raise the price for services to people with private insurance. That places like Tucson and Kansas City have relatively lower prices for private insurance reflects the absence of a single large dominant system in those cities.

‘“Price has been ignored in public policy,” said Dr. Robert Berenson, a fellow at the Urban Institute, who was unconnected with the research’, in the Times article. Other health policy experts, such as Princeton’s Uwe Reinhardt, have been warning about this for decades. In the effort to pass the ACA, and please both providers and insurers, this point was in fact ignored, and it is the source of most of the common legitimate criticism of the ACA – that in many places decent health insurance policies bought through the health exchanges are unaffordable. With higher prices in these regions, insurers pass on the cost to their customers.  This is illustrated in the NPR story “Obamacare Deadline Extended As Demand For Health Insurance Rises” on December 18, 2015, which documents both the success of ACA measured by the large increase in the number of people signing up for coverage and their frustration at the frequently-high cost of this coverage. Of course, this is completely unrelated to the criticisms leveled at ACA by the Republican candidates for President and their allies in Congress, whose “solution” – abolish ACA – is Marie Antoinette-like. While the French queen is reputed to have said, in response to being told that the peasants had no bread, “then let them eat cake!”, Republicans, hearing that many people cannot afford health insurance on the exchanges even with subsidies, or get Medicaid in states (that they control) which have not expanded it, respond “let them pay out of their own pocket!”

The issues and solutions are clearly laid out by the reliably insightful Dr. Don McCanne is his “Quote of the Day” on this topic. A solution cannot come from a jerry-rigged program that allows either insurers or health systems or both to maximize their profit. It needs to come from a system that starts with price controls, most effectively by a single-payer system such as Medicare. There are, as he notes, still risks – mainly that health systems may under-utilize services when they cannot make profit, leading to lower quality of care. But we can guard against this both on the regulatory end, by measuring quality outcomes and holding providers responsible, and through the market because the incentive to not provide services to Medicare patients because they can be more profitably provided to the privately -insured (the “opportunity cost”) goes away.

The infatuation of both policy makers and providers for integrated health systems is not entirely misplaced. The potential savings from shared data and not repeating tests, and more importantly for caring for people in the most clinically appropriate setting (inpatient, ER, outpatient surgery center, primary care, long-term care) is a real positive feature of these systems. But to the extent that these providers are allowed to use their market muscle to raise prices to insurers which are passed on to beneficiaries, it becomes a real negative.

The key feature of a good health system is that it is not focused on balancing the financial interests of big insurers and big providers, but that it puts the benefits to patients, to the people’s health, first.

Saturday, December 12, 2015

Medical interventions we shouldn't be getting: issues of cost, health, and equity

Many of us are concerned about the cost of medical care, particularly in the US. We are also concerned about the care that we, ourselves or our families and friends receive. We want the diagnostic tests and treatments we get to be likely to benefit us and to not cause us harm. Or, if there is chance that we may be harmed, that the probability and degree of benefit exceeds the probability and degree of harm. Sometimes there can be a tension here: we want to spend less on health care globally (and certainly less on the portion that just goes into the pockets of big corporations, like pharmaceutical companies and device manufacturer, insurers and huge health provider networks), but don’t want to scrimp on things that may benefit us. Especially when we are well-insured and not paying for it directly out of our pocket.

Fortunately, there are many times when these two different interests come together, when tests or treatments that are costly are also of no benefit to us, and may cause us harm. The problem is that sometimes our doctors recommend them to us anyway, or, frankly, we ask for them (because we’ve read about it on the Internet or seen an ad on TV or because a doctor recommended it for our Cousin Shirley who had the same thing, or at least something that sounded a little the same). How can we, as patients, know when our doctors are recommending a test or treatment that is not indicated for us, will not benefit us, may even cause us harm? How do we know when we are appropriately advocating for ourselves as opposed to asking for something inappropriate? It is not easy, but we can try.

One answer is NOT to reject all medical recommendations for diagnosis or treatment, to assume that something “natural” is always right. Medicine can do a lot; it can treat a lot of illnesses that can cause you to suffer or die prematurely. There are many tests that are proven useful or valuable for making diagnoses. Another is to not reject things simply because they cost money (difficult when you don’t have much and they cost a lot), but also not to ask to have them because you have money (or insurance). To a great extent it is asking your doctor questions, asking about the degree of benefit, hoping that they are up to date, trying to question what the likely benefit is from another test or treatment, especially if costly. It also, sadly, means understanding if the doctor themselves, or a company they work for, stands to make profit on the test or treatment, since this creates a conflict of interest (but certainly does not mean that it always or usually isn’t appropriate).

There is an entire medical literature on medical overuse, and your doctor should be familiar with it, or at least the most egregious overuses that are in her/his field. A recent article in JAMA Internal Medicine , “Update on Medical Practices That Should Be Questioned in 2015”, by Morgan et al., reviews some of these.[1] The authors reviewed the literature on articles published in 2014, and came up with 104 that at least one of the 3 thought was very important, and 33 that all agreed were very important. The article reports on the “Top 10”, because this is a nice, round number and because it seems to be a manageable number of things for doctors to remember. Just these 10 represent a lot of excess tests and excess expense.

I will summarize some of these 10 (the ones I feel are most relevant; you can find them all in the article), both to remind doctors and other health care providers who might not have seen it, and because other readers may have had these tests recommended. Before doing so, it is important to review the difference between a screening test, done on the general population (or a subset of it) who don’t have any symptoms, and a diagnostic test, which may be the same test done on someone with symptoms related to what that test tests for. Frequently a test is useful for diagnosis for someone with symptoms but in someone without may not only be of little use, but find “false positives” that end up leading to more testing with more cost and greater risk.

  • There Is No Benefit to Screening for Asymptomatic Carotid StenosisNone, for screening. Ultrasound or other tests may be of value for people with TIAs or other symptoms.
  • Screening Pelvic Examinations Are Inaccurate in Asymptomatic Women and Are Associated With Harms That Exceed Clinical Benefits. Again, none, for screening. You can’t screen for ovarian cancer. I have been trying to convince my residents of this for years, despite their being often taught otherwise by OB-Gyns. This is the part where the provider puts his/her hands inside. It is NOT the same as a Pap smear, which is recommended for screening.
  • Head Computed Tomography Is Often Ordered but Is Rarely Helpful. Often finds clinically insignificant abnormalities. Also, often repeated with no added benefit. (“A retrospective cohort analysis reviewed 130 patients admitted for any cause at least 7 times during a 1-year period to a tertiary care center. Patients received a mean of almost 7 CT scans, including 3 head CT scans. More than one-third (36%) of head CT scans were ordered to evaluate for altered mental status. Only 4% (7/127) of head CTs had clinically significant findings that resulted in a change in management.”)
  • Thyroid Cancer Is Massively Overdiagnosed, Leading to Concrete Harms.  Study is from S. Korea, where they screen for it.
  • There Is No Benefit to Paracetamol (Acetaminophen) for Acute Low Back Pain. “The median times to recovery were 17 days in both of the paracetamol groups and 16 days in the placebo group.” This does not mean we should be using opioids, though. In fact, when we give opioids for good reasons, like postoperative pain…
  • Postoperative Opioid Use Continues Past the Postoperative Period.


Some overuse comes because practices once felt to be appropriate, but now shown not to be, are being taught to trainees, who either don’t know or are loathe to disagree with their instructors. Some comes, consciously or not, from the potential for providers to make money. Another recent JAMA Internal Medicine[2] study looked at family medicine and internal medicine residents who saw “secret shoppers”, standardized patients who were requesting inappropriate imaging. About 25% of the time the tests were ordered, which could be seen as either bad or good (75% of the time they weren’t). The study also showed no difference between those residents who got significant feedback and those who didn’t, and similar (and generally good) techniques of communicating to the patients why these were not indicated. An interesting wrinkle is that some of the tests chosen (MRI or CT for new-onset uncomplicated low back pain) were imaging studies not recommended by the American Academy of Family Physicians and American College of Physicians (Internal Medicine) as part of the “Choosing Wisely” campaign, but are not recommended against by the  American Academy of Orthopaedic Surgeons in their “Choosing Wisely recommendations. Of course, orthopedists stand to benefit from doing surgery on these patients. (By the way, among the recommendations of the AAFP is not doing screening pelvic examinations!)

What is the social justice issue here? After all, tests and treatments that are not indicated should not be done on anyone, regardless of financial or insurance status, racial or ethnic characteristics. Indeed, one could argue that those with more money or better insurance would be the most likely recipients of extra tests. But poorer, or less well-insured, people often don’t get the tests and treatments that they actually need (which has often been discussed in this blog; see for example Dead Man Walking: People still die from lack of health insurance, November 17, 2013), while tests and treatments are overused for others. This creates the (accurate) perception of a two-class health care system, and the (sometimes, perhaps, less accurate) perception that the disadvantaged are, in addition to not getting care they need, are not getting discretionary care that benefits the wealthy. Also, the more money spent globally on unnecessary medical care, the less available for necessary non-medical initiatives (housing, food, education) that would actually create better health.

A single payer system would not in itself achieve this goal, but it would create far more equity in the allocation of healthcare dollars, and at least eliminate the profit motive that might impact the recommendations of some providers and healthcare institutions.





[1] Morgan DJ, Dhruva SS, Wright SM, Korenstein D, Update on Medical Practices That Should Be Questioned in 2015, JAMA Intern Med. 2015 Dec 1;175(12):1960-4. doi: 10.1001/jamainternmed.2015.5614.

[2] Fenton JJ, Kravitz RL, Jerant A, et al., Promoting Patient-Centered Counseling to Reduce Use of Low-Value Diagnostic Tests: A Randomized Clinical Trial, JAMA Intern Med. Published online December 07, 2015. doi:10.1001/jamainternmed.2015.6840

Total Pageviews