Yes, it's the insurance and drug companies, but it's the health systems also!

Much of the focus in discussions about the high (extremely high!) cost of the US healthcare system is on administrative costs (over 34%, per Annals of Internal Medicine article by Himmelstein, Campbell and Woolhandler)[i] and the profits taken by insurance companies and drug companies. This is totally right on, and are major reasons why a single-payer #Medicare4All system would save enough money to not only cover the tens of millions of Americans who currently do not have health insurance, but to provide decent, comprehensive coverage to the majority of Americans who have inadequate health insurance. This, of course, includes those who have marginal health insurance (like Blake Collie, an 8 year old boy with a cerebral aneurysm, whose parent bought a Christian insurance plan that was all they could afford, but would not pay for the treatment; the advice was “trust in God”), and those who have Medicaid, now aggressively being cut back, especially with new Trump Administration policies that allow states to slash it. And all the rest of us who pay large premiums but only find out what our insurance doesn’t cover and what our co-pays will be when we get sick. And then we get sicker.

In addition, we have Medicare Advantage (MA, also known as Medicare Part C) a deal which essentially can make a Medicare patient (for a little extra premium) an HMO patient.  For the patient this comes with the typical HMO advantages (such as vision and hearing and other coverages, and usually drug coverage so you don’t need a separate Part D insurance plan, and often little or no co-pay) and disadvantages (limited networks of doctors and hospitals, often poor out-of-area coverage), but it has real negative impact on the health system. MA programs get special treatment from the Center for Medicare and Medicaid Services (CMS), including a big increase in payments if they can demonstrate, with all the wizardry, bells-and-whistles, and large staff combing and padding the Electronic Health Record (EHR) that their patients are sicker. In fact, they are demonstrably less sick than traditional Medicare patients; indeed, the older and sicker MA patients are encouraged, subtly or not, to transfer to traditional Medicare. A recent article by Richard Kronick on the Health Affairs blog demonstrates that MA programs are being overpaid by $200 Billion. This is real money, and it is not by accident, as discussed by Don McCanne in his Quote of the Day.

But the less-discussed contributor to our high health costs are the hospitals and health systems that make big money. This is true even when those health systems are ostensibly not-for-profit. The for-profit hospitals and health systems are at least open about it, and they pay taxes. Non-profits do not pay shareholders, and also do not pay taxes. The presumed reason for this is because of the public service that they provide to their communities. But while this may be true of many small-town and rural hospitals, which are also in the most danger of closing and leaving their communities bereft of hospital care, it is often not at all the case for large urban health systems. They make money. For example, an interview on the NPR program “1A”, otherwise focused on the Peak Health System in Summit County, CO that has had some success in reducing costs in that rural tourist county, the Colorado insurance commissioner Michael Conway discusses the proposal for a “public option” in his state and notes that asking hospitals to give back a little is not too much for the urban hospitals making $2 billion a year (about 28 min in).

And, since they do not have shareholders to pay, the health systems reinvest most of the money they make into expanding hospital services or building new buildings. This could seem like a good thing, except that the choice of what services to expand is often (usually) based not on what services the community needs most, but what services will – make them more money! And, also, hopefully make them more desirable destinations for high-margin services than the other urban hospital systems with which they compete. So if, as is usually the case, cancer care and heart disease care and orthopedic care are big money-makers, they build fancy new cancer hospitals and heart hospitals and orthopedic hospitals to try to draw well-insured patients away from their competitors. Thus, we get redundancy and overcapacity in these high-end services in competitive cities, with each shiny new cancer center seeking to lure patients from the one that’s a few years older across town.

Meanwhile, these urban communities do have other, less lucrative, needs that are scarcely ever the target of major investment. In a rational system, the money made on those “profitable” services could be used to invest in and subsidize lower-profit (or in some cases money-losing) services that are in great demand in the community. These certainly include primary care, mental health/behavioral care, drug addiction treatment, and virtually any care delivered to poor or uninsured people. A reasonable health system would just subsidize this care, and not bill and dun people repeatedly for money that they do not have, and cannot and will not pay, ruining their credit, attaching their wages, and challenging their ability to pay for other things, like shelter, food, and clothing.

But privately-run health systems almost never work this way (even if “non-profit”) because their boards like them to make money. And reward them for making money. An article in GQ in April, 2019, reports that CEO salaries at our big health system are doing very well indeed; at the 62 largest, the 2018 salary averaged $18 million! The other “C-suite” executives (COO, CFO, CMO, etc.) are also very well paid; it would not be rare for a major hospital system to have 10 executives making over $1M a year. These CEOs are sometimes doctors, but often accountants or MBAs – basically they run their hospitals as a business, and often have no other context to relate to. When thinking about community benefit, they think of “community” as the “community of well-insured”, the “community of suburbanites”, and of course, especially the “community of potential donors” (cool that they can even get people to give them money and take a tax write-off!). They almost never think of the “community of need”. They are sometimes briefly interested as long as government will give them money for helping the needy, but return-on-investment (ROI) is always measured in dollars, not population health.

This is what happens when the private sector is given control of an industry; they pursue their own benefit. It is unconscionable that we do this in areas like health care which are needed by everyone. Virtually all decisions are made with their eyes on the bottom line. Is CMS going to pay for more residents? Let’s get more cardiology fellows to do procedures and make us money, not more family medicine residents who will go out and meet the needs of people, even if not in our hospital.

There are people who are concerned about government-run health care, and this spills over to their concern about government-financed health care, such as Medicare for All. The problem is that there is a status quo, and that status quo is destructive to the health of our people.

It needs to change.

---

[i] Himmelstein DU, Campbell T, Woolhandler S, “Health Care Administrative Costs in the United States and Canada,

2017”, Ann Int Med, doi:10.7326/M19-2818, published online Jan 7, 2020.

Primary care specialty choice: student characteristics

.
I have written about both the characteristics of medical schools (recently, in A New Way of Ranking Medical Schools: Social Mission, June 20, 2010; also Rankings of Medical Schools: Do they tell us anything?, September 5, 2009) and of medical students, including the people being accepted into medical school, the specialty choice of those graduating, and what aspects of both schools and students are associated with the kind of doctors students become (Who will care for the underserved? The role of off-shore medical schools, June 2, 2010, Primary Care and Residency Expansion, January 7, 2010, "Uncomplicated" Primary Care?, October 8, 2009, Medical Student Selection, December 4, 2008). A recent study, “Primary care specialty choices of United States medical graduates, 1997-2006” (Academic Medicine June 2010;85(6):947-58) by DB Jeffe, AJ Whelan and DA Andriole from the Washington University School of Medicine in St. Louis provides further information to inform this discussion.

The authors used two surveys administered annually by the Association of American Medical Colleges (AAMC) to medical students: the Matriculating Student Questionnaire (MSQ) given when students begin school and the Graduation Questionnaire (GQ) given at graduation. They were able to match the questionnaires of nearly 2/3 (64.9%) of medical students graduating in the 10 years 1997-2006 to look at the degree to which the answers students gave to questions on the MSQ and GQ would predict their choice of specialty. It takes a while to get through the description of how they grouped questions, and “weighted factors”, but there are a number of important findings. Most of them not surprising, but this study provides additional data support for things we have been seeing.

Fewer graduates are entering primary care. For some reason the AAMC’s GQ counts 6 “primary care” areas, including, in addition to the usual family medicine, general internal medicine, and general pediatrics, also obstetrics/gynecology, internal medicine subspecialities (IMSS), and pediatric subspecialties (PdSS). However, this study separates them out. The percent of students entering the first 4 of these dropped from 1997 to 2006 (GIM from 15.7% to 6.7%, GP from 10.2% to 6.6%, FM from 17.6% to 6.9%, and OBG from 8.2% to 6.1%), while the subsubspecialties increased (IM from 6.8% to 11.4%, Peds from 2.2% to 4.4%). This increase in the last 2, however, did not compensate for the decrease in the first 4 so there was a net decrease from 60.7% to 42.1% for all these “primary care” specialties. The increase was in, then, surgical specialties, and more significantly in specialties such as radiology, anesthesiology, and emergency medicine. And the trend continues – although not part of the study, the article reports that the total % of students entering these 6 specialty areas in 2008 was down to 30.3% from the 42.1% in 2006. Remember, this is not just real primary care – it includes medical and pediatric subspecialties!

The more important part of this article is its correlation of certain demographic and attitudinal characteristics with specialty choice. Being female is important: over the 10 years of the study 45.1% of graduates were women, but they were over-represented in each of the 6 fields except IM subspecialties: 77.3% of OBG, 72.6% of GP, 58.2% of PdSS, 50.8% of FM and 49% of GIM. Indeed, the authors suggest that one of the reasons for the continued downturn in entry of medical students into PC fields is that the % of women in medical school, which was increasing during the study period, tended to compensate somewhat for the decreased interest in PC among men. Now that the % of women in medical school has stabilized, at roughly 50%, while interest in PC continues to go down, this is no longer having the same compensatory effect.

Compared to white students, underrepresented minority (URM) students were somewhat less likely to choose the PC specialties of GIM, GP, and FM, while Asian/Pacific Islander (As/PI) grads were less likely to choose FM, OBG, GP and PdSS, according to the text, but there are subtleties to this. For example, white students were 68% of the total but 75.3% of those entering FM; however, the % of URM students in FM (14%) was also greater than in the overall cohort (12.7%). The difference is that As/PI were much lower in FM, 9.9% compared to 18.2% of the overall cohort. In this report URM students include black, Hispanic, and American Indian/Alaskan natives, while the As/PI group includes some groups that are truly underrepresented in medicine and are usually counted by Federal grantmakers as URM, as well as some groups, e.g., South Asian, Japanese, Chinese, that are in fact over-represented in medicine. Many of these are among the 15.7% of students who had one or more parents who was a physician or the 24.1% more who had a non-physician professional parent. This is important because “…a student’s having a physician parent had a pervasive negative effect on graduates’ choice of any generalist-primary care specialty…” while those with non-physician professionals as parents were more likely to choose GIM, GP, and IMSS.

Though highly-touted as a deterrent to entering PC (and, perhaps, of more significance now than in the early part of this study period) debt had only a “modest” effect; students with higher debt were less likely to choose GP, PdSS, GIM, or IMSS, but more likely to choose OBG, and choice of FM was essentially unaffected. The probability of students from public medical schools entering PC was much greater than from private schools, a consistent finding of all studies, presumably reflecting the curriculum and emphasis of those schools as well as the characteristics of the students they select. The authors linked a number of questions to assess students’ “choice of medicine as a career goal” (including the importance of innovation and research, social responsibility, and prestige), and students’ “perceptions about medicine and medical practice” (including altruistic beliefs and the belief that the demands of medicine interfere with family/other interests). All were high, and few led to big differences in specialty choice. Among the most significant of those that did were “intention to practice in an underserved community” (more PC, especially FM), and “interest in academic faculty positions” (very low FM -- the odds of being interested in academic practice compared to full-time non-university practice was only 12% for someone entering FM, compared to GIM’s 86%, IMSS’ 185% and PdSS’ 316%!).

Among areas that the questionnaires did NOT look at was the probability of entering rural practice; other studies have shown that, except for family medicine, essentially no students are entering practice in rural areas (see Primary Care and Rural Areas, April 28, 2010, Medicare Costs in Rural America: A case of reaping what we haven't sown?, March 26, 2009, Ten Biggest Myths Regarding Primary Care in the Future, January 15, 2009 ). Despite their assertion (probably correct) that “…the predictive validity of planning to practicing in an underserved community at graduation has been established,” intention to practice in an underserved community is not a surrogate for intention to practice in a rural area, because practicing in a rural underserved community means living in a rural community, while one can practice in an urban underserved community and live in a more upscale neighborhood. This is particularly true for pediatrics (see Primary Care, Pediatrics, and Physician Distribution, May 21, 2009).

So what does all this tell us that is new? Not much. There is a dramatic decrease in the number and percent of students entering true primary care specialties, combined with some increase in the number entering obstetrics/gynecology, internal medicine and pediatric subspecialties and a large increase in specialties that are in no way primary care. It tells us that students who are interested in and attracted to, and are wooed by, “high-status” research and specialty care medical schools are less likely to enter primary care; that being the child of a physician or having a high family income makes one less likely to enter primary care; that students interested in caring for the underserved and women are more likely to enter primary care (although not rural practice). So what does this mean “we” should do?

Well, that depends on who “we” are. If “we” are the selective, high-status medical schools who are low on social mission (see A New Way of Ranking Medical Schools: Social Mission, June 20, 2010) and “we” value what “we” do, maybe “we” shouldn’t do anything. The same might be said if “we” are the schools that tend to be high on social mission and production of PC physicians. If, however, “we” are the American people, who need more primary care and rural physicians, or the politicians and policy makers who actually can have some influence, “we” need to make policies that reward schools that select the students whose demographic and attitudinal characteristics make them more likely to enter primary care, underserved, and rural practice. And those schools whose curriculum (formal and informal) and faculty attitudes and relative status-within-the-institution favors those same outcomes. And pay, on the back end, more money to those who do what “we” don’t have enough of (primary care, underserved and rural practice) than those who do what “we” have too much of already.

Whether or not “doing the same thing over and over again and expecting different results is the definition of insanity”, it is sure not going to accomplish any change in the physician workforce.
.