Health Policy

So what can we do about health care costs?


Last week Jeff Goldsmith wrote a great article in part explaining why health care costs in the US went up so much between 1965 and 2010. He also pointed out that health care has been the same portion of GDP for more than a decade (although we haven’t had a major recession in that time other than the Covid 2020 blip when it went up to 19%). However, it’s worth remembering that we are spending 17.3% of GDP while the other main OECD countries are spending 11-12%. Now it’s true that the US has lots of social problems that show up in heath spending and also that those other countries probably spend more on social services, but it’s also clear that we don’t actually deliver a lot more in services. In fact probably the most famous health economics paper of the last 50 years was Anderson & Rienhardt’s “It’s the Prices, Stupid”, which shows we just pay more for the same things. Anyone who’s looked at the price of Ozempic in the US versus in Denmark knows that’s true.

But suspend disbelief and say we actually wanted to do something about health care costs, what would we do?

There are 4 ways to cut health care costs

  1. Cut prices
  2. Cut overall use of services
  3. Reduce only unnecessary services
  4. Replace higher priced services with lower priced ones

Number 3 or reducing only unnecessary services is the health policy wonks dream.

The Dartmouth school, originating with Jack Wennberg, has done a pretty good job convincing the health policy establishment that there is massive practice variation across the nation (and even within cities and individual hospitals), and that while this leads to higher costs, it doesn’t result in better outcomes. In fact outcomes where there are more services and spending tend to be worse. Dartmouth does have its critics like Buzz Cooper, and maybe all the explanation of variables in health care spending is caused by well meaning doctors ministering to the inner city poor, but it’s not hard to find overuse bordering on fraud. There have been a ton of well meaning attempts to both educate patients to choose wisely and to get doctors to behave better (or at least report their data), but there’s a new report out showing that Dartmouth had it roughly right every day. (This recent NYTimes one is about cutting babies’ tongues to make them breastfeed more easily).

Overall there have been some reductions in some measures, like hospital admissions but many of those have been replaced with other services, and in general practice variation has not gone away. Could it happen? Maybe, but 50 years of evidence makes it look unlikely. Don’t forget that the Obamacare authors were faithful disciples of Dartmouth but not much of that philosophy ended up in CMS policy.

Number 4 or replacing higher priced services with lower priced ones is the Silicon Valley health tech dream cross-bred with the Dartmouth school’s love of primary care. I will admit to being a fan of this movement. If we can replace higher priced people (doctors) with lower priced people or non-people (AI) we should be able to deliver the same things we are doing today at a lower cost. For example, in the field of psychotherapy there’s currently a great shortage of therapists. One thing that’s being done is replacing therapists with lower qualified coaches. But the end game is to use AI-powered chatbots and avatars to do the same thing. 

A related attempt is to deliver preventative services using technology. This is now paid for by Medicare – it’s called remote physiological monitoring (RPM). While its introduction has been a tad bumpy, it intuitively makes sense. If you can start tracking the care of relatively sick people while they are at home and relatively healthy, surely you can pick up issues before they get worse, intervene with medication changes and other services in their homes, and therefore prevent hospital admissions and improve outcomes. In fact, given how cheap tracking technology is, and the advances in AI, can’t you monitor everyone (based on their level of acuity) and give them a personal AI health coach? I call this the “continuous clinic” and it’s a great idea if I say so myself. The problem is that it’s not going to happen easily in a medical world that manages its process in terms of office visits and hospital admissions and gets paid on those metrics. We simply don’t have the right type of new organizations to put this together. And if you believe John Glaser and Sara Vaezy’s recent piece in the HBR called Why the Tech Industry Won’t Disrupt Health Care, we’re unlikely to get them. (I think John & Sara hope that the incumbents will reform themselves, but they would say that, wouldn’t they!)

Which leaves us with 1, cutting prices, and 2, reducing overall use of services. 1 & 2 are what the rest of the OECD does. 

Virtually every country in the OECD has some form of central price controls. Even if they have multiple paying entities, like Germany, there’s one agreed price schedule. Or, as in the UK and Scandinavia, there’s a regional or national budget. The US also has such a national price control, but only for some people over 65, given that Medicare Advantage now covers half of that population, and only for some services. Notably it doesn’t cover drugs, although that will slightly change in the near future given CMS’ new ability to negotiate the prices of some drugs. 

To this point in the US, any attempt to squeeze down on Medicare prices produces two effects. One is violent disagreement on behalf of provider organizations, which spend more money lobbying than basically any other industry in America. Almost always this means that Congress balks at imposing any real cuts. The other is that providers find ways to transfer those costs onto patients unable to negotiate. You’d think that the patients’ representatives (insurers and employers) would resist that but RAND has shown that they are basically price takers, paying more than double what Medicare pays for the same thing. Again this could change, and there’s some recent legislative activity that has a few people very excited, and has spurred some lawsuits about fiduciary responsibility – ironically one from an employee of a drug company. But we remain a long long way from a German/Japanese/French style price schedule.

Which leave us with 2, reducing overall use of services. The name for this in US health political  (if not policy) circles begins with another R, rationing. The stories of Canadians flooding across the border to access American health care were always basically bullshit, but like today’s stories of critical race theory, transgender drag queens corrupting our youth, and millions of migrants invading the southern border, it doesn’t take much to wind up the Fox News crowd as the Democrats found out. In 2009 the very wonky issue of when women should get mammograms became death panels very quickly. (BTW if you want to read a lot more about Canada, here’s a classic THCB piece I wrote in 2003. Not that much has changed)

This all means that obviously and transparently reducing services, presumably by creating a UK style cost-benefit analysis commission, is unlikely to happen. We have tried outsourcing that to the private sector, particularly in Medicare Advantage. But the combination of naked greed and stupidity from the MA plans and the use of scary AI, will probably put paid to that soon enough now the trial attorneys have got hold of it.

So to summarize, we pay about double what most other countries pay in $$ terms and about 50% more as a share of our (much bigger) GDP. And of course we lead the league (still) in the number of uninsured people and those who are practically uninsured, or facing bankruptcy from medical bills. There are four ways we could fix it, but none of them seem that promising.

And I don’t see a way this changes any time soon.

Matthew Holt is the publisher of The Health Care Blog