Health Affairs published a new study that examines the value proposition of Academic Medical Centers (AMCs)***. AMC’s tend to be high cost providers for both the very unusual and rare care that they are the unique providers of in a region and common care. I’ve been skeptical about the value proposition of AMCs for routine care for a long while in the context of ACA network designs:
Ideally, the insurance companies that want to mimize their claims pay-outs want to have Seattle Children’s or any other high cost specialty hospital in network for a la carte services such as organ transplant and regional trauma centers of excellence, but out of network for pneumonia or routine elective surgery or setting broken ankles….
I was wrong. I am surprised by these results and I now need to significantly move my priors.
We examined more than 11.8 million hospitalizations in the period 2012–14 for Medicare beneficiaries ages sixty-five and older and found that, after adjustment for patient and hospital characteristics, high-severity patients had 7 percent lower odds, medium-severity patients had 13 percent lower odds, and low-severity patients had 17 percent lower odds of thirty-day mortality when treated at an academic medical center for common medical conditions, compared to similar patients treated at a nonteaching hospital. For surgical procedures, high-severity patients had 17 percent lower odds of mortality, medium-severity patients had 10 percent lower odds, and there was no difference for low-severity patients…
When all of these results are taken together, we found that the better outcomes at AMCs appear to apply to all patients, not only the sickest ones with the most complicated conditions.
Those are all very meaningful differences in mortality for the Medicare population.
Now onto the policy question — we have been moving towards narrow networks as a means of cost control. This is quite common in the ACA Exchange markets. AMCs are expensive in both the per unit cost and as a magnet for people with significant medical histories that require high cost care. Insurers that can maintain network adequacy from both a regulatory and marketing perspectives without including an AMC will be able to maintain a pricing advantage over insurers that include an AMC in their network (all else being equal) and thus they will be able to cherry pick the healthiest and cheapest portion of the population with attendant risk adjustment outflows.
The current risk adjustment formula is based on the average state wide premium so if an insurer with no AMCs in network can claim most of the healthy, low morbidity enrollees while driving down the state wide average premium, the risk adjustment transfers to the high cost, AMC containing network insurer would be insufficient to pay for the high cost care that is best or only done at an AMC. This is not an uncommon play in the ACA market. It either drives out the broad networks, or raises rates significantly on the sickest people.
Yet, this study, and others indicate that AMCs provide significant value over the alternatives. We should want to pay more for more health which includes low mortality odds.
I am not sure what this means for policy, but it is a study that should make us re-examine risk adjustment and narrow networks at the very least.
*** Burke, L., Khullar, D., Orav, E. J., Zheng, J., Frakt, A., & Jha, A. K. (2018). Do Academic Medical Centers Disproportionately Benefit The Sickest Patients? Health Affairs, 37(6), 864-872. doi:10.1377/hlthaff.2017.1250
Amir Khalid
Do teaching hospitals at public universities in America get any sort of subsidy from the state that reduces medical fees?
Another Scott
Interesting, but not really surprising. Teaching hospitals should be the best at their craft (they teach, after all). But there is the complication, as you say, of attracting the sickest and most difficult patients.
I haven’t read the paper, but I would be interested in knowing why the mortality is lower. Better techniques? Fewer mistakes? Newer hardware and techniques? Lower infections? More sensible work hours for staff? Less soul-crushing bureaucracy? More automation of paperwork? Adequate staffing levels? Not enough really good physicians to go around to staff lower-prestige hospitals?
Thanks.
Cheers,
Scott.
BradF
David
“All this evidence is consistent with the view that residents bear the cost of their own training, which would mean that GME funds are treated as general monies going to their institutions; in fact, these funds are often used in ways that are difficult to trace, assess, and justify.”
Even within the AMCs and their better outcomes, at least by this study, there is no free lunch. Give the non-AMCs the same low-cost labor or transfer the value of that labor to the same facilities, and the leg up advantage may evaporate.
Brad
MomSense
@Another Scott:
With the usual caveats about anecdata, in my experience the staff just have had more experience. At my local hospital the surgeon has done procedure A five times in her/his career. At the teaching hospital the surgeon did that procedure five times last week.
I had a partial knee replacement the same week as my neighbor some years ago. I went to a fancy teaching hospital in Boston and he went to the local hospital. His failed in three months and I’m still going strong. And my neighbors transition to a full replacement didn’t go well because the surgeon had compromised so much bone in the partial replacement surgery. This is a known risk that makes it worth going to the fancy teaching hospital.
Feathers
Another factor is that a patient usually has to do some maneuvering to get themselves into a teaching hospital. So you are likely to get more knowledgeable and motivated patients, especially for the procedures that don’t have to be done st an AMC.
Personally, I saw the very real difference in mental health treatment when a co-worker in crisis went to the place her insurance chose, versus my doctor who deliberately chose a teaching hospital program for me, and kept me going while I waited for a spot to come open. Harder to do comparisons here, but the differences in treatments and outcomes were huge.
RepubAnon
That’s been my experience – the teaching hospitals cost more per procedure, but because they have additional expertise, overall costs are lower due to fewer unnecessary procedures.
FlyingToaster
We just got our BCBS enrollment packet (from HerrDoktor’s employer) for next year. Same broad network, but with a tack-on copay of $250 for 8 hospitals. Now this is Boston,where most hospitals are teaching hospitals, but the 8 most expensive (MGH, Brighams, Dana-Farber, Childrens, etc.) have the copay slapped on, where hospitals in the same network (Newton-Wellesley, NorthShore) don’t. Nor does BethIsraelDeaconess, which is in the same complex as 3 of the surcharge hospitals.
Our local hospital is a Harvard facility, but doesn’t have a co-pay, and neither does Tufts/NE Med. We’re only affected if WarriorGirl has to go to the emergency room from summer camp (both camps send to MGH).
I found it odd that they seemed to cherry-pick which hospitals to require a co-pay for.
gvg
this is good for me personally as I work for a University with the teaching hospital. My insurance always includes it. I think most insurance around here does as it’s pretty much the driver of the local economy and a large percentage of the population works for either the University or the hospital.
the other local hospital has a good reputation too and better for some things and I can go to them also. We aren’t a typical small town though. that’s one of the problems with figuring out what’s best. Places aren’t alike. Peoples life experiences drive their votes but other areas have different mixes.
Barbara
@Amir Khalid: They do get a subsidy but it’s not enough. The funding of teaching hospitals takes the form of training subsidies for physicians. There is a whole world of perverse incentives and counterproductive consequences of training too many specialists and not enough primary care physicians, alongside the fact that some hospitals (in NYC, for instance) try to load up physicians in training with the kind of work that can be done by nurses and others who, perversely, are better paid than physicians “in training” in high cost jurisdictions. This world is so byzantine that it defies the understanding even of people like me. But the bottom line is that for all they receive in subsidies, AMCs tend to be much more expensive than their non-university counterparts. Some of that probably relates to market position, but some of it relates to real structural cost differences.
Ruckus
@Another Scott:
I may have some of the answer. But I have to go to work. Will check in later and leave my thoughts on this.
David
As a physician who works and trained at an AMC, not surprising. I’m constantly amazed at what passes for care at the community hospitals that transfer patients to us.
prostratedragon
If a patient has a condition that requires long-term management of some kind, I think the teaching hospitals more likely to have staff that know what protocols have been successful, and the systems to help the patient implement them. I’m thinking of some common ailments like heart failure that are not really curable, but for which years of additional life at some quality can be had.
Cheryl from Maryland
Not surprising at all — both for experience and for the ability to think outside the box. My husband’s litany of health issues include some zebras (neural sarcoidosis). After wasting 5 years to get a proper diagnosis, which we got at a teaching hospital, we don’t think about elsewhere. Fortunately, we live in the Baltimore/DC area.
Barbara
@David: In fairness, the entire model of care at most non-academic medical centers, i.e., local hospitals, is based on a bunch of independent actors mostly operating without regard to each other, and using the hospital as a central hub for their own patients. Interdisciplinary coordination is hard to arrange. My anecdata is of my husband’s elderly aunt finally going to VMC Medical Center after a prolonged spell of not feeling well, and finding that of the 11 medications she had been prescribed by various physicians who were not inter-conscious, she only really needed four. Some community hospitals might as well be academic medical centers, and many hospitals are moving to a model that looks more like an AMC, with employed physicians. But modern hospitals and the “business” of medicine in general are organized for the benefit and convenience of physicians, not patients.
PST
@MomSense:
I believe that this gets at the essence of the difference, rather than some inherent superiority of the AMC model, and is one reason community hospitals sometimes have excellent programs in a particular specialty. To do so, they need to have a sufficient volume to sustain a small group of well qualified specialists. Notwithstanding MomSense’s friend’s bad experience, that specialty is often joint replacement. Lots of communities have aging populations and no nearby AMC vacuuming up all the available hip patients.
I wonder how the distinction between AMCs and community hospitals will be affected as the former increasingly buy up the latter. At first, these were thought of a feeder hospitals who would send their serious cases to the AMC. But these days, there is a trend toward moving particular specialties to the acquired hospital and staffing it with doctors on the faculty. For example, University Hospital will decide that all its behavioral health inpatient care will now be at Community General and all its ophthalmology will be at Regional, both former community hospitals that retain that character in other respects. In a sense, AMCs are getting spread out over a collections of formerly independent hospitals across a region.
jl
Thanks for interesting link. More anecdata, but the difference between AMC and community hospital is obvious to me from my experiences with care family and friends have received at the two different kinds of institutions. Several community hospitals were severely understaffed, poorly organized and quite chaotic in how they functioned.
One specific case was an oldster that was having short but intense seizures at irregular intervals. It made quite a difference in care when an intern could find the time to stay in the room and watch for one to occur and simply report what happened to an attending (local teaching hospital), versus me having to wait for a seizure to begin (which as I layperson, I had a hard time recognizing, the onset was subtle), run out into the hallway and flag down a nurse who was running (slight exaggerating, actually a ‘high trot’ is a better description) from room to room to tend to the patients. So, very anecdata, but I have to wonder if a variable like non-profit versus for-profit, or how many times the ownership of community hospital changed hands in last five years might explain some of the difference.
jl
Regarding specialist MDs, I don’t think the problem is the number, but the cost. the US really stands out in paying its specialist MDs far more in real terms than any other country. From what I’ve read, specialists in US also have far more control over procedural coding and physician reimbursement system than other countries.
Ruckus
@Barbara:
My experience with both the VA and when I had a doc who also taught at OSU med is that the interaction between most of the docs and their students and the time they spend discussing the patient is invaluable, especially that most of the rest of medicine spends the absolutly least amount of time possible with any patient. That interaction leads to less instant decisions, most of which are probably correct but without any feedback only time tells. And that may be too late.
jl
Finally, to be nice and totally bash on community hospitals, it does seem that low severity surgeries can be done at community hospitals with no loss. Also, interesting that seems no difference in high and medium severity stroke and sepsis. Would be interesting to see why that would be. Maybe timeliness of first standard-of-care treatment after onset of symptoms is the most important factor there, and no systematic difference between AMC and community hospitals there.
Also, I missed that they did adjust for for-profit status in the analysis. And, did adjust for multiple tests (21!) with a very conservative method, so unlikely to have picked up chance associations, even for ratios between 0.90 and 1.00.
jl
Though, the authors were bad and naughty in one way, for anyone reading the paper. The test results for difference between AMC and community hospitals are reported in a misleading way in the tables and graphs, making it look like there are more statistically significant differences than the actual statistics show. So, need to disregard the charts and actually read the results section to get the bottom line that their numbers actually support.
L85NJGT
Unionized nurses, custodial and housekeeping are in-house and union, on premise translators and social workers, better quality administrators – there be turtles all the way down.
A lot of it is fundamental management focus and operating discipline; of course action item #1 by a larger system when they acquire a community hospital is to dump the administrative staff. My local hospital never recovered from its first M&A – I think it’s at #3 now and the AMC system is booting any non-affiliated practices from the attached MOBs just to return to some base level of accountability.
I suspect a high-low mix of fewer and bigger regional trauma centers being fed by stand alone ASCs, multi-practice MOBs, birthing centers, etc., would be a better model, but the community hospitals are tied at the hip with state and local politics, so here we are.