A recent paper** at AEA highlights the role of financial incentives in clinical decision making.
Medicare’s prospective payment system for long-term acute-care hospitals (LTCHs) provides modest reimbursements at the beginning of a patient’s stay before jumping discontinuously to a large lump-sum payment after a prespecified number of days. We show that LTCHs respond to the financial incentives of this system by disproportionately discharging patients after they cross the large-payment threshold. We find this occurs more often at for-profit facilities, facilities acquired by leading LTCH chains, and facilities colocated with other hospitals….
the average LTCH keeps patients about a week longer than they would if reimbursements were not tied to their lengths of stay.
To translate this out of academic economicese, Medicare pays long term acute care hospitals a little bit at the start of a course of treatment. After a while, there is a huge lump sum payment that is supposed to cover the entire bundle of services is triggered if a patient is still in the LTCH. A patient who is in the LTCH the day before the trigger date has only been a source of costs and low revenue for the LTCH. A patient who is in the LTCH the day after the lump sum trigger is a source of costs and big revenue for the LTCH. A patient who is in the LTCH for six months after the trigger date is a money losing patient.
LTCH hospitals want to push as many patients as they credibly can to the trigger date that releases a whole lot of revenue. Once revenue is triggered, the incentive is then to get the patients out the door as quickly as possible to minimize new costs.
If we were to assume that LTCH are run without profit motive and with pure altruism, we would expect to see a discharge pattern that does not spike immediately after the revenue release day. If we were to assume that economic incentives matter, we could expect at least some patient stays to be stretched to the trigger point.
This study sees the second story. LTCH hospitals respond to clear economic incentives.
Any payment structure that has controllable attributes that lead to a significant change in revenue or costs will see gaming. People, including doctors, respond to financial incentives. The response may vary between individuals and industries, but step functions create strong incentives where the decision to keep someone an extra day or two or to hurry up their discharge by the same day or two will occur.
As we move towards even more complex payment structures, we need to think through the incentives that the payment system creates. And then we need to look at the trade-offs and decide whether or not those trade-offs are good for the patient and society or if they are merely means of extracting rent.
**
Eliason, Paul J., Paul L. E. Grieco, Ryan C. McDevitt, and James W. Roberts. 2018. “Strategic Patient Discharge: The Case of Long-Term Care Hospitals.” American Economic Review, 108 (11): 3232-65. DOI: 10.1257/aer.20170092
SP123
Why is there also a weekly pattern to discharges, both before and after the payday? Is it some mentality that patients should stay exactly X weeks? Are LTCH patients usually admitted and discharged more often on a certain day of the week?
Ella in New Mexico
The way LTACH’s were originally set up–as “post ICU rehabilitation” in which they promise Medicare that by being selective, one more month of ICU level care with some additional focused treatments respiratory, physical, occupational, speech) will result in lower costs and better outcomes–is clearly a great idea on paper but not really possible in real life.
I’ve seen it in my 10 years in ICU, believe me. But we DESPERATELY need these facilities–we actually need more of them–in the healthcare algorithm of hospital to home transition.
Just like nursing home care: We don’t want to pay these facilities to do the costly things that make for good, quality healthcare. So we blame them for their failures and for their attempts to get around arcane funding loopholes instead of making good outcomes the basis for payment.
p.a.
Kind of off topic, but I wonder how issues like this, which are obvious to even a high school economics student, are not the result of incompetence, but of regulatory capture. The benefitting parties were not just part of the rule/lawmaking process, but dominated it, whether openly or sub-rosa.
Wag
One of my favorite axioms— “Tell me how you will measure (reward) me, and I will tell you how I will perform,”
dr. bloor
@Ella in New Mexico: Yep. That big payday in the middle of the graph isn’t just turning a profit on the folks on the left side of the spike, it’s covering the losses incurred tending to the folks on the right side of Day Zero.
Barbara
@Ella in New Mexico: Outcomes based payment switches the incentives around and still leaves hospitals with the ability to arbitrage payment systems, mostly by prioritizing patients with more manageable conditions.
daveNYC
Discovering that if you pay people a wad of cash if they do X makes them do X is not exactly groundbreaking research. Looking at the abstract it doesn’t seem like they attempted to calculate how many healthy people are being kept for longer stays than they need in order to get that payment or how often people kept past the payday are sent home a day or two early in order to save money.
Ohio Mom
Off the immediate topic but when I saw this on Boing Boing, I thought of yesterday’s discussion: “Leaks reveal the health care industry’s playbook for smearing and spinning Medicare for All out of existence by 2020.”
Don’t know how to do a link but it’s eady enough to get to boingboing.net and scroll down.
Anyway, obviously Medicare for All is a powerful rallying cry that is scaring the insurers.
WereBear
@Ohio Mom: Thanks, And it should scare the insurers.
Eric
@daveNYC: Figure 3 in the PDF seems to imply that there are many patients that are staying longer than needed. For the 2002 data, there is no spike at any timepoint. For the 2004 data, there is an apparent spike in length of stays after the 2004 threshold. This spike is even larger in 2013 (at the new 2013 LOS threshold) and a dip in the number of patients discharged right before the threshold.
Figure 4 then shows that most of these discharges are to home or skilled nursing facilities (not to acute care hospitals or death).
I wouldn’t be surprised if the paper did show how many patients are being kept for longer stays than they need . . . I just don’t have time to read the 34 page report. Section 5B might do exactly this.
cervantes
Solution: Put them on a global budget corresponding to the demand in the catchment area. That’s what most reasonable countries do.
Ken
Unrelated issue. Just noticed Court Listener shows a new document was added this morning to Texas v US case, but the document is available only thru Pacer, which requires ID and is limited to official usage. Anyone know if the new document is the long awaited decision on PEC protections?