Menu
Categories
When Researchers Need to Spend Some Time Actually Studying the Things They are Researching
September 22, 2014 Economics Problems

In my inbox this morning is this working paper:

Should Hospitals Keep Their Patients Longer? The Role of
Inpatient and Outpatient Care in Reducing Readmissions
by Ann P. Bartel, Carri W. Chan, Song-Hee (Hailey) Kim  –  #20499 (HC HE PE)


Twenty percent of Medicare patients are readmitted to the hospital within 30 days of discharge, resulting in substantial costs to the U.S. government.  As part of the 2010 Affordable Care Act, the Hospital Readmissions Reduction Program financially penalizes
hospitals with higher than expected readmissions.  Utilizing data on the over 6.6 million Medicare patients treated between 2008 and 2011, we estimate the reductions in readmission and mortality rates of an inpatient intervention (keeping patients in the hospital for an extra day) versus providing outpatient interventions.  We find that for heart failure patients, the inpatient and outpatient interventions have practically identical impact on reducing readmissions.  For heart attack and pneumonia patients, keeping patients for one more day can potentially save 5 to 6 times as many lives over outpatient programs.  Moreover, we find that even if the outpatient programs were cost-free, incurring the additional costs of an extra day may be a more cost-effective option to save lives.  While some outpatient programs can be very effective at reducing hospital readmissions, we find that inpatient interventions can be just as, if not more, effective.

Now, I don’t know them, but most economists I am familiar with attend seminars, teach classes, etc. from the comforts of their own office or at some hotel. But when I see findings like, “if heart patients stayed in the hospital for one more day as compared to sending them home, outcomes would improve,” I get weak hearted. First this is bad economics. Unless these are dynamic general equilibrium models, there is no way to figure out if in fact these partial equilibrium results hold. Why? Have we asked what happens to the other patients when patients who were released are staying in the hospital one more day? Now, you might say, “they come back anyway, so this doesn’t matter,” but it does. When a patient is released, they are in less immediate danger than those coming out of surgery and other emergency situations. So keeping these folks for an extra day, even if medically necessary for them, may not be medically optimal “for all.” Second, and my family would have very good reason to appreciate this, is that it’s not just as simple as “keeping patients for another day” on various hospital flaws. We all talk about medicine as if there are an endless stream of doctors, mid-level providers,, nurses, care coordinators, cleaning staff, etc. But there are not. And a little birdie tells me that our recovery units and other critical care units are already swamped with patients, very likely understaffed by professionals at all levels, and generally pretty intense places to be. It’s not like there are lots of empty beds and equipment sitting there readily available for patients – there is a reason patients are sent home, the hospitals need the space. There are lots of important details I am not permitted to share, but needless to say (I hope) if but rest assured there are a host of other bad incentives running around our hospitals, and these researchers are now ignoring any and all of those things when concluding, “just keep folks in the hospital for another day” and wholly disregarding whether surgeries are necessary, whether the extra people can be cared for reasonably well, and more.

This problem is not simply the province of this paper, but a good number of new ones I come across.

 

"4" Comments
  1. >> “Moreover, we find that even if the outpatient programs were cost-free …”
    I assume that they mean the thing was priced at $0?

    And the premise seems to ignore the obvious reality of trade-offs. I’m sure that we could also improve outcomes if everyone had a personal physician who gave them a full check-up each week.

  2. I am also pretty sure that if a personal physician were checking on us every week we’d see a lot more unnecessary medicine administered too … my knee hurts at the moment … it must be cancer!

  3. Interesting. As much as economics has become a hobby for me, I have much to learn still. That said, my understanding is that hospitals are under pressure by payors (e.g., the government) to push patients out the door as fast as possible to save costs (hence the rise of outpatient programs noted–anything to save the price of a day in the hospital). We see this throughout the system where days post-procedure spent in ICU, for example, have decreased over the years. So, my first read was that maybe cutting down hospital time was a good idea taken too far, and that there was never any optimization with thought toward downstream effects (readmission, for example).

    My question to you is this: aren’t the scarce resources you outline already reflected in the price of a day in the hospital? Certainly keeping all patients an extra day even for a limited number of procedures would increase demand and thus price, but can’t that be factored in (although then maybe that would cause patients to be sent home earlier after other procedures if the prices were higher)? Why is the current number of days the correct number of days? Maybe this is bad economics, but at least they are doing some math and thinking about outcomes rather than simply trying to get patients home as fast as possible.

    In the end, we’re not a single-payor system, so optimizing “for all” doesn’t happen. From an individual patient’s perspective (and from the perspective of his/her payor), would I pay for another day now to reduce chance of readmission later? I guess it depends, and largely on the factors outlined in their paper, right?

    • I happen to think your questions are spot on – but given the way hospital bills are calculated we are nowhere near thinking about marginal cost pricing. And beyond that, I think there are real staffing problems (factor limits to use fancy terms) in nursing, particularly when there are uniform labor agreements requiring that a surgical ICU nurse, for example, be compensated the same way as a nurse on a step down unit. The staffing, in my estimation, is way off.

      As far as pushing patients out the door too fast, I can see that. But let me add a few things. First, if patients don’t aim to get out the door, medically, they seem to be less likely to leave. You’d be stunned at the number of patients post-recovery who lose the desire to do anything. Second, there are incentives to keep patients in house for a long time too. This includes measuring outcomes based on time of survival post-surgery and not on quality of life post surgery. Finally, it might be the case that doctors are performing too many surgeries right now, so perhaps asking people to stay in house for a longer period of time will make it harder for them to decide to cut.

Leave a Reply
*