• Chart of the day: The hospital productivity problem

    The ACA calls for annual reductions in Medicare payments to hospitals to the tune of (an expected) 1.1% below what they would otherwise be (PDF). This (expected) number is the law’s so-called “productivity adjustment,” and is the rate at which private, non-farm multifactor productivity* is expected to increase. Essentially, Medicare is telling hospitals that they had better increase their productivity at least at the rate that the rest of private industry does (apart from farms) or else payments will fall behind. This, along with other assumptions, gives rise to a frightening chart (see Figure 1, here).

    The question is, can hospitals increase productivity at this rate? For good reason, as far as I can tell, the answer is “no.” For example, here’s a comparison of percent changes in economywide multifactor productivity to two different methods of computing changes in hospital multifactor productivity from 1990 through 2005. (For the purposes of this chart, “economywide” means “private, non-farm,” per the chart’s footnote.)

    economy-hosp productivity

    The paper (PDF) that is the source of this chart discusses the challenges of estimating hospital productivity. I suppose one could argue that both ways it does so cause computation of changes to be biased downwards, but I don’t know how that argument would go. So, in the absence of such an argument, it looks pretty likely that hospitals have never been able to increase productivity at the rate of the rest of the economy. That’s not surprising to me.

    With that as a starting point, for what reason do we think hospitals will suddenly be able to increase productivity at least at the rate of the rest of the economy under the ACA? I guess the typical answer is to point to ACOs, bundled payments, pay-for-performance, etc., but I don’t buy that they will close the gap. I would buy that they might narrow the gap.

    Moreover, is it even fair to single out one industry and demand that it begin to track the average productivity increase of the rest of the other industries? After all, productivity growth certainly must vary across industries. It’s not unreasonable that some would be systematically low over long periods of time, given, say, their structure or the nature of technological developments. (See Baumol’s cost disease. I blogged through his book here.)

    Can hospitals really grow productivity as quickly as all other (non-farm) private industry? How? What’s the best answer to this hospital productivity problem?

    * “Multifactor productivity” means the “the change in a level of outputs relative to the change in a level of two or more inputs.”


    • I think you are confusing, a bit, the rate of change in productivity with the absolute level of productivity. Your chart shows, historically, that the rate of change of productivity in health care has been lower than for the economy as a whole. But why is that? Maybe there’s something about healthcare that makes it intrinsically harder to improve productivity. Or maybe its that we have traditionally had the incentives all wrong, and healthcare providers just haven’t invested in productivity improvements.

      From all of the examples of waste and inappropriate treatment, it sure doesn’t seem farfetched to me that the healthcare industry could ramp up their productivity improvement quite a lot. In fact, I would think that they could do quite a lot better than the economy average, at least initially.

      I know this is hard. We can’t even get providers to wash their hands consistently. But maybe the financial incentives in the ACA can help.

    • Has the medical-industrial-complex entered into the level of outsourcing that the rest of Western world business has gone through in order to “improve productivity”? From the obvious – bookkeeping, billings, collections, insurance claims/disputes, lawsuit management, etc. – to the less obvious – reading all X-Rays, CT, MRI, Ultrasound, PET scans, etc. How enthusiastically would radiologists compete on price with their Bangalore counterparts?

      • Don–The outsourcing of radiology has been repeatedly suggested and repeatedly shot down. There are legal and regulatory barriers (i.e., the doc in India has to be board certified in America which is extremely difficult without training here) and more importantly than radiologist (whom few like), the plaintiff’s attorney’s lobby is very, very against this (no ability to sue an Indian physician).

    • I agree with your underlying thought adding that the incentives presently being used IMO cannot increase productivity. Have we seen an increase in productivity in education? I don’t believe ACO’s will help based upon our experience with HMO’s that have virtually the same incentives.

      In fact I find it more likely that ACO’s will increase costs and do little to affect productivity except possibly in the most superficial way. Take note of how Medicare (single payer of those over 65) pays for colonoscopies. I believe about twice as much to hospitals as to OP clinics while at the same time the government is trying to stop OP surgical clinics from being built. Why not go with the less expensive price which in many cases is more convenient for the patient.

      By the way have you ever thought about how productive we are in the telecom industry? Telecom companies can buy smart phones for about $20 in China. They can offer the phones for free with service and can even guarantee the phones at very low or no cost since their cost is ~$20. When the phone breaks they don’t have to fix it rather they can simply replace the phone. When we can do that with humans we might see the same type of productivity increases we see in industry.

      If we really want to decrease costs then we have to consider giving up on old ideas and go from a fixed phones to cellular. Just substitute the word phone with the word hospital and cellular with the word outpatient. Right now in certain types of care we are moving in the opposite direction. We might have to give up on the idea that the government, physicians, hospitals, insurers should determine how health care is provided to the individual patient and get the patient involved. Additionally, we might even have to change our ideas regarding licensure. We certainly have to stop politicizing medical care.

      • We consistently see that when docs have financial ownership in OP facilities, utilization increases. That convenience leads to increased spending. Also, the patients getting inpatient colonoscopies tend to be much sicker. My colonoscopies today were all very sick folks, including the one i cancelled to send off to the cath lab. The ten I did last week at the surgicenter were all pretty healthy folks. Faster, easier, needed no real attention after.


        • Steve–while self-referral and utilization are real problems, the major explanatory figure for the increased costs of the U.S. system seems to be prices (see Ezra Klein’s blog and this one for a much more detailed examination of that claim).

          If prices are the most significant problem, then encouraging outpatient clinics/more care settings in order to raise competition and pressure larger providers seems to be the best (even if or especially because docs will hold ownership stakes).

        • Steve, whether sick or well those colonoscopies done in the hospital seem to cost a lot more, probably around double. As you well know there are guidelines as to when Medicare will pay for a colonoscopy. Are you suggesting that those that fit within the criteria for colonoscopy or those that have symptoms should not have colonoscopy because the numbers are rising? As you know better than I that procedure is supposed to detect cancers of the colon earlier and thus saves lives.

          You seem concerned that the ease and other things that OP clinics provide might cause utilization to rise when to date many people have not had their first colonoscopy that is recommended. Maybe utilization should rise and we would have less untreatable cancers of the colon. Is there proof of significant abuse of the system when colonoscopies are done in surgicenters compared to hospitals? Hospital monopolies of the past caused quite high pricing that relaxed when outpatient centers started competing. Think radiology and laboratory and present day colonoscopy.

          Maybe you are paid too much for your service or maybe we just can’t afford to pay for asymptotic colonoscopies. Look at how a relatively recent change in a guideline made one previously relatively poorly paid medical subspecialty into one of the highest paid. I am sure your subspecialty medical association pushed for the preventive care (early diagnosis) labeling and payment by Medicare. I am not saying you don’t deserve the money you earn rather that we have to start thinking more about how things will be paid for in the future and recognize that technology has changed medicine and thus the medical system needs to change or adapt. The top down government approach hasn’t permitted that change to occur and the ACA hasn’t either. It is basically more of the same.

    • Per Jeff Goldsmith and others, many hospitals have been on a construction binge and an administrative-high tech hiring binge for over a decade.

      If they just pared back their facilities and staffing to the levels of 2000, then measured productivity would soar. Same is true for many types of expensive equipment that are duplicated (or ‘triplicated’) in the community.

      Hospitals have survived largely by upcoding and learning to manipulate highly graded fee schedules. This is true especially in ER’s but elsewhere too.

      Hospitals are a ‘target-rich’ environment for anyone who is determined to lower costs, and willing to risk the wrath of highly paid hospital employees.

      • Hmm. Assuming that spending binge really was all waste, these all sound like one-off cuts to me. Measured MFP would grow quickly for a while, but eventually hospitals will run out of things to cut and then we will really see the lower payments start to bite.

        Besides, while trimming waste is noble enough, we should remember that these are one-off static gains that would be insignificant compared to a permanent increase in the MFP growth rate by even a few tenths of a percent.

      • Bob, (1) For hospitals to build duplicate equipment means there is a market for that equipment. There are ways to control that type of decision making and it requires less pressure from government rather than more.

        (2) Many hospital systems do not rely solely upon billing. They rely upon large donors or they would be in the red.

        (3) Acute care hospitals require many expensive things at their fingertips. When a person is dying one cannot wait for the bureaucracy to act. That happened in a well known case in Canada, Natasha Richardson death following a ski accident. (I am not attacking the Canadian system, just using this as a well known example.) Bureaucratic methodology killed her because the bureaucracy acted so slowly. There is little question in my mind that she should have lived in a country like Canada that has excellent physicians.

        Can you imagine having to transfer a critically ill patient to another hospital over and over again?

        (4) Hospital safety is based upon duplication.

        Let me just add, for a long time almost everything that didn’t require a hospital could be performed outside the hospital at much lower rate. Maybe we need to rethink our ideas about hospitals. They are like shopping malls where individual vendors ply their trades.

    • -This reminds me of the comments from Sen. Warren, in which she provided her estimate of what the current minimum wage should be based by summing the aggregate productivity growth from the 1960’s to the present, then multiplying the real minimum wage in the 1960’s by that number. Her calculations indicated a minimum wage something like ~$22 per hour.

      This was a truly fascinating chain of reasoning, since it presumed many things that are unlikely to be true. Foremost amongst these is the presumption that productivity growth has been uniformly distributed across all industries from the 1960’s onwards. Unfortunately for Sen Warren and all others who made this argument, productivity growth has been anything but uniform across all sectors. In fact – it has been concentrated amongst relatively few sectors in where the returns to capital investment in mechanization/automation have been high. This category includes relatively few sectors with significant minimum wage employment. Productivity gains in semi-conductor fabrication have been astronomical. Productivity gains in folding and displaying clothing at clothing retailers have been much more limited. Forcing clothing retailers to pay their employees as their productivity has increased as much as the employees involved in semi-conductor fabrication will quickly bring about the end of the business of selling clothing in retail settings as we know it.

      The same fallacy underlies the ACA policy cited above. Disaggregate all of the sectors and you will see dramatic gains in some sectors, few to none in the others. The gains will be concentrated in sectors where the returns on capital investment are highest, and lowest where the returns on capital spending aren’t sufficient to cover the cost of the initial investment – like, say, robots that can fold, stack, and display knit sweaters twice as fast as your average teenager.

      The scope for mechanization and automation in caring for the sick, the wounded, and the disabled is limited,and it’s insane to expect – much less mandate – productivity gains in this sector that equal or exceed average productivity growth figures for the economy as a whole. This would be true even if we could actually measure medical/hospital productivity well enough to reduce it to a reliable number, but we’re a long ways away from that.

      This piece of the ACA is based on a fallacy, wrapped in a cluster&*^%, inside a non-sequiter. Won’t end well.