Code Red: Two Economists Examine the U.S. Healthcare System

March 15, 2013

The PROMIS of Market Forces

Filed under: Uncategorized — David Dranove and Craig Garthwaite (from Oct 11, 2013) @ 10:01 am

On Monday, my health economics students were treated to a guest lecture by NU Professor David Cella, the director of the Patient Report Outcome Measurement Information System (PROMIS) project. The PROMIS project is the culmination of decades of research into measuring the health of patients through carefully worded questions about physical, social, and emotional well-being. Through an intelligent, interactive questionnaire, PROMIS can deliver accurate and reliable measures of the quality of life at minimal time and expense. (You can try out PROMIS at I believe that PROMIS will be a key element of an efficient, market-based, health economy.

For decades, markets and regulators alike have been stymied in their efforts to assure quality while controlling costs. Fee-for-service incentives increase costs without commensurate increases in quality. (This is Enthoven’s famous “flat of the curve argument.”) Capitation, prospective payment, even shared savings, can lead to too much cost cutting at the expense of quality. Then there is the matter of helping patients find the right providers for their specific needs and giving providers the right incentives to boost quality, whether that means that physicians keep up with the latest research, nurses receive extra training, or integrated systems make sure that patients are assigned to the most appropriate staff providers on staff. There is a broad consensus that a major obstacle has been our inability to adequately measure and reward quality.

The current thinking, as exemplified by ACOs, is to reward quality through a lengthy list of pay for performance measures. This may improve upon the status quo, but it can never work to anyone’s real satisfaction. Providers will respond to incentives as they always have, in ways that we will find wanting. As providers focus on rewarded behaviors, they will give short shrift to unrewarded behaviors – there is already a lot of evidence that this is occurring. Payers could respond by lengthening the list of rewarded behaviors, but this will never prevent multitasking unless payers remove from providers all discretion about how to deliver care; for all intents and purposes, payers will become the providers.

The only way out is to measure what really matters, outcomes, and let providers find the best way to improve them. But outcomes data in medical records is inadequate. Can the asthmatic patient climb a flight of stairs? Can the hip replacement patient walk without pain? Is the cancer patient depressed? Medical records won’t answer these questions. PROMIS can.
Imagine being able to measure, with considerable precision on a standardized scale, the mobility of an asthmatic or hip replacement patient, the mental health status of a cancer patient, or the fatigue of a morbidly obese patient. And then imagine being able to convert these measures into a single score analogous to a Quality Adjusted Life Year. It is not difficult to use these numbers to answer questions of fundamental importance: Which orthopedic surgeon’s patients see the biggest improvement in mobility? Which providers and which drugs are best at combating clinical depression? Which endocrinologists and dieticians are doing the best job helping their obese patients to cope? We might even ask, in fact, we should ask, bigger and bolder questions. Which ACOs generate the most QALY gains for the patients? Which health plans are best for which types of patients? I believe that PROMIS scores work best at this high level, where it is not necessary to assign credit to any particular provider.

If we can answer these bold questions at the highest level of care, ACO or health plan, then the market will take care of itself. ACOs and health plans will be accountable for cost and quality and will have limited ability to game the performance measures. With this accountability, ACOs and plans must find ways to balance cost and quality throughout the vertical chain or they will fail to compete in the market. (This assumes, of course, that there are competing ACOs and plans.)

ACOs and plans will find it difficult to assign credit to individual providers using PROMIS. They may turn to P4P measures, but these will still invite gaming. How does the ACO promote quality? How can the health plan improve its PROMIS ranking? The answer is really very simple.

It is well known in the economics of organization that the best way to hold individuals accountable for quality is through a combination of hard and soft incentives. ACO managers who are responsible for the quality of their organizations, and health plan managers who are responsible for the quality of their networks, should feel free to use P4P process measures. But they should accompany them with a qualitative review. If managers suspect that their medical staff is gaming the P4P rules, they should act on those suspicions. Ask the medical staff what they think of each other’s quality and act on this as well. Health plans should drop providers from networks, or reduce their rates, based on quantitative and qualitative measures. If managers act consistently and in good faith, doctors and other providers will learn to trust them and strive to be their best. This is how it works in most professional services firms, where rewards are usually based on a combination of hard and soft incentives. Healthcare organizations must move towards the same model or forever be stuck in a world of command and control, and gaming and inefficiency.

The problem with healthcare management these days is that there is too little management. Managers address problems using hard-edged incentives that have never worked in isolation and never will. PROMIS gives us an opportunity to hold health care organizations accountable for their overall performance. But the managers of these organizations must trust themselves to manage, to use their own judgment, and not be automatons. When this happens, we will at long last see market forces fulfill their promise.


  1. David
    Assessing PROMIS in 20% of population generating 80% of costs–with attendant cognitive, language, habilitative limitations, makes instrument limited answer.

    I am not saying these tools dont serve a purpose–of course they do–but they are not one size fits all.

    I could not link to review, but how does PROMIS differ from SF-36 btw?


    Comment by Brad F — March 15, 2013 @ 11:57 am

    • PROMIS can be administered to the entire population and therefore used to track wellness as well as illness, though it does not distinguish between wellness and, shall we say, extreme wellness (e.g., athletic performance.)

      Think of PROMIS as a sleeker and much easier to use version of SF-36. It is a simple yet accurate way to recover health utility states. Several insurers as well as Medicare are taking their first steps towards incorporating PROMIS and I believe that entrepreneurs with imagination will find ways to use PROMIS to dramatically transform how we evaluate our providers and payers.

      The link is

      Comment by dranove — March 15, 2013 @ 12:53 pm

  2. Elderly folks, think duals (I care from them frequently), would have a heck of time with some of these questions. I emphasized the 80/20 aspects because we want to shine a light on the costliest beneficiaries–treatments for these folks have the greatest impact on system resources and costs, as you know.

    All in all, harder than you think. Instruments for surrogates need validation as well. But then again, maybe when you refer to entrepreneurial spirit, you imply those kinds of fixes.

    Thanks for engaging

    Comment by Brad F — March 15, 2013 @ 6:31 pm

    • Points taken. And I can add that PROMIS affords new opportunities for gaming. (I think about how my car dealer always implores me to give it perfect scores on the survey of customer satisfaction.) PROMIS is one of several hard measures that may be used, but ultimately I think it is the soft measurement by good managers that will drive quality.

      Comment by dranove — March 15, 2013 @ 6:55 pm

RSS feed for comments on this post.

Blog at

%d bloggers like this: