Monday, March 21, 2011

Maryly proceeding off course

Maryland, the only state with a hospital rate-setting process, also has an interesting financial incentive program related to quality indicators. It is described here. Some excerpts:

This initiative, which commenced July 1, 2009, links payments to hospital performance on a set of 49 Maryland Hospital Acquired Conditions (MHAC) across all-payers and patients in the State.

During fiscal year 2008, these hospital-based preventable complications were present in approximately 53,000 of the State’s total 800,000 inpatient cases and represented approximately $500 million in potentially preventable hospital payments.

The MHAC methodology provides a system of payment incentives based on a hospital’s actual number of complications versus a statewide target rate for each of the 49 MHAC categories. Under this approach, hospitals face strong financial incentives to reduce complication rates. They will also be armed with a sophisticated data analysis tool that will enable them to systematically help achieve this collective goal of reducing complications.

The Washington Post recently (March 19) published a letter to the editor from Robert Murray, the Executive Director of the state's Health Services Cost Review Commission, which offered more detail:

The [MHAC] method of applying hospital rewards and penalties is based on measuring each hospital's performance and determining whether the complication rates are lower or higher than, or on par with, expected rates. The expected rates of complications for each hospital are calculated using statewide average rates for the type and severity of illnesses of the patients treated by a given hospital. Therefore, hospitals with more complex patients are not disadvantaged because their expected complication rates would be higher than those hospitals with less complex patients.

The MHAC approach to funding the rewards and imposing penaities is revenue-neutral and does not raise money for the state through fines; for poorer-performing hospitals, a portion of their approved increase in prices for the current year has been withheld and redistributed to the better-performing hospitals based on performance in the previous year.

In this article, you can see some of the objections to this scheme:

The head of the Maryland Hospital Association says the complication list is too broad and that part of a reported drop last year in the overall rate of complications may simply have been hospitals doing better record-keeping. One leading patient safety expert says the Maryland program – and other national efforts – are moving forward despite insufficient evidence to truly measure and verify the types of preventable complications that should be targeted.

"There is so much pressure to drive down cost and improve quality that politics have gotten ahead of the science," says Dr. Peter Pronovost, a professor at Johns Hopkins University School of Medicine and winner of a MacArthur Foundation "genius grant” for his work on improving hospital safety, often through the use of simple checklists. "There’s a gap between regulators, who say the measures are good enough and clinicians, who say they’re not."

From this vantage point, I am hard-pressed to see how a "focus" on 49 metrics makes much sense. That is unlikely to stimulate a sensible approach to process improvement. Also, the dollars at stake are de minimis -- 0.5% of total inpatient hospital revenue in the state or about $60M -- unlikely to act as much of a financial incentive. Dr. Pronovost has it right. Government regulation of this sort is invariably crude and off-point. It would be much better if the medical profession demonstrated that it is capable of self-regulating in a way that persuasively exhibited a commitment to quality and safety and to patient involvement in the design and delivery of care.


Anonymous said...

Although Dr. Pronovost's comment about accuracy of metrics is important, it is dwarfed by your point that it is 'unlikely to stimulate a sensible approach to process improvement'. Hospitals will inevitably focus on those 49 metrics instead of conceiving of a transformative, system-wide revision of their work - as laid out in Charles Kenney's book "Transforming Health Care."


Paul Levy said...

I don't even think they will be able to focus on the 49. There are just too many items. To use an systems analogy, this is an approach that focuses way too much on the noise and not enough on the signal.

Anonymous said...

Given the modest and siloed improvements made by medicine in self-regulation, these CMS style solution sets should be expected. Until ambitious QI becomes central and explicit to business strategy, there will be bean counting by external stakeholders, and the time and energy drain required to manage and respond. It also tends to relegate quality to specialists rather than explicitly the role of all. It is the tradeoff of teaching to the test, but failing at innovation.

Anonymous said...

"It would be much better if the medical profession demonstrated that it is capable of self-regulating in a way that persuasively exhibited a commitment to quality and safety and to patient involvement in the design and delivery of care."

Well that is the crux of the issue, isn't it? The healthcare profession has not shown that it is capable of self-regulating. The all too often mantras of "my patients are sicker" or "the data are wrong" prevent any forward progress. The inability to develop a single meaningful metric that accurately and fairly captures quality will continue to prevent such progress. It may just be an impossible task. Every provider/researcher/agency/specialty group is quick to develop metrics in their niche area, but having dozens (let along hundreds) of metrics will continue to plague our ability to succintly measure quality.

What if we could only define and use 1 metric to measure healthcare quality? What would that measure be? Maybe if we get everyone to focus in on answering that question and someday attain some semblance of consensus, we'll make actual forward progress.

Anonymous said...

How about preventable harm?

Garrett Bean said...

Preventable harm would be a good start, but our patients do not pay us solely to avoid harm. We need a measure that will assess the effectiveness of care as well as encouraging pro-active medicine.

Anonymous said...

To establish baseline for ACO contract rates, insurance companies have (1) risk scores, and will create (2) health scores of some sort (transparency anyone?) for individuals that bonuses to providers will be based upon. Providers will combine these with (3) clinical and patient participation scores to map progress (albeit in a limited way initially) to receive ACO benefits. From a patient's point of view, (4) quality of life or functional capacity scores might be most salient. Other research is attempting to get at individual health through an (5) 'allostatic load' measure, and there are other global assessments that could clearly be in demand. Where do different global measures intersect, and what delivery quality measures predict the greatest change?

What if we ask the patient what they want when they come in the door and ask them if they got it on the way out?