From 3M Health Information Systems
A national quality strategy relentlessly pursuing improvement of eight outcomes: Is now the time?
Everyone, except for the measurement industry, seems to agree that we have too many metrics1 for measuring performance. An increasing number of individuals and organizations are also saying that the metrics are too process oriented and that, in fact, we should be focused on outcomes2. We’ve articulated this position before in previous blogs. With the Trump Administration taking office, it is likely that there will be changes in the Affordable Care Act (ACA). Recently, the National Academy of Sciences released a report entitled, Metrics that Matter for Population Health. As summarized in the introduction, the report was designed to “explore the status and uses of measures and measurement in the work of improving population health.” It begins with an overarching set of comments, some of which are highlighted below. Our observations, which are noted in italics, form the core of this blog.
Here are the comments from the report:
- Proliferation of metrics creates confusion. Agreed
- The best measures drive action and are linked to interventions. We would add: only if linked at least in part to financial incentives.
- The greatest opportunities to improve population health reside outside the traditional health sector; therefore, good measures are needed to catalyze action (and collaboration) among those sectors. Agree but only to a point.
- However, indicators work best in catalyzing population health action in multiple sectors when they reflect collective needs and priorities determined by community stakeholders and have been measured at a human scale. This all depends on what these words translate into.
- The metrics realm requires a shift from “data first” to “purpose first.” Again, this all depends on what these words translate into.
Teutsch noted in this report that a 2012 IOM report (IOM, 2012a) identified three areas of population health for which metrics were needed: health outcomes, the health of communities, and health-adjusted life expectancy (HALE), which create a summary measure of the health of the total population (page 8). To make it simple, while the report acknowledges at the outset the proliferation of metrics and how that creates confusion, it continues to highlight (directly or indirectly) hundreds if not thousands of measures.
So in this blog, we don’t need to rehash the arguments against the proliferation of metrics – except we still have no idea how to kill this beast. All ideas are welcome. We’ve also discussed the second point regarding the need for linking metrics to interventions, but to restate our three-legged stool: pay for better outcomes, encourage collaboration (both by bringing providers together and via financial incentives) and disseminate information. Payers, legislators and the public should have all the information they need to identify, pay for and demand better outcomes. This is a prerequisite to driving health system collaboration toward goals that are set by those receiving care rather than those delivering it.
While, in theory, we agree with the third point we encourage that we walk before we run and start with measuring outcomes in the health sector. As we’ve highlighted many times, there are two types of health outcomes – those that that can be translated into dollars and those that cannot. Those outcomes that can be translated into dollars—largely but not completely, such as preventable hospital admissions—are within the scope of the traditional health sector. In contrast, increasing confidence or empowerment, which cannot be translated into dollars, very much lies beyond the traditional health sector. But, again, we need to walk before we run.
Items number four and five become meaningless verbiage unless we start with the health sector and incorporate financial incentives that tie the traditional health sector to other sectors (such as housing, education, prisons, road safety). If we seek to impact public health then the purpose must be to improve health outcomes – understanding that we start with the traditional health sector and move out from there.
In these days of tweets and social media, we need to craft messages that are simple. Improving health outcomes is simple and is something everyone, liberal or conservative, can sign on to. Both the New York and Texas Medicaid Programs have committed themselves to improving outcomes. But we cannot risk drowning out this simple message with hundreds of measures. We need no more than the following:
- Potentially preventable admissions, readmissions, ER, outpatient services, and complications (five measures right there, each of which can be translated into dollars)
- Patient activation, empowerment or confidence (using either the Hibbard or Wasson measures)
- Change/ Improvement in mental or physical health status
- Risk Adjusted Life expectancy or mortality
Underneath these eight metrics, we need a tool that any type of individual consumers (patients, families, physicians, integrated delivery health care systems, payer, housing authority, educational system) could utilize to construct a summary quality of care index composed of domains that are of interest to the individual consumer. While directed to classically understood “individual consumers,” the product should be applicable to any consumer, including health professionals. We could have hundreds or even thousands of metrics underneath these eight metrics.
But let’s be simple and start with this core set of eight measures today.
Richard Fuller, MS, is an economist with 3M Clinical and Economic Research.
Norbert Goldfield, MD, is medical director for 3M Clinical and Economic Research.
- Stempniak, Marty Don Berwick Offers Health Care 9 Steps to End Era of ‘Complex Incentives’ and ‘Excessive Measurement’ http://www.hhnmag.com/articles/6798-don-berwick-offers-health-care-9-steps-to-end-this-era-of-greed-and-excessive-measurement
- Chapter 3 Measure Quality of Care in Medicare p. 39 http://www.medpac.gov/docs/default-source/reports/chapter-3-measuring-quality-of-care-in-medicare-june-2014-report-.pdf?sfvrsn=0