To improve quality, whom we measure is as important as what.

by Joe Rotella, MD, MBA, FAAHPM

The growing awareness in America that our healthcare system is unaffordable and fails to deliver consistent quality has led us to healthcare reform, mandatory quality reporting, pay-for-performance, and Accountable Care Organizations (ACOs). We need good measures of quality healthcare so we can direct our limited resources where they provide the most value. We would do well to pay particular attention to the quality of care near the end of life. As Ellen Goodman noted in her recent Harvard Business Review article “Die the Way You Want To”, 25% of Medicare expenditures are incurred by the 5% who are in their last year of life. Surely we can deliver more value by increasing quality of life and decreasing futile medical interventions for these patients.

It makes sense that we should measure quality in order to improve it, but figuring out exactly how can be a real devil. We often focus on what we should measure—the numerator, but whom—the denominator—is just as important. Working on the AAHPM Quality and Practice Standards Task Force this past year has been an eye-opening experience for me. Wading through ponderous technical documents describing a host of proposed national quality measures, I was struck with how hard it is to design a really good one. The National Quality Forum (NQF) recently recommended a set of quality measures for hospice and palliative care. The NQF only recommends measures with well-defined methodologies validated for the specific populations under consideration. Many employ a narrow denominator, e.g. hospice patients rather than dying patients. It’s a good start, but I hope policymakers and payers will be careful in how they incorporate them into mandatory quality reporting systems. Measuring only hospice patients in order to improve the quality of end-of-life care is like searching for a lost dollar bill only where the light is good. The real payoff may be lurking in that dark corner.

For example, the best set of quality measures we have for end-of-life care so far, the Family Evaluation of Hospice Care (FEHC), has only been validated for hospice patients. When we apply the FEHC to hospices, we can compare the quality of Hospice A with that of Hospice B. That might be good enough if most dying patients received hospice care, but the National Hospice and Palliative Care Organization reports in its 2011 Facts and Figures that only 42% of those who died in 2010 were enrolled in hospice. So what about the quality of end-of-life care for the majority of patients who die in hospitals, skilled nursing facilities and homes without the benefit of hospice care? How can we measure that?

Current and proposed mandatory quality reporting programs perpetuate the silos in our healthcare delivery system. Since hospitals are designed for treating acute, potentially-reversible problems, they report post-discharge, patient-rated satisfaction surveys that totally miss the experience of the many patients who die during their stay. Likewise, we see skilled nursing facilities as places for rehabilitation, so we require a Minimum Data Set that focuses on restoration of function, even though many patients languish and die there. Since hospice is the place for dying, that’s where we’ll mandate reporting of end-of-life quality measures, but that’s not going to improve the quality of dying where most of it happens.

The Medicare final rule creating ACOs requires them to report on 33 quality measures linked to how much they can share in any savings. Despite the striking disproportion of Medicare dollars spent in the last year of life, none of these measures has anything to do with comfortable dying or the appropriate utilization of hospice or palliative care. In the original 127-page proposed ACO rule, the word “hospice” appeared once and “palliative” not at all. Comments from the AAHPM and others must have been heard, because the terms appear about a dozen times in the final rule. It’s a step in the right direction, but we have a long way to go. We need researchers to develop and validate quality of life and quality of dying measures for all relevant care settings. Once available, we should push for their inclusion in all mandatory quality reporting and pay-for-performance systems. We’ll never get the maximum value for the Medicare dollar without it.