I’ve had the distinct privilege, as an AARP-sponsored scholar-in-residence at the National Center for Complex Health and Social Needs to continue evolving the ideas spurred by my DrPH dissertation research on innovating ways to replicate models of care for vulnerable populations that while more effective, are also more difficult to spread. Last week, Health Affairs Blog published a post that I wrote with colleagues from the Camden Coalition of Healthcare Providers on this subject. To give it a read click here. There are some exciting new approaches that we can take to this problem and a path to do so that can build a field of practical knowledge we need to transform our health care system for the better.
Health service research is usually expected to jumpstart rather than stall innovation in health care delivery. Except when it doesn’t. And when it doesn’t, it can deter policy makers, funders, and health leaders from maintaining the constancy to purpose required for breakthrough innovations. In the June 26, 2013 issue of JAMA, Joynt et al. used Medicare claims data and a health systems research classification schema to analyze the percentage of Medicare admissions that are “preventable” vs. “nonpreventable.” They did this for “high cost” and “non-high-cost” Medicare beneficiaries. A relatively small percentages of Medicare hospitalizations were deemed to be “preventable” (10-17%) according to this methodology. This conclusion is a buzzkill for generating commitment to invest in innovations in preventive heatlh. Worse yet, it’s wrong.
For one, the schema for classifying admissions as “preventable” or “nonpreventable”, promulgated by AHRQ is a model that has outlived its usefulness in the age of population health. As the late George E.P. Box (statistician with major contributions to quality control, time-series analysis and Bayesian inference) said, “All models are wrong, but some are useful.” And some are not. Such appears to be the case with this AHRQ schema. On what grounds can one make such an assertion? Two; face validity and empiric experience from a 12-year intensive randomized controlled trial of a new model of advanced preventive care.
With respect to face validity, several categories of “nonpreventable” conditions per the AHRQ schema are clearly preventable according to many studies in preventive and public health research. A few examples; orthopedic fractures related to falls, ischemic heart disease, stroke, kidney failure, peripheral vascular disease, and others; many of which are among the most powerful drivers of health care costs. Because these conditions are not easily and quickly prevented solely by a short-term increase in visits to physicians they are classified as “nonpreventable”. Is that the extent of our thinking about the possibilities for preventing these forms of human suffering and their associated hospitalizations and costs? Is that all we’ve got?
The authors do acknowledge that, “… while a proportion of these very expensive inpatient episodes may be potentially preventable, … , their prevention would likely require a long time horizon and substantial investments in population wellness.” Then to complete their kiss of death for investments in prevention, they go on to say … “Such investments are critically important for ensuring the health of the population, but the time frame needed to see cost savings is likely years, not weeks or months.” Anyone familiar with the realpolitik of health policy in the U.S. today would interpret this to mean … this area of health system innovation is dead and won’t be funded for the foreseeable future. We want savings and we want it now – not years from now – and there isn’t enough savings through prevention in the near term to make it worthwhile. At least that’s how these authors chose to interpret their data.
High quality research shows that the most innovative of preventive programs for vulnerable Medicare beneficiaries can yield significant results in a much shorter time. Ironically, Joynt et al. cite not one, but two such papers related to the Medicare Coordinated Care Demonstration as evidence that their “… findings may shed light on why many recent efforts to control costs for these medically complex, high-utilizing patients, including the Medicare Coordinated Care Demonstration (MCCD) programs, have failed to do so.” This is ironic because these same publications also highlight one of the original 15 programs in the MCCD, Health Quality Partners (HQP), as successful in reducing hospitalizations and cost among complex, high-cost Medicare beneficiaries.
From the 2009 JAMA paper that Joynt et al. cite, HQP’s positive impact among high risk beneficiaries was described as follows; ” … for this subgroup [highest severity cases] both differences were large (-29% for hospitalizaitons and -20% expenditures) and statistically significant (P=.009 and P=.07, respectively).” And from the more recent June 2012 Health Affairs article “… Health Quality Partners, reduced hospitalizations by 30 per 100 beneficiaires (33 percent; p=0.02).” Not cited by Joynt et al., was the finding from a July 2012 PLoS Medicine article (also reported in CMS’ Fourth Report to Congress on the MCCD) that the HQP program was associated with ” … a 25% lower relative risk of death (hazard ratio [HR] 0.75 … the adjusted HR was 0.73 (95% CI 0.55-0.98, p=0.033).”
And finally, the Fourth Report to Congress on the Evaluation of the Medicare Coordinated Care Demonstration of March 2011, reported that HQP’s program had no statistically significant impact on “potentially preventable hospitalizations” (using the same AHRQ criteria as that used by Joynt et al), yet produced a 39% decrease in overall hospital admissions (p<0.01). Either HQP’s program prevents the unpreventable or we seriously need to rethink our assumptions about what is preventable as we pioneer new models of care.
Is it unusual that only HQP’s program has had such effectiveness within the MCCD? By most innovators’ standards a 1 in 15 ‘hit rate’ of success is very high. Think about how many compounds were screened to find a few promising HIV or cancer therapies. Or the number of lightbulb designs tried by Edison. Do health service researchers, policy makers, and health bureaucrats understand the nature of innovation? Do they see the relationship between Einstein’s self-assessment “It’s not that I’m so smart, it’s just that I stay with problems longer,” and the need for constancy to purpose to solve our most challenging problems in health care? Do they understand that just labeling a bunch of different programs having different design and performance characteristics with the same categorical name (care coordination, care management, etc.) doesn’t make them all the same? If not, our health care woes are just beginning.
Note: This post was written by me (Ken Coburn, MD, MPH) and originally posted at http://www.hqp.org/blog/?p=547 on 12/15/2013. It is reproduced here, again. Join HQP in preventing the unpreventable.
In the June 2012 article in Health Affairs by Brown et al., “Six Features of Medicare Coordinated Care Demonstration Programs That Cut Hospital Admissions of High Risk Patients” a subgroup of patients was defined using criteria available in Medicare claims data;
[(HF, CAD, or COPD) AND ≥1 hospitalization in prior year]
OR [(diabetes, cancer (not skin), stroke, depression, dementia, atrial fibrillation, osteoporosis, rheumatoid arthritis/osteoarthritis, or chronic kidney disease)
AND ≥2 hospitalizations in the prior 2 years]
Members of this subgroup participating in the Health Quality Partners (HQP, http://www.hqp.org) program had -33% fewer hospitalizations (p=0.02), -30% lower Part A & B Medicare expenditures (with program fees excluded) (p=0.045) and -21.5% lower net costs (program fees included) (p=0.15). All terrific stuff and since the emphasis of this particular analysis was to identify common elements of successful programs, using complex subgroup definitions for that purpose is fine. However, there are significant real-world challenges in trying to use such a complex eligibility criteria for program implementation and scalability.
In the HQP experience, it remains hugely challenging to cobble together a patchwork of collaborative data sharing agreements with hospitals and primary care practices in order to serve a geographic region. Complex criteria sets such as these make that job harder. Having worked many years with the authors of this article I know that they too are fully aware of and appreciate this concern, but the inexperienced reader might confuse or meld these two separate issues: finding common elements of successful programs vs. defining the “best” target population for scaling effective care management interventions.In tables in the Appendix to the article another, simpler subgroup is defined as;
HF, CAD, or COPD
Just having one or more of these 3 conditions meets this subgroup criteria; no other prior hospitalization usage, other co-morbidities, etc. This group is a lot easier to “find” prospectively with data readily available in primary care practices (their billing data). In the demonstration, HQP randomized 695 individuals meeting these criteria (43% of all those in the study) vs. just 273 (17% of those enrolled) of the more complex subgroup above. Results for this simpler, more easily identified subgroup? For HQP’s program, not bad; -25% fewer hospitalizations (p=0.005), -20% lower Parts A & B Medicare expenditures (-$220 per person per month) (p=0.02), and -10% net savings when program fees were included (-$116 per person per month) (p=0.22).
There are plenty of challenges to scaling highly-effective care management programs like HQP’s. One challenge we can and should avoid is making the criteria set for eligibility needlessly restrictive and difficult to implement – especially when the evidence supports a wider population of people who can benefit. With each larger scale cycle of testing, the criteria can be further refined (and coned down, if necessary), but in the meantime, we should encourage the use of target group criteria that are feasible to implement and support system redesigns with the greatest possible chance of successfully transforming our health care system for the better.
This same blog article is also posted on the HQP blog at http://www.hqp.org/blog/
HQP’s nurse care management model is such an effective preventive intervention that it saves lives among chronically ill older adults. I’m proud to be the lead author of the study undertaken by the remarkable team at HQP and published in PLoS Medicine. It’s a compelling proof of concept for a whole new approach to Advanced Preventive Care. Read the full report here; http://dx.plos.org/10.1371/journal.pmed.1001265