3M Health Information Systems
Helping to pay for a patient’s funeral: Socioeconomic disparities and population health
Since completing my training in internal medicine in 1976, I’ve been seeing low income patients as a primary care internist two days a week. Not infrequently, we take up a collection among the clinic staff for any number of issues that impact the lives of our patients. For people who live in the margins, the most heartrending of these situations is not being able to afford the burial of a loved one. More than once, I have seen patients face this tragedy, and one of the most poignant instances occurred a few weeks ago with the death of the husband of a diabetic woman, both of whom were patients of mine. We were able to collect a few hundred dollars to help her pay for the funeral, but not enough to avoid cremation, despite the fact that her husband had explicitly stated in his will that he wished to be buried.
Thinking about the challenges facing these patients (and even the clinic’s medical assistants, many of whom are single moms and often one step away from poverty themselves) made me consider the issues involved in reimbursing facilities for their care. Population health management and accurate capitation rates depend on risk adjustment of patient populations. However, most current risk adjustment approaches lack clinical detail that if included, could account for a significant portion of the impact of socioeconomic factors on the health of patient populations. To account for these factors, our research group believes that risk adjustment methodologies should take advantage of currently available information, including the following three categories:
- severity adjustment of chronic illnesses that are, in turn distinguished from each other when clinically distinct
- health status
- pharmaceutical data
Chronic illnesses not only need to be stratified by severity of illness (as we have done, for example, with our Clinical Risk Groups methodology), but there should be clinical distinctions that specify the clinical detail that is available. For example, a patient with cerebral palsy needs to be stratified by severity of illness, but if the individual is in foster care, he or she should be identified in a separate risk category. Without this detailed approach to risk categorization, it is inevitable that managed care organizations will assiduously pursue adverse risk selection—in other words, create incentives to provide treatment for individuals who are least likely to generate high medical expenses and to limit services to high-utilization populations.
In speaking with one large Medicaid organization enrolling dually eligible individuals (Medicare and Medicaid), I learned that the HMO had a specific program for hemophilia case management that attracted a large number of these patients. Not surprisingly, they stopped the program as their reimbursement rates did not recognize this specific category of highly complex individuals. My patient who recently died was under active treatment for a cancer that required an intestinal resection, and thus he did not have bowel movement through his rectum. In addition, he had diabetes, hypertension, and a learning disability—all of which can be captured using diagnostic data. Unfortunately, most risk adjustment methodologies do not capture this information, resulting in highly suspect conclusions when it comes to capitation risk adjustment and tracking populations for outcomes.
Unfortunately, health status is not considered at all in the dually entitled capitation programs that CMS is encouraging, although many researchers—including our research team—have shown that health status is important and routinely collected for all individuals who have a significant chronic illness(es) . In fact, all managed care organizations are collecting this information, at least as measured by the health professional .
If policymakers wish to avoid additional adverse risk selection they should, at a minimum, require both payers and providers to include a health status data element in their risk adjustment methodologies so that plans can be made to expeditiously incorporate health status assessment into their analysis. In addition, worsening of health status (e.g. worsening of diabetes) is one outcome measure that should be tracked by managed care organizations . As our research demonstrated, there are significant differences in rates of progression of diabetes and hypertension in managed care plans.
There are two approaches to incorporating pharmaceutical data into risk adjustment. The easiest and most statistically oriented approach is to simply do a regression analysis. One will find, not surprisingly, that the prior use of expensive medications predicts a future use of expensive medications (and overall resources). One will find, not surprisingly, that the prior use of expensive medications predicts a future use of expensive medications (and overall resources). Such an approach rewards managed care organizations for the use of expensive medications. A categorical clinical approach that assigns, for example, the use of insulin to severity level 2 – diabetes, rewards managed care organizations that are appropriately encouraging use of the entire therapeutic armamentarium. Just as important, such an approach can be utilized to easily track managed care organizations that are stinting on approvals for medications.
There are additional types of information that could be made available, such as homelessness and literacy, which I will focus on in my next blog, along with the practicality/utility of collecting these new data elements. The last blog in this three-part series will discuss statistical and financial strategies that can mitigate some of the impact of socioeconomic differences and encourage managed care plans to take on the sickest of the most disadvantaged individuals.
From both a quality of care and, to a certain extent, a financial point of view (there will be little savings for excellent case management for some of the sickest patients), it is critical that there be a more thorough dialogue on socioeconomic differences and risk adjustment. Hopefully this blog along with the engagement of others will contribute to this important policy question. Better recognition of these issues can only result in better care for my low income patients, one of whom died a few weeks ago without a decent burial.
Norbert Goldfield, M.D., is Medical Director for 3M Health Information Systems.
 Medicare Payment Advisory Commission, Report to the Congress: Medicare and the Health Care Delivery System (Washington: MedPAC, June 2013), chap. 3.
 Fuller RL, Goldfield NI, Averill RF, Eisenhandler J, Vertrees JC. Adjusting medicaid managed care payments for changes in health status. Medical Care Research and Review. 2013 Feb;70(1):68-83.