For the last 10 years, my wife and I have volunteered as physicians at the Capital City Rescue Mission Medical Clinic in Albany, New York. The clinic is a safety net medical provider for those without health insurance. It is totally supported by the Capital City Rescue Mission; no patient is billed for the services provided. The clinic provides both urgent care as well as continuous care for chronic medical conditions like diabetes, hypertension, and COPD. We serve a diverse patient population which has changed over time. Initially we saw almost exclusively the residents of the shelter or the homeless living on the street. We also saw a smattering of people from other countries visiting family; these folks had no health insurance but had medical problems or needed refills of their medicines.
But recently this has changed dramatically. As many of the undocumented immigrants in the US made their way to New York, the governor of NY and mayor of NYC distributed these immigrants widely throughout the state and this included Albany. These folks have essentially no health insurance (NY state does give them “emergency Medicaid” which pays only for truly emergent care). So they have found their way to our clinic and are now the majority population. It is not unusual on any given clinic day to see patients from 4 or 5 different countries speaking 4 or 5 different languages.
Taking care of these patients can be very difficult. Some of the challenges are directly related to the fact that our patients do not (and maybe never will) have health insurance. Getting subspecialty care is nearly impossible. That hernia that really should be fixed won’t be until it is incarcerated. Getting expensive drugs paid for is even more difficult. Take for example the patient I recently saw with hepatitis C. A curable disease. Or the diabetic who has failed metformin but doesn’t really need insulin. Guidelines all suggest semaglutide. Forget about it.
But there are yet other ways our poor population is cheated out of optimal care. Having our diabetics do dietary modifications is nearly impossible because they eat whatever the mission is serving, and it’s usually carbohydrate rich. If you don’t have access to an electrical outlet you cannot use a nebulizer to treat your severe COPD. If you sleep in a different place every night, many of them unsafe, your things often get stolen (it doesn’t matter that there is no street market for lisinopril). If you cannot read pill bottles or don’t speak English (and have nobody to read them or translate them for you) it is impossible to be compliant with your complex medical regimen.
Being poor is really bad for your health. My examples may seem like hyperbole but they are not. And although they may represent a far end of the spectrum, even less dramatic social determinants can have a profound impact on health outcomes.
“Social determinants of health (SDOH) are the conditions in which people are born, grow, live, work and age that shape health.” (https://files.kff.org/attachment/issue-brief-beyond-health-care). Health related social needs (HRSN) are “social and economic needs that individuals experience that affect their ability to maintain their health and well-being.” (https://www.cms.gov/priorities/innovation/key-concepts/social-drivers-health-and-health-related-social-needs). HRSN are those social determinants that impact a specific individual (so HSRN are derived from an analysis of SDOH).
There are lots of definitions of SDOH. The KFF publication cited above offers a reasonable classification; SDOH domains include:
1. Economic stability (employment, income, debt, etc)
2. Neighborhood and physical environment (transportation, housing, etc.)
3. Education (literacy, language, etc.)
4. Food (hunger and access to healthy options)
5. Community and social context (support systems, discrimination, etc.)
6. Health care systems (health coverage, provider availability, quality of care, etc.)
It is not at all difficult to identify potential HRSNs in each of these domains. For example, being unemployed, homeless, illiterate, and being in a discriminated group all could lead to a bad health outcome. It is obvious these are not mutually exclusive. It is also obvious that many boil down to poverty. They all contribute to the huge problems we have in the US with health equity.
Exactly how much each HRSN impacts an individual’s health outcomes is impossible to determine. And whether multiple HRSNs have an additive effect or synergistic effect is also unknown. In cancer, for example, HRSN impact virtually every step in the cancer journey, from cancer risk through screening through treatment and through survivorship. ASCO has nicely summarized the issue in a recent position paper (Social Determinants of Health and Cancer Care: An ASCO Policy Statement | JCO Oncology Practice (ascopubs.org)). In many studies, poverty and education are major drivers of poor cancer outcomes.
It's one thing to acknowledge that SDOH are important and quite another to do something about it. The types of interventions required to address many HRSN are just not part of what oncologists do. In fact, addressing HRSN has not really been what physicians do period. But this is changing.
The proximate solution for some of the HRSN lies in community resources. Transportation, food insecurity, and insurance navigation are all problems for which solutions may exist and the key is making the right connections. And there are efforts at both the national level (via CMMI) and the state level (via Medicaid) to address this. Primary care is leading these initiatives.
Oncologists are a little downstream from these primary care efforts, but given the importance of SDOH in cancer outcomes oncologists need to get involved. This was recognized by CMMI in the EOM, the current oncology alternative payment model which requires that participating practices start to assess patients for HRSN. Note that the model does not require providing any sort of solution; rather it requires oncologists to just ASK.
Only recently CMS formalized a way for doctors to get paid for documenting SDOH (Using HCPCS Code G0136 for Social Determinants of Health Risk Assessment | AAFP). Using a standardized screening tool for SDOH every six months, a doctor can bill a G-code for reimbursement. Several of these tools are identified in the AAFP article I have referenced; it is estimated that they take 5-15 minutes to complete and do not need to be done at the time of the physician visit. Participation is voluntary. The physician fee schedule has a reimbursement rate of $18.66. Since this just took effect this year, it’s a little early to tell what impact it will have. But this requires only documentation of the problem, not necessarily attempting to solve it.
At a recent meeting I attended, Alan Balch of the Patient Advocacy Foundation, Todd Goldstein of Memorial Sloan Kettering, and Norm Shore of Patient Discovery discussed a pilot being done at MSKCC in their BMT unit. Patient Discovery is a software tool for social needs assessment. The tool is completed by patients prior to their office visit and can be reviewed at the time of the visit. MSKCC has found that using this technology can really streamline the data collection process and unburden staff. Patients and staff both really liked it, and it allowed personalization of the needs assessment as well as formulation of a plan that was patient specific. They have plans to expand the program. This use of technology should provide a roadmap for all practices to collect this data. But this project also shows the need for clinical staff and social workers who can connect these patients with community resources, so called “Needs Navigators” (as opposed to Clinical Nurse Navigators). We don’t know whether health outcomes will be impacted.
These efforts at data collection of course do not necessarily solve the multiple HRSN, and many of the available solutions will only apply to certain providers in certain geographies. But it is clearly a start. A more comprehensive up-to-date catalogue of available solutions needs to be maintained and made widely available.
There is no doubt that many of the HRSN cannot be addressed in this fashion. Some of these most vexing problems need societal approaches. We are at the beginning of tackling these problems. We don’t know the magnitude of the problems, we haven’t figured out which ones are even amenable to solutions, and we certainly don’t know which ones have the greatest influence on our patients’ health care outcomes. But there is one area that as health care providers we can attempt to remedy, and that area is discrimination.
There have been countless articles in the medical literature documenting that there is bias in how medical care is delivered. Some of this bias is longstanding and at least to me is not ill intentioned. I’ll give you an example. I was taught that hypertension in blacks does not respond to beta blockers and that diuretics are preferred. As it turns out, this may not be completely correct. It’s not that beta blockers are ineffective but this racial distinction was based on faulty observational data or even anecdote and treatments were prescribed (or altered) based on race. The truth is that we still don’t have great studies to answer this question. But over time this old clinical axiom has become completely irrelevant given the vast number of anti-hypertensives, many of whom have a far better side effect profile than beta blockers (and many of which are generic and cheap). Beta blockers aren’t the drug of first choice in any patient subset. But by definition this clinical practice involved racial bias.
There are other biases where the impact may not be as benign. Recently, there was a debate about whether prostate cancer is a more aggressive disease in black men. Prostate cancer is certainly more common in American blacks and they have a worse outcome. Is this because of worse biology? Or is it because of substandard treatment? This is a very complicated question. I am aware of the data that there is different biology (just as, for example, there is more triple negative breast cancer in the black female population). But there is also data that treatment may not be as aggressive; this might be because of lack of insurance, inability to pay the out-of-pocket expense, lack of access to top notch treatment facilities or even patient choice. But if a doctor believes that the black patient is doomed to a poorer outcome, will they recommend less aggressive treatment just because the patient is black?
And then there are just plain troubling differences in treatment of minority populations. There is less use of molecular profiling in blacks. But a disproportionately high share of the Medicaid population are minorities, and even today many Medicaid plans do not cover molecular testing (note that the CMS National Coverage Determination doesn’t impact Medicaid coverage policy). There is less use of chemotherapy in black women with breast cancer. And of course, fewer minority patients are enrolled in clinical trials. Why? Is it implicit bias? Is it racial prejudice? Is it beyond the control of physicians?
I strongly believe a more thorough understanding of these social determinants that impact patient health outcomes is critical. It’s not just race. There is good evidence that the LGBTQ population can be subject to the same biases. The first step is critically evaluating these many disparities without introducing a political agenda. As a practicing physician I always wanted to deliver the best care for my patients. If I am doing something for the wrong reasons I would like to get it right.
I will close by pointing out (in an apolitical fashion) that a major advance in resolving HRSN was the expansion of Medicaid. Since Medicaid was expanded, numerous studies, including several in oncology, have demonstrated improved health outcomes. I am distressed that 10 states still have not expanded Medicaid, and many of these states have numerous other SDOH challenges. I am not naïve; Medicaid expansion costs money, and usually that means higher state taxes. But I feel that Medicaid expansion is just the right thing to do.
American physicians are constantly reminded that compared to the rest of the world, we are doing a lousy job taking care of our population (especially in light of how much money we spend). In fact, just recently the Commonwealth Fund reminded us of how poorly we performed in the Covid era (https://www.commonwealthfund.org/series/mirror-mirror-comparing-health-systems-across-countries). Now I would argue that this report is hopelessly biased and rehashes the same old arguments. But I would suggest that many of the deficits identified in their report trace back to SDOH. So, the problem has less to do with medicine and more to do with poverty and other social factors.
In my next post, I will address how to get the best possible care (at any cost). Specifically, I will discuss second opinions and the use of centers of excellence, especially in cancer care.