• Community health workers and hospital readmissions

    Regular readers will recall my many posts on the lack of sensitivity of Medicare’s hospital readmissions measures to socioeconomic status. See, for example, this or that. Very attentive readers with an exceptional memory may also recall that I participated in an evaluation of the evidence relating to community health workers (CHWs). Results of that evaluation are here (PDF) and more is linked to from here.

    A few documents relevant to these two threads, and their intersection, crossed my desk in the last week. Below are a few, relevant excerpts. (Emphases added.)

    From Shreya Kangovi, Judith Long, and Ezekiel Emanuel in JAMA:

    Low-income African American patients [] are up to 43% more likely than their higher-income white counterparts to find themselves back in the hospital within weeks of discharge. As a result, the cost of care for these disadvantaged patients is high, as illustrated by the population of low-income patients who are dually eligible for Medicare and Medicaid. Dually eligible individuals cost twice as much as other Medicare beneficiaries largely because they are 4 times as likely to be readmitted to hospitals for ambulatory care–sensitive conditions. […]

    Yet poor health status is only part of the reason for the readmission of patients []. Besides disease burden, factors perpetuate the revolving hospital door for low-income patients: lack of access to medical resources such as a regular source of care, competing socioeconomic issues such as homelessness or food insecurity, and social isolation. […]

    Existing postdischarge interventions frequently fail to help patients []. First, they do not target patients with low socioeconomic status. In fact, many postdischarge services are only available to patients with insurance. Second, even when such services are offered, low-income patients use them at a low rate owing to mistrust of clinical personnel like nurse practitioners and home health aides. Most importantly, many postdischarge interventions are fundamentally clinical interventions, delivered by a workforce trained to address clinical issues. Paradoxically, intensifying clinical follow-up care [] might actually increase admissions; outpatient medical providers often do not have the tools to address the underlying social causes of poor health and have no choice but to refer these patients back to the hospital when they inevitably fall ill. […]

    [T]he CHW workforce, may be able to reach marginalized patients and link them into nonmedical support systems. They can help to address material needs for resources like food or housing as well as social needs, such as the need for purpose or socially meaningful interactions. CHWs share socioeconomic status with the individual patients they serve; this shared life experience affords CHWs a high level of what sociologists call an “empathic understanding” of their patients; they have experienced similar stressors as their patients and have a knowledge of local community resources that clinical staff may lack. Therefore, CHWs may have an enhanced ability to provide both emotional and instrumental support.

    From a NEJM Perspective by Prabhjot Singh and Dave Chokshi:

    The Affordable Care Act (ACA) includes levers to shift our health care system’s focus toward comprehensive, high-quality care for populations. Through structures such as accountable care organizations and incentives such as readmissions penalties, hospitals are increasingly responsible for the care of patients both in and outside the hospital. For example, hospital systems have invested in care coordinators, aiming to reduce readmission rates by stratifying patients according to risk level and tailoring their discharge interventions. As these systems look further beyond their own walls, they may see opportunities for lower-cost, CHW-based programs to demonstrate superior value.

    Beyond reducing readmissions, CHW programs may help to address the root causes of preventable chronic disease. Social exclusion, poverty, marginalization, and the built environment contribute to the high burden of chronic disease, particularly in low-income communities. But social services addressing these social determinants of health are too often fragmented. CHWs who can integrate knowledge of the local social service milieu with knowledge of patients’ individual circumstances can create a vital link for vulnerable populations. In concert with social workers, CHWs can mobilize social support, create avenues for family members to engage in the care process, and strengthen long-term community relationships that help patients sustain healthful behaviors.

    From a Commonwealth Fund Issue Brief that is definitely worth a full read, by Clifford Marks, Saranya Loehrer, and Douglas McCarthy:

    Because the measure used for Medicare’s penalty is not adjusted for patients’ socioeconomic status (SES), and because patients with lower SES experience higher rates of readmissions, safety-net hospitals on average receive higher penalties under the current regime. While adjusting for SES could address this concern, such a move would simply hide and perpetuate a disparity that we as a society should be working to rectify, the panelists noted. […]

    Experts [] noted the futility of discharging vulnerable patients into communities lacking strong networks of primary care and the community support systems necessary to aid patients in their recovery.


    • Aaron Carroll had (a while ago) presented some evidence showing that racial heterogeneity in the US was not driving a significant component of the cost difference in areas in the U.S.

      This study seems to show data that supports the opposite view (though in a very limited sense given it only is relevant to readmissions).

    • Time for one from the Wayback Machine:


      “Question for advocates of the readmission penalty (no snark intended – seriously):

      Have you ever worked in a hospital setting where a big chunk of the patients come from the underclass? What would happen to Mayo’s readmission rates if you stuck Mayo smack in the middle of a neighborhood that’d been an epicenter of poverty and social dysfunction/despair for decades?

      There’s an immense number of things functions that the personal traits and social network around a married Mormon dental hygenist in SLC performs to keep her from returning out of a hospital that an unmarried, unemployed ex-custodial worker in Roxbury, MA doesn’t have and never will. They can come in equally sick and get exactly the same care, but does anyone believe that they’ll be equally able to look after themselves and equally well looked after once they walk out the door?

      Unless there’s some mechanism that accounts for these differences in a meaningful way this particular “stick” is simply going to mean hospitals like Boston Medical Center – which care for lots of patients who have more in common with the later example – are simply going to get penalized for caring for poor people, who are going to return more often for things that doctors and hospitals can’t do anything about. If this isn’t done carefully it will simply result in fewer resources for the hospitals and patients that need them the most.”

      • I agree that is a big problem and I also work in and around Boston Medical Center and watch these issues go on. You can spend a lot of resources to attempt to turn this around, but you are not going to get reimbursed for them.

        We have our own take on these issues just out in Health Economics Policy and Law:


        • I believe that Maryland is trying to address the “you are not going to get reimbursed” part through the State Innovations Models initiative. The state is hiring CHWs through local non-profits or local departments of public health. The state is trying to estimate the savings that this intervention will cause (as compared to an estimated Medicaid baseline spending trend), and it is trying to use the savings to continue funding the CHWs. Basically, getting medical system dollars to fund the intervention.

          The whole concept of estimating a baseline and estimating savings ranges from somewhat to very fishy to me. But it does at least attempt to address the reimbursement problem for non-medical services that improve health. That said, there is also the problem of scale: I’m not sure we have implemented CHW interventions on a large scale yet, and there are bound to be hiccups when anyone tries to do so.