COVID-19 has taken a severe toll on socioeconomically disadvantaged populations and racial and ethnic minority groups. It has disrupted traditional health care delivery and strained health systems to the breaking point. The need to address behavioral, economic and social determinants of health – in pursuit of the Triple Aim of better health outcomes, improved patient experience, and reduced per capita cost – has never been so compelling.

In a new study published in Health Services Research, we add to the growing evidence base that community health workers (CHWs) can help meet these challenges. We used pooled data from three randomized trials to show that a standardized CHW program reduced hospitalizations and decreased fragmentation of hospital care, effects that persisted three months after the intervention ended.

CHWs are trained and trusted individuals who often share a socioeconomic background and demographic characteristics with their patients and can therefore provide support consistent with patients’ values and needs. Using participatory action research, including interviews with patients, the Penn Center for Community Health Workers developed IMPaCT (Individualized Management for Patient-Centered Targets), a standardized CHW model focused on partnering with patients to support them in meeting their individualized needs. In three prior randomized controlled trials, IMPaCT has improved a variety of health outcomes, including access to primary care, chronic disease control, and patients’ perceived quality of care.

Our latest study used pooled data from these trials to demonstrate IMPaCT’s efficacy in reducing hospitalizations. Because hospitalizations account for a third of U.S. health care expenditures, reducing preventable hospitalizations is a desirable goal for all health systems working to achieve the Triple Aim. Our study has three main findings:

First, we found that patients who received the IMPaCT intervention had a significant and sustained reduction in hospital-based care. Over 9,398 observed patient months, the total number of bed days per patient in the IMPaCT group was 66 percent of the total in the control group (849 days versus 1256 days). This reduction was driven by fewer hospitalizations per patient and a shorter average length of stay.

Second, we found that patients who received the IMPaCT intervention had decreased fragmentation of hospital-based care, as measured by the proportion of patients with hospitalizations outside of their primary health system (18.8% in the intervention group, versus 34.8% in the control group). Hospitalizations across multiple health systems have been associated with increased hospital costs, longer lengths of stay, and higher mortality rates. Our study suggests that the IMPaCT intervention could help prevent these negative outcomes by strengthening patients’ perceived connection with both their primary care team and their health system as a whole.

Third, we found that reductions in hospitalizations began within the first 60 days of the intervention and persisted for at least three months after the intervention ended. This suggests that IMPaCT was successful in reducing hospitalizations not by encouraging patients to delay needed inpatient care, but rather by addressing the social, economic, and behavioral factors affecting their health and connecting them with long-term medical and social supports that could help them stay out of the hospital.

Figure. Cumulative number of hospitalizations for patients in the intervention and control groups in the nine months following trial enrollment, including the six-month IMPaCT intervention period and a three-month post-intervention period. Source: Vasan et al., Health Services Research, 2020.

These results add to a growing evidence base showing that holistic, theory-based interventions using tailored, longitudinal, relationship-based support can generate a substantial reduction in hospitalizations. This reduction in hospital utilization has the potential to generate significant cost savings; a recent economic analysis found that every dollar invested in the IMPaCT program would return $2.47 to the average Medicaid payer within the fiscal year.

Policy Implications

Our study shows that CHW programs can reduce preventable hospitalizations, thereby preserving needed acute care capacity for patients with severe manifestations of COVID-19. In addition, because CHWs can be rapidly trained and deployed in high-need areas, CHW programs have the potential to help create needed jobs, play a critical role in contact tracing, and improve outreach to the communities most severely impacted by the virus.

At a time when many health systems are financially strained and reimbursement policies are being necessarily reevaluated, public and private payers should implement sustainable mechanisms for reimbursing CHWs for the wide range of services they provide. These findings add timely and relevant evidence that rigorously developed and evaluated CHW programs can play a critical role in improving health outcomes and reducing health care costs.

The study, “Effects of a standardized community health worker intervention on hospitalization among disadvantaged patients with multiple chronic conditions: A pooled analysis of three clinical trials,” was published in Health Services Research on July 8, 2020. Authors include Aditi Vasan, John W. Morgan, Nandita Mitra, Chang Xu, Judith A. Long, David A. Asch, and Shreya Kangovi.