The benefits of care coordination across clinicians, specialties, and settings have been a central focus of health care payment and delivery system reforms. Better care coordination, it has been argued, is a remedy for a fragmented health system and could lead to improved health outcomes, superior patient experiences, and lower costs. But research on the impact of care coordination programs thus far has yielded mixed results, and it is far from clear that the promise of these programs has been or will be fulfilled—particularly when it comes to controlling overall medical costs.1-3 A new study by Berkowitz and colleagues4 adds to this conversation, helping to answer some questions while raising others.
The authors use a difference-in-differences design to evaluate cost and utilization changes for Medicare and Medicaid beneficiaries in the Johns Hopkins Community Health Partnership, a comprehensive, multipronged care management initiative funded by the Centers for Medicare & Medicaid Services. This intervention had 2 main components: an acute care intervention (ACI) and a community intervention (CI). The ACI was a bundle of services—early discharge planning, patient education, medication management, and follow-up after discharge—at 2 John Hopkins–affiliated hospitals aimed at smoothing the transition from inpatient to postacute and outpatient care. The CI used a predictive model to identify outpatients who were at the highest risk for hospitalization and deployed teams of community health workers, behavioral health specialists, and nurse care managers to work with patients and primary care practitioners to coordinate care.
With some exceptions, the results of both interventions were positive but varied significantly by the population studied. Medicaid patients receiving the ACI had lower emergency department use and fewer follow-up visits, resulting in savings of $4295 per beneficiary-episode (inpatient stay plus 90 days following hospitalization) compared with a control group drawn from neighboring hospitals. By contrast, Medicare patients experienced a reduction in follow-up visits but an increase in hospitalization and readmission rates. Despite this increased utilization, the authors calculate savings of $1115 per episode. This is a counterintuitive finding that warrants further explication. Many care coordination programs are designed to lower costs by reducing hospital use and improving outpatient follow-up; the authors suggest an alternative pathway to decreasing costs related to less intensive acute resource use (although this was not empirically demonstrated).
The CI appears to have been successful for Medicaid patients but not for Medicare patients. The intervention was associated with statistically significant reductions in avoidable hospitalizations, emergency department utilization, and readmissions for Medicaid patients—resulting in savings of $1643 per beneficiary per quarter compared with a propensity score–matched control group. But there were no significant utilization or cost improvements for Medicare patients.
There are few rigorous cost-effectiveness evaluations of care coordination programs. Prior research has yielded conflicting results overall, in part because of the highly heterogeneous nature of the populations and interventions studied.1 One recent randomized clinical trial examined the effect of home-based nurse care coordination on frail elderly Medicare patients and found a net cost savings of $296 per beneficiary per month.5 Other trials have found that care coordination can lower rates of emergency department visits and hospital readmissions, particularly for older high-risk patients—but these did not include formal cost-effectiveness analyses.6,7 Meanwhile, the Community Preventive Services Task Force did find sufficient evidence of cost-effectiveness of community health worker interventions for cardiovascular disease prevention and diabetes prevention and management.8
The present study, while instructive, is not without limitations, and its results should be interpreted with caution. First, the study is not a formal cost-benefit, cost-effectiveness, or return-on-investment analysis. For example, the evaluation did not incorporate the cost of the intervention itself (including in-kind costs). Analytic constraints imposed by the Centers for Medicare & Medicaid Services may have limited the generalizability of the economic findings. Second, the initiative took place in Maryland, which implemented its hospital global budgeting efforts during the study period, potentially limiting the study’s implications for states not engaging in similar reforms. As the authors note, global budgeting may also have had disparate effects on participating hospitals vs comparators.
Still, there is much we can learn from this study, given its large sample size, diverse patient population, and the range of care coordination services evaluated. Practitioners are still trying to determine how best to identify patients who need care coordination; how to weave together transitional and more longitudinal interventions; and what works with respect to addressing unmet social needs. Berkowitz and colleagues provide support for a growing recognition that no single care management program is likely to be effective for diverse high-risk patient populations differing in their needs and circumstances. For instance, the ACI and CI had substantially different cost and utilization effects for Medicaid compared with Medicare beneficiaries. Programs should therefore start with a detailed understanding of their own high-risk patients, followed by segmentation of that population into coherent subgroups for specific interventions.9
The study also offers fodder for other research questions that should be rigorously evaluated. Both interventions studied used a bundle of services, but it is not clear which ones were most responsible for the demonstrated cost reductions—an important consideration in resource-poor settings where patients may not have access to an entire suite of services. It is also not clear how the costs (or savings) of care management interventions are spread across a health system or over a patient’s entire continuum of care—that is, to what extent costs are not saved, but simply shifted. Finally, while the study’s interventions focused on cost and utilization effects, we should bear in mind that more fundamental aims of care coordination are to improve patient experience, quality, and health outcomes.
Although most analyses of care management have found modest savings, if any, there are pockets of promise. The relative success of the Johns Hopkins Community Health Partnership suggests that comprehensive inpatient and outpatient care coordination can be effective—even, and perhaps especially, in socioeconomically disadvantaged areas like East Baltimore, Maryland.
Published: November 2, 2018. doi:10.1001/jamanetworkopen.2018.4295
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2018 Khullar D et al. JAMA Network Open.
Corresponding Author: Dave A. Chokshi, MD, MSc, NYC Health + Hospitals, 199 Water St, 31st Floor, New York, NY 10038 (email@example.com).
Conflict of Interest Disclosures: None reported.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Khullar D, Chokshi DA. Can Better Care Coordination Lower Health Care Costs? JAMA Netw Open. 2018;1(7):e184295. doi:10.1001/jamanetworkopen.2018.4295
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: