How to cut Medicaid costs without hurting care

As a primary care physician serving Medicaid patients, I recently witnessed a preventable hospitalization that perfectly illustrates America’s health care dysfunction. My homebound patient with heart failure needed a routine lab test to adjust their medication. While commercially insured patients can access electronic lab orders, my patient’s Medicaid plan contracted exclusively with a laboratory requiring physical forms by mail. The resulting delay in getting his labs and subsequent medication adjustments led to heart failure exacerbation and a preventable hospitalization costing taxpayers thousands.

This case exemplifies a larger pattern: Nearly 40% of emergency department visits and hospitalizations among Medicaid beneficiaries are potentially preventable through timely primary care access. These avoidable events cost American taxpayers approximately $36 billion annually. As Congress considers Medicaid cuts to control federal spending, it should focus on eliminating bureaucratic barriers that paradoxically increase costs.

Past cost-containment efforts ironically created costly inefficiencies. When Medicaid plans reduced outpatient payment rates in 2015, they inadvertently created “phantom networks” of providers who nominally accepted Medicaid but rarely saw these patients. The resulting appointment delays pushed patients toward more expensive emergency care as conditions worsened.

Similarly, to manage prescription costs, many Medicaid pharmacy benefit managers implemented procedural denials — rejecting prescriptions for administrative rather than medical reasons, such as for trying to pick up a prescription a day too early. My health care delivery organization, Waymark, conducted research that found these denials, more common in Medicaid than other insurance types, may reduce immediate drug costs but often trigger ED visits and hospitalizations — for example, when a seizure medication pickup is rejected on Friday night and the patient has a seizure on Sunday, before the physician’s office is able to resolve the issue Monday morning. For key medication classes, prescription denials increased net costs by $624 to $3,016 per patient annually, far exceeding any short-term savings from the denial itself.

Even well-intentioned eligibility restrictions can backfire. Arkansas’ experiment with Medicaid work requirements neither improved employment nor health outcomes, but increased private medical debt and delayed care.

Evidence suggests there are more effective approaches to improve Medicaid efficiency. Accelerating value-based payment models, which hold providers responsible for quality improvement and cost-effective care delivery, represents a clear opportunity. Currently, only 21% of Medicaid payments are structured in value-based arrangements, compared with 43% in Medicare Advantage (one area where Medicare Advantage is succeeding). Given Medicaid’s fragmented state-based structure, federal guidelines promoting value-based models that incentivize better care at lower costs are essential.

Unfortunately, most value-based arrangements in Medicaid today do not actually incentivize meaningful improvements in care delivery. These arrangements, which are sometimes referred to as “value veneers,” are nominal value-based contracts that are mostly traditional medical care reimbursements, with modest pay-for-performance bonuses for achieving low-bar quality metrics.

These contracts don’t meaningfully alter how care is delivered. Doctors still rely on patients to show up to their clinics and get their exams, labs, and medications as per usual, and there’s little incentive to monitor patients between visits to prevent a hospitalization, or find patients who didn’t show up to the primary care appointment.

Instead, states should replace these old styles of contracting and health care delivery with genuine value-enabled models in which health care organizations are responsible for transforming care delivery. This would include services like homeless patient outreach under bridges and tunnels, nutrition and physical activity counseling integrated into the clinic, and comprehensive behavioral health support to address the barriers to getting mental health or substance use treatment as a patient with Medicaid. A true value-based approach would incentivize prevention and proactive care that results in reduced overall costs by keeping people healthier, rather than reward the current system of reactive, visit-based billing that ironically pays more when people are sicker and hospitalized.

Technology adoption represents another untapped opportunity. State Medicaid agencies should replace outdated risk prediction systems like the Chronic Illness and Disability Payment System, which research shows has very low predictive power for identifying at-risk Medicaid patients. Instead, modern machine learning tools can now predict avoidable emergency department and hospital visits in Medicaid with over 90% accuracy, enabling early intervention.

For example, Medicaid patients with diabetes who take insulin often get low blood sugar during the last week of every month — not because of biology, but because the Supplemental Nutrition Assistance Program (SNAP, formerly “food stamps”) are typically distributed at the beginning of the month, so patients have less food and take the same amount of insulin and their sugars drop too low, resulting in a preventable ambulance ride or emergency department visit. Proactively providing food vouchers via a social worker, and insulin education from a pharmacist, can prevent such cycles. The tools to ping care management teams to be truly proactive rather than reactive about such issues can focus resources on patients before they become high utilizers of acute care.

Few Medicaid plans utilize such predictive technologies, instead typically waiting until patients become “superutilizers” or “high cost claimants” before intervention — the health care equivalent of waiting for cancer to metastasize before treatment.

Additionally, Medicaid programs should redirect funds toward community health worker programs as a standard part of the health care workforce. These interventions connect patients with trained community members who address social and health needs that traditional medical visits can’t manage. Research shows these programs generate strong returns on investment — $2.47 for every dollar invested according to randomized trials, and up to $3.73 when community health workers are integrated with multidisciplinary teams including pharmacists, therapists, and care coordinators. These programs simultaneously reduce costs and create stable jobs in underserved communities.

Waymark’s recent study in Virginia and Washington demonstrated that community-based care teams using predictive technology reduced avoidable ED visits by 20% and hospitalizations by 48%, generating over $200 in monthly per-patient savings while creating local jobs for teams providing proactive preventive care.

For fiscally responsible policymakers, the path forward is clear: You can reduce Medicaid expenditures without creating counterproductive barriers to care. By making Medicaid more efficient through smarter policies and technological innovation, America’s largest safety net program can deliver better care at lower costs to those who need it most.

Sanjay Basu, M.D., Ph.D., is a primary care provider and co-founder of Waymark, a public benefit company focused on improving access and quality of care in Medicaid. He is the author of “Transforming Medicaid” (2024).