Hospitals are in a tough spot. Drastic changes to the healthcare landscape, coupled with big increases in patient cost-sharing and the rippling effects of inflation on basic needs like housing and groceries, have created additional pressure on hospitals and households. Hospitals face rising costs, increases in claim denials, ongoing staffing shortages and tariff uncertainties — not to mention absorbing the fallout from recent cuts to Medicaid and Affordable Care Act coverage programs. This places providers in the unenviable position of needing to collect more from patients who, in many cases, are already struggling with their own financial security.
But the status quo has not been working well for patients, either. In a 2023 national survey, nearly four in ten respondents — including almost 30 percent of respondents with employer-based coverage — reported delaying care due to costs. For many people, having health insurance no longer means they can afford to use healthcare. Employer-sponsored insurance (ESI) premiums are outpacing wage increases and continue to grow. Over half of people with ESI are enrolled in high-deductible health plans (HDHPs), leaving them exposed to high out-of-pocket costs. This means that even insured patients are struggling to pay their medical bills, and providers are increasingly struggling to collect from them. One 2022 analysis of patient accounts showed that between 2018 and 2021, bad debt attributable to insured patients soared from 11.1 percent to 57.6 percent. All of this contributes to higher amounts of uncompensated care. Unfortunately, these trends will likely only be exacerbated by recent changes to Medicaid and the healthcare safety net resulting in millions more uninsured patients.
These harsh realities will require hospitals to think more holistically about revenue cycle management as a safety net strategy for connecting patients to care. Quickly and efficiently connecting patients to financial assistance can help prevent patients from accruing medical debt, potentially improving health outcomes. For example, one study of a large health system found that access to financial assistance — including for insured patients with relatively high incomes (350% of the federal poverty level or FPL) — increased healthcare utilization, including ambulatory care, and “increases the detection and management of treatment-sensitive conditions,” such as diabetes. Reducing emergent care by providing preventive care that manages chronic illnesses, for example, promises better patient outcomes and shifts patients to lower cost settings. In this way, such strategies may not just improve patient care — they may also improve hospital finances. Early analysis of medical debt and mortality rates in 2,943 counties has found a correlation between medical debt and increased mortality for all causes of death. Yet research suggests that many patients remain unaware of financial assistance. Making it easier for patients to qualify for financial assistance can connect them to the right treatments faster, build trust and foster loyalty while also improving the long-term health of communities and financial stability of hospitals.
What can hospitals do?
Using predictive analytics to presumptively screen and qualify patients for financial assistance (presumptive eligibility tools) may be a vital strategy to navigating an increasingly challenging environment. Many hospitals and health systems have already begun incorporating predictive modeling offerings from third-party vendors to better manage caseloads and avoid overtaxing their teams. The universe of hospitals using predictive analytics will likely grow as self-pay patient volumes increase and demand for financial assistance grows. Studies have warned that these tools, largely designed for collection purposes, can be used to bar people from accessing needed financial assistance. However, they also, with guardrails, can be deployed to maximize financial assistance for patients while reducing administrative burden for providers. Nascent reporting illustrates that predictive analytics can save hospitals from spending on debt collection, allowing them to direct their capacity to caregiving.
What constitutes best practice for using predictive analytics in this context is still evolving.
Developing guardrails for how and when predictive analytics can be deployed to maximize patient access to financial assistance and reduce administrative burdens for hospital staff is critical and needs consensus support. As the sole nonprofit organization relieving patient medical debt in bulk, Undue Medical Debt is uniquely positioned to understand the complex issues that face key stakeholders including hospitals and other providers. The current system and incentives do not set up providers or patients for success. It is not working for anyone. This research is an effort to identify better practices — not one-off solutions — that hold potential to improve both patient and provider experiences. What follows is an overview of emerging issues and a framework for addressing them in ways that benefit patients and providers alike. Specifically:
- How do hospitals and their vendors currently use predictive analytics to make decisions about financial assistance and collection
- What are the opportunities and known challenges in deploying predictive modeling to presumptively screen people for financial assistance, including ethical considerations
- What are a set of emerging strategies to maximize the benefits — while minimizing the potential risks — of predictive analytics as a screening tool for hospital financial assistance
Over 6 months, we conducted structured interviews with a range of stakeholders including revenue cycle vendors; hospital finance and community benefit leaders from academic medical centers, large national health systems and small regional systems; patient and legal services advocates; and industry thought leaders, including policymakers and community leaders. We asked these stakeholders to identify challenges, opportunities and lessons learned from implementing presumptive eligibility approaches. This is an ongoing discussion as this field develops and matures.
Our work identifies the following emerging strategies:
Strategy 1: Workflow Bolster existing financial assistance policies and procedures to include a robust presumptive eligibility screening and determination approach.
Strategy 2: Technology Adjust current financial assistance policies and procedures to fully leverage predictive analytic tools.
Strategy 3: Evaluation Conduct routine data and landscape scans to understand how best to target, tweak and scale financial assistance policies and presumptive eligibility initiatives.
Strategy 4: Communication Proactively share data and trends with patients, partners and policymakers.
Strategy 5: Leadership Develop champions among senior leaders and board members who will invest in the technology, workforce and infrastructure to upgrade financial assistance policies and procedures as an integral part of delivering quality care.
Hospitals — and patients — face unprecedented strain in 2025. Using predictive analytics to presumptively screen uninsured and underinsured patients for eligibility for financial assistance or public coverage programs, like Medicaid or CHIP, are critical steps in connecting patients to timely care, increasing revenue and (for tax-exempt hospitals) meeting community benefit obligations. These functions are already part of many hospitals’ revenue cycle workflows. But increasing demand stemming from uninsured and underinsured patients will require hospitals to streamline these processes and scale the work. Predictive analytics can help hospitals shift more patient bad debt accounts to charity care or to convert uninsured patients to insured status, helping revenue cycle teams protect critical revenue streams. This would allow leadership to redirect staff time to collection efforts that are more likely to succeed, speeding up the timeframe for communicating the financial assistance determinations or more affordable coverage options to patients and giving them peace of mind in a fraught time. More efficiently helping patients get needed resources is possible with planning, collaboration and upfront investment.
We are in a moment of tremendous uncertainty and disinvestment in healthcare programs. Financial assistance, a bedrock of the safety net, will be critical for patients everywhere in accessing needed and timely care. Hospitals will need tools to manage an inevitable growth in the uninsured and underinsured. Predictive analytics is one tool that can alleviate workloads and get patients to resources. However, it is not a panacea, or without implementation hurdles. While there is a clear case for leveraging these emerging technologies, the data they provide us highlights a deeper issue — the unaffordability of health insurance. We need larger-scale public policy interventions that shore up the safety net in order to fully address the challenges healthcare providers and patients are facing today — but in the meantime, we owe it to patients to find ways to preserve access to care.
More on Presumptive Eligibility
Presumptive Financial Assistance Series
- PART 1 — Intervening Upstream
- PART 2 — Intervening with Accuracy
- PART 3 — Assessing the Cost
