Good Data, Better Outcomes: Why Data Quality Is Central to Medicaid Modernization

Government programs today run on data. Medicaid eligibility, unemployment insurance, SNAP benefits, and dozens of other federal and state programs rely on automated systems and cross-program records to determine who gets what. The quality of this data determines how well these programs function and how fairly they serve the public.

Yet recent discussions among policy leaders have highlighted recurring concerns about data quality in real-world applications. These include inconsistent records, missing information, and the difficulty of reconciling data from multiple sources. While these are not new problems, they have become increasingly relevant as states and federal agencies seek to modernize operations and reduce reliance on burdensome manual processes.

Proven solutions exist. Data cleansing tools can standardize and correct records, while systematic comparison across sources helps identify which information is most reliable. The best path forward is not to avoid data sharing out of fear of imperfections, but to invest in the work that makes data both accessible and reliable.

High-quality, well-cleansed data opens powerful opportunities for states. It enables robust ex parte verification, allowing states to confirm eligibility without requiring individuals to self-attest and submit onerous paperwork that beneficiaries sometimes forget to submit. States have varied in their application of either process. Ohio has led the charge, while others are lagging far behind. The increased usage of an ex parte approach, moreover, can only be expected to climb as states re-establish up-to-date records with enrollees and shake off the artificially inflated presence of COVID-era auto renewals.

Ex parte is only as effective as the underlying data. If states attempt to move into an ex parte environment equipped with poor data, they can expect a cascade of problems. These include slowed processing times, increased appeals, and a bloated backlog, leading to an increased administrative burden and, ultimately, a frustrating experience for beneficiaries who depend on timely support. So important is its quality that even if data is 85% accurate, a state relying on ex parte as the sole verification method could see up to 15% of people lose their healthcare or SNAP benefits. 

Finding the Balance

The most reliable approach blends human oversight with automated data tools. Relying solely on self-attestation or entirely on data are both flawed strategies. We should be leveraging our data to help with the process and assist with verification for self-attestation, but we shouldn't be entirely reliant on it.

Few examples illustrate this risk better than Michigan’s experience over a decade ago. The state deployed a risk-scoring tool for unemployment insurance decisions. Flawed implementation contributed to benefit freezes affecting around 100,000 applicants in a short period, leading to widespread issues including lost homes, vehicles, and eventual litigation. As Michiganders experienced, bad or inconclusive data cannot only create inefficiency; it can produce real harm and erode confidence in public systems.

How Conflicts of Interest Result in Unreliable Data -- and Denied Care

Data quality problems are not always the result of negligence. In a healthcare system shaped by competing, often conflicting, financial interests, the data itself is sometimes skewed by those conflicts. Organizational conflicts of interest lead to higher costs and decreased quality of care. 

When an insurance company determines whether to cover a service, a process called prior authorization, there is a clear incentive to deny care. Plans will sometimes claim that there is a fire wall between them and the entity reviewing the service request. However, when the entity is  a subsidiary of the same corporate parent, there is, at most, only the appearance of an objective review process.

Thus, a recent report from the HHS Office of the Inspector General (OIG), found “Differences in denial rates between for-profit and nonprofit [Medicare Advantage insurers] suggest that financial incentives may be partially driving higher denial rates.” Take United Healthcare’s subsidiary NaviHealth, for example, which handles prior authorization requests for several health insurers including United Health Care. NaviHealth denied requests for admission to skilled nursing facilities at an alarmingly high rate. So much so that independent Medicare appeals contractors such as Maximus later overturned 97% of skilled nursing care denials issued by NaviHealth when enrollees appealed.

Without eliminating conflicts of interest in our health system, data cannot be trusted to guide decision-making.

What Modernization Requires

The path forward requires strategic investment in data quality infrastructure alongside thoughtful policy design. Cleansing and cross-validation efforts, combined with appropriate human oversight, enable programs to move faster without sacrificing accuracy or fairness. This balanced approach also protects taxpayers by reducing improper payments and fraud, while safeguarding beneficiaries from avoidable denials or disruptive delays.

When implemented correctly, this approach enables us to deliver the right benefits to the right people at the right time. For CAMI, that is the measure of true modernization: technology and data harnessed in service of accuracy, accountability, and the people such programs are intended to serve.