Equipo Health points to hidden risk in diagnosis-code scoring

Equipo Health is spotlighting research showing that diagnosis codes alone can miss important patient conditions, affecting risk scores, reimbursement, and care decisions. The company says AI-powered clinical review can help healthcare organizations find coding gaps and build a more complete view of patient risk.

Why it matters: - Risk scores shape reimbursement, care management priorities, performance evaluation, and financial risk in value-based care. - When diagnosis codes miss conditions documented elsewhere in the medical record, healthcare organizations can underestimate patient complexity. - Inaccurate risk assessment can also affect compliance, audit readiness, and the resources organizations think they need to care for a population.

What happened: - Equipo Health highlighted research showing that risk assessment based only on diagnosis codes can miss hidden patient risk. - The company said its AI-powered clinical intelligence tools are designed to identify potential coding gaps, undocumented conditions, and other relevant clinical details. - Equipo framed the issue as a broader challenge for healthcare organizations trying to make risk assessment more representative of the patients they serve.

The details: - A 2018 study in The American Journal of Managed Care reviewed 1,596 men treated at two Veterans Affairs medical centers. - Researchers compared EHR problem lists with manual review of clinical notes and found the problem lists captured only 23% of congestive heart failure cases, 42% of chronic obstructive pulmonary disease cases, and 46% of diabetes cases. - Comorbidity scores built from clinical notes predicted long-term mortality, while scores based on problem lists did not. - The study concluded that EHR problem-list data can distort patient case mix and should be validated before being used for risk adjustment. - A 2025 study in Computational and Structural Biotechnology Journal analyzed 10.6 million free-text EHR entries from 102,525 patients in Finland. - Using natural language processing, researchers identified 31,987 instances of falls in clinical text versus 4,090 through diagnosis codes. - The same study identified 7,059 instances of incontinence in clinical text versus 3,873 through diagnosis codes. - Conditions found in clinical notes showed stronger associations with mortality than conditions found through diagnosis codes alone. - A 2022 Health Affairs study on Medicare risk adjustment for Alzheimer's disease and related dementias found 22.7% of beneficiaries identified through clinical review lacked a matching dementia diagnosis in claims data. - The same study found 34% of beneficiaries with a dementia code did not have the condition confirmed through clinical review. - Medicare spending for beneficiaries with clinically identified dementia that was not captured in claims was $14,619 higher per beneficiary than for those correctly identified in claims. - Equipo said its Foresight AI Risk Analyzer Engine reviews unstructured clinical documentation to surface possible gaps and relevant information for risk assessment. - Equipo said its platform combines clinical, operational, and patient engagement data to support population health management, risk analysis, care coordination, reporting, and referral coordination.

Between the lines: - The research suggests the main problem is not a lack of data, but a lack of connection between structured codes and the broader clinical record. - AI can surface likely gaps at scale, but clinical judgment and coding validation still determine whether those findings should change a risk score. - For health systems, the issue is both financial and operational: a missed condition can affect payment, performance measurement, and which patients get follow-up.

What's next: - Healthcare organizations are likely to keep expanding risk review beyond diagnosis lists to include notes, assessments, medical history, and encounter documentation. - Equipo is positioning its healthcare intelligence platform as a way to connect risk analytics with care management and patient outreach. - The company says this approach can help organizations review possible documentation gaps, support population assessment, and prioritize care more effectively.

The bottom line: - Diagnosis codes remain essential, but they do not always tell the full story. - Healthcare organizations that want more accurate risk adjustment may need a broader, AI-assisted view of the clinical record.