Chart abstraction in RADV is the careful review of a patient’s medical records to identify all diagnoses that should affect the Risk Adjustment Factor, along with when those conditions started or were active and the evidence that supports them. Abstractors comb through provider notes, test results, imaging, treatment plans, and other documentation to confirm that a condition is present during the measurement period and that there is concrete evidence linking it to the appropriate risk-adjustment category. This process ensures the diagnoses used to calculate the RAF are properly supported, reproducible, and auditable, which helps prevent payment errors and maintains accuracy in risk-adjusted payments. While clinicians contribute to the documentation, chart abstraction relies on trained abstractors or coders who pull and verify the information from the medical record, rather than simply assigning RAF from diagnosis lists or conducting patient interviews.

The Medicaid Integrity Program is a federal effort established under the Social Security Act to protect the Medicaid program from improper payments. Its main purpose is to identify, deter, and recover fraud, waste, and abuse by conducting audits and reviews of Medicaid providers, including managed care entities, using data analysis andProvider billing review. It supports states with audits, training, and guidance, coordinating with CMS and other program integrity activities to strengthen compliance and oversight. This focus is specific to Medicaid and differs from initiatives like hospital expansion funding, veterans’ health services, or Medicare Part D oversight, which fall outside MIP’s scope.

In RADV, two sources of evidence are used to validate diagnoses on submitted claims: encounter data and chart documentation. Encounter data are the data submitted in claims about patient encounters; chart documentation is the actual medical records used to validate diagnoses. The encounter data provide billing-level information—the codes, dates of service, and provider details that feed into risk scores—while chart documentation offers the clinical proof that those codes are accurate and supported by the patient’s records, notes, tests, and findings. If the chart documentation does not support a diagnosis listed on the encounter data, the diagnosis may be disallowed or adjusted.

In CMS risk adjustment, the risk score is shaped by the set of diagnosed conditions captured for the year plus whether chronic conditions from the prior year are recaptured in the current year. When additional conditions are reported, the model adds the corresponding condition weights (HCCs), which increases the overall risk score because more predicted costs or resource use are inferred. If chronic conditions from the previous year are not recaptured in the current year, those conditions no longer contribute to the current calculation, so the risk score can decrease. Demographics and other factors can influence the score as well, but the essential idea is that more diagnosed conditions raise the score, and failing to capture prior chronic conditions lowers it.