Oracle Health’s New AI Tools Target Coding and CDI
Oracle Health unveiled five AI revenue cycle modules at its September 2026 summit, including autonomous coding and real-time CDI tools built to catch gaps before claims go out.
·
-
-
Maverick Medical AI Moves CDI to the Point of Care
Maverick Medical AI’s Sept. 8 launch puts real-time physician queries at the point of care, testing how far CDI can shift left of the retrospective chart review.
·
-
ACDIS/AHIMA 2026 Query Guidelines Tackle AI Queries
ACDIS and AHIMA just updated the query practice guidelines CDI programs live by — and for the first time, they spell out rules for AI-generated queries.
·
-
AHRQ Funding Collapse Puts CDI’s Evidence Base at Risk
AHRQ’s near-halt on grantmaking is starving the sepsis and patient-safety guidance CDI teams depend on daily — coding leaders need a plan before it breaks.
·
-
Clinical Validation Denials: Where Agentic AI CDI Fits
Payers increasingly split DRG denials into coding downgrades and clinical validation disputes. Here’s how agentic AI helps CDI teams tell the two apart.
·
-
CMS’s Malnutrition Mandate Raises Coding Audit Stakes
CMS is mandating malnutrition quality reporting by 2028 while OIG audits keep finding malnutrition codes unsupported. Here’s what coders need to know.
·
-
OIG Audit Flags $15.2M in Sacroiliac Injection Coding Errors
A new OIG audit found Medicare improperly paid $15.2M for sacroiliac joint injections, nearly a quarter misclassified as therapeutic instead of diagnostic.
·
-
AI Surgical Coding Accuracy: What Procode’s Study Shows
A new peer-reviewed study found Procode’s AI hit 86.7% surgical coding accuracy on operative reports, more than double GPT-5 and human auditors combined.
·
-
Agentic AI Brings Real-Time CDI to Behavioral Health Coding
Behavioral health coders face rising audit risk and thin margins. Agentic AI now flags documentation gaps in real time, before a note or claim goes out.
·
-
LLM Hallucinations in Medical Coding: What emtelligent’s Study Found
A new study shows general LLMs fabricate nearly 1 in 3 medical codes. emtelligent’s clinical AI hit 89.85% F1 with zero hallucinations—here’s what coders need to know.
·