The autonomous mid-cycle keeps the patient story intact

The middle revenue cycle has one job: carry an accurate account of care from documentation through coding and onto the claim.

Today, that story passes through separate teams, systems and review queues. Each handoff creates another chance for context to be lost.

Health system leaders see both the problem and the opportunity. In this report, 45% say revenue cycle accuracy and efficiency are where AI will deliver the greatest value, more than double any other response.

That makes sense. The record shapes reimbursement, quality reporting and risk adjustment. When it misses the complexity of the care delivered, the consequences move downstream.

Earlier automation solutions worked best in simpler outpatient facility coding, where records are shorter and patterns more predictable.

Inpatient care is different. A single stay can generate more than 60 documents, with important details spread across notes, labs, orders and discharge summaries. Rules and keyword searches can’t reliably assemble that clinical picture.

Large language models (LLMs) can. They can interpret the full record in context, connect evidence documented at different points in the encounter and be fine-tuned to a health system’s patient population, clinical criteria and documentation practices. That advance makes broader autonomy finally possible. 

The longer-term opportunity is one LLM-based platform that follows the patient story across the entire mid-cycle: inpatient and outpatient, clinical documentation integrity and coding. During the encounter, it can surface documentation opportunities. After discharge, it can code encounters within approved confidence thresholds. Pre-bill optimization can review the broader population and catch what remains. Every result stays tied to the source record, and each health system decides where autonomy has earned its trust.

Some cases will still need experienced professionals. The difference is that their time can go to the cases where judgment matters, instead of reviewing work that the technology has already proven it can handle. 

The destination is not just a faster version of today’s revenue cycle. It is a fundamentally different operating model, where AI handles an increasing share of the work and healthcare professionals can focus their expertise where human judgment matters most.

Learn more about revenue cycle and AI in healthcare in HFMA’s The Hospital of the Future Part II report.