The RCM Staffing Shortage Will Not Resolve Itself

Sixty-three percent of healthcare providers report RCM staffing gaps. That figure has held steady for three years, through hiring pushes, compensation increases, and vendor augmentation arrangements that added headcount without changing the underlying dynamic. The shortage is not a pipeline problem waiting for the labor market to correct. It is a structural condition produced by a revenue cycle architecture that requires humans to do work that should not require humans at all.

Understanding the difference matters, because the two diagnoses lead to completely different responses. One leads to another round of hiring. The other leads to a fundamentally different operating model.

Why Hiring Against the Shortage Makes It Worse

The work that consumes RCM staff capacity in most health systems is not judgment work. It is assembly work. Before a prior authorization can be submitted, someone gathers the clinical record from the EHR, identifies the payer's current criteria from a separate system, verifies coverage status from a third, and assembles a submission packet by hand. Before a denial can be worked, someone reconstructs the context of the original encounter, cross-references the payer's denial reason against their current policy, and drafts a response.

None of that work requires clinical judgment or payer expertise. It requires access to information that exists across disconnected systems and a person willing to assemble it. Physicians and their teams spend an average of 14.6 hours per week on prior authorization alone, according to the AMA's 2025 survey. That is not a measure of authorization complexity. It is a measure of how long manual context assembly takes at scale.

When organizations respond to staffing gaps by hiring, they add capacity to perform that assembly work faster. The bottleneck does not move. It scales with headcount, breaks with turnover, and gets harder every year as payer rules multiply and plan variation grows.

The Turnover Problem Hidden Inside the Staffing Problem

Healthcare RCM turnover rates run significantly above the national average, and every departure takes institutional knowledge with it. A billing specialist who has spent three years learning the quirks of a specific payer's authorization requirements, the documentation patterns that pass a particular insurer's medical necessity review, the appeal language that works for a given procedure, carries that knowledge out the door when they leave.

In a manual operating model, that knowledge lives in people. When the people leave, the organization loses the knowledge and then spends months rebuilding it through trial and error on live claims. The denial rate climbs. The appeal win rate drops. A new hire starts the learning curve from the beginning, and the cycle repeats.

This is the compounding cost of a staffing-dependent revenue cycle that never appears as a line item. It shows up as a denial rate that bounces back after every improvement effort, as AR that ages past the point of easy recovery, and as write-offs that represent revenue nobody had the capacity to chase.

Flow addresses this directly. When prior auth, coding review, denial management, and payment reconciliation run on a connected data foundation where clinical records, payer policy, and claims status exist in one place, the institutional knowledge that currently walks out with departing staff is embedded in the system rather than carried by individuals. Flow's AI agents handle the rules-based majority of revenue cycle work proactively, and certified RCM experts handle the judgment calls, payer relationships, and edge cases that require human expertise. Neither operates blind. The system works regardless of who is in the seat.

What the Alternative Operating Model Looks Like

The organizations reducing their dependency on RCM headcount are not doing it by eliminating their revenue cycle teams. They are doing it by eliminating the assembly work that consumes those teams, and redirecting skilled staff toward the work that actually requires them.

When agents handle the rules-based majority, staffing requirements stop scaling linearly with volume. Adding providers or locations does not require adding billing staff at a corresponding rate, because the agents absorb the volume increase without proportional headcount growth. Flow works with any EHR and does not require organizations to replace or restructure their existing systems to deploy. The operating model changes. The infrastructure underneath it does not have to.

The Learning Problem That Headcount Cannot Fix

There is a second dimension to the staffing problem that hiring cannot address regardless of how many people are added. In a manual operating model, the lesson from a denial never reliably reaches the authorization decision that caused it. A coder who identifies a documentation pattern that triggers denials has no mechanism to push that insight upstream to the scheduling team. Each workflow learns in isolation, and the same root causes recur.

Through the Outcome Intelligence Loop™, every denied claim, coder override, appeal result, and recovered dollar feeds a shared intelligence layer automatically. Prior auth outcomes inform coding decisions. Coding patterns tighten claim submissions. Denial trends close back to access. The system learns from every outcome without requiring a human to translate the lesson across workflow boundaries.

That compounding effect is what a staffing strategy cannot produce. More people working the same disconnected workflows produce more output of the same kind. A connected system produces a different kind of output entirely: fewer denials entering the queue, because the conditions that created them were addressed upstream before the claim was filed.

The Right Question for 2026

The staffing shortage will not resolve through hiring because the shortage is not fundamentally about people. It is about an operating model that requires people to perform work that a connected system should handle. Health systems that diagnose the problem correctly in 2026 will spend the next three years building operating leverage. Those that respond with another hiring cycle will spend those years managing the same bottleneck at increasing cost, on margins that averaged 1% in 2025 and have no room to absorb it.

The question is not where to find more RCM staff. It is what the revenue cycle should still require staff to do. A 30-minute briefing on how Flow's agents and experts divide that work in production environments is the fastest way to answer it concretely, before a budget cycle closes around the wrong solution.

Book a demo.

‍