The State of RCM in 2026: What Revenue Cycle Leaders Are Watching
The State of RCM in 2026: What Revenue Cycle Leaders Are Watching
Two years after the Change Healthcare cyberattack paralyzed clearinghouse traffic for weeks, the revenue cycle industry is still recalibrating. New payer behaviors, evolving regulatory obligations, and an ongoing talent shortage are compressing margins for billing departments and RCM vendors alike. Here is a grounded look at what experienced RCM leaders are tracking heading into the second half of 2026.
Clearinghouse Resilience After the Change Healthcare Attack
The February 2024 ransomware attack on Change Healthcare — then processing roughly 40 percent of U.S. health insurance claims — remains the sharpest operational shock the industry has absorbed in decades. Practices that depended on a single clearinghouse discovered their exposure the hard way: delayed claims, frozen eligibility checks, and cash-flow gaps that lasted weeks.
The lasting shift is a move toward redundancy. Large health systems and mid-size RCM shops accelerated dual-clearinghouse contracts, routing commercial claims through one vendor and government payers through another. Smaller practices, which historically accepted single-vendor convenience, are now asking harder questions during contract renewals. Vendors offering real-time failover — automatically rerouting 837 transactions if a primary clearinghouse goes dark — have a credible selling point that did not exist before 2024.
The attack also accelerated conversations about cybersecurity posture in billing departments specifically. Payer portals, clearinghouse credentials, and remittance data are all targets. Expect more third-party security assessments baked into RCM contracts by 2027.
AI in Claim Scrubbing and Autonomous Coding: Mature vs. Hype
AI-assisted claim scrubbing is genuinely mature. Predictive edit engines that flag likely CO-16 (missing/invalid information) and CO-97 (payment adjusted because the benefit for this service is included in the payment for another service) rejections before submission have been commercially available for years. The reduction in avoidable first-pass rejections is measurable, and most enterprise-class practice management platforms now include some version of it.
Autonomous coding — where a model reads clinical documentation and assigns CPT and ICD-10 codes without a human in the loop — is a different story. Vendors are deploying it selectively in high-volume, low-complexity settings: radiology reads, laboratory panels, straightforward evaluation and management encounters. For complex specialties, or anywhere MDM documentation drives code selection (99213 vs. 99214, for example), fully autonomous coding carries audit exposure that most compliance officers are not ready to accept.
The honest framing for 2026: AI handles the commodity work and flags the edge cases. It does not replace the certified coder who reviews those flags. RCM leaders buying AI tools should demand audit trail transparency — which encounters were auto-coded, with what confidence score, and which were escalated to human review.
The No Surprises Act GFE Workload
The No Surprises Act’s good-faith estimate (GFE) requirement for uninsured and self-pay patients has been in effect since January 2022, but enforcement and workflow maturity are still uneven. For RCM teams, the operational burden is real: scheduling staff must identify self-pay encounters early, collect service and cost data from all co-providers when applicable, and deliver the GFE before the scheduled service.
The convening provider requirement — where the primary provider must coordinate cost estimates from co-providers (anesthesiologists, assistant surgeons, facility fees) into a single bundled estimate — remains the hardest piece to operationalize. Many practices are still handling this with manual outreach and spreadsheets.
Payers have also introduced their own complication: price transparency and GFE workflows need to align with the Advanced EOB requirements that apply when a patient has coverage. The two frameworks share design logic but differ in technical spec and timing. Billing departments with a clean self-pay workflow are not automatically ready for the covered-patient version.
Staffing, Payer Behavior, and What to Watch Next
The staffing shortage in medical billing has not resolved. Experienced coders and AR follow-up specialists remain difficult to hire and retain, particularly in markets where health systems compete directly with remote-first RCM vendors. Offshore teams remain a functional option for many tasks, but language-sensitive denial management and peer-to-peer clinical reviews are hard to offshore cleanly.
On payer behavior: three trends stand out.
Silent downcoding — where a payer’s adjudication algorithm automatically reduces a submitted E/M level without issuing a formal denial — has become more common. An 835 remittance shows CO-45 (charges exceed fee schedule) rather than a CO-97 adjustment, but the net payment reflects a lower-level code. Detecting this requires systematic reconciliation of submitted vs. paid procedure codes on every ERA, not just dollar amounts.
Prepay clinical edits by commercial payers, particularly UHC and Aetna, have expanded. Certain procedure codes now route to a pre-payment clinical review queue that can add 30 to 60 days to adjudication. Medicare’s Targeted Probe and Educate (TPE) program operates similarly on the government side.
AI-driven denial generation is the counterpart to AI-assisted claim scrubbing. Payers are deploying their own models to identify claims with documentation patterns associated with upcoding or medical necessity gaps. Appeals need to address the specific edit logic, not just general documentation quality.
For a broader industry view, the Healthcare Financial Management Association (HFMA) publishes regular benchmarking and regulatory updates that RCM directors should have on their reading list.
If your team is navigating any of these pressures — clearinghouse strategy, denial pattern analysis, or coding compliance — reach out to discuss how we can help.
This post was drafted by AI and reviewed by our editorial team. Last updated 2026-05-30.