Connecting AI Islands Across the Revenue Cycle

Hi and Happy Tuesday.

In our previous NL, we covered Q1 and the rollout of plug-in AI across the revenue cycle.

This week, we are focused on Q2 and connecting those tools so work moves in a straight line, with no manual handoffs and no lost information.

1. Prior authorization automation at scale

The Burden:

Physicians handle ~40 prior authorizations per week, spending 12-13 hours on them. 29% report serious adverse events linked to delays.

90%+ Of physicians say prior auth delays care

The window: A recent Humana announcement confirms plans to eliminate about one-third of prior authorizations for outpatient services by January 1, 2026, and to publish authorization metrics while speeding up decisions.

That creates a window in Q2 to replace manual “chasing” with automated chains that:

• Run eligibility and benefit checks up front.

• Auto-assemble clinical documentation

• Monitor status and push updates to worklists

Q2 Target 1:

Build automated chains for 2-3 high-volume service lines

Track: Measure reductions in staff time per authorization and delays tied to missing information.

2. AI-driven worklist triage and routing

The Bottleneck:

76% of RCM leaders say denials management is their most time-consuming task

41% of providers face denial rates ≥10%

The problem: Queues sorted by date, not recovery potential. High-value accounts wait behind low-value ones.

• Rank by collectability, age, and payer behavior

• Create separate queues for high-dollar accounts

• Flag accounts near the timely filing limits

Q2 Target 2:

Deploy AI queues for denials and AR teams

Track: Dollars resolved per FTE. Average resolution time vs 2024 baseline

3. Payment posting and underpayment detection

The Leak:

Providers lose 1-5% of net revenue to underpayments—up to $99K annually per provider

1-11% of net revenue lost to underpayments and manual errors

The fix: Document AI and rules engines now parse remits, match contracts, and flag variances at scale.

• Automate remittance file parsing

• Match line items to contract terms

• Flag payment variances for follow-up

Q2 Target 3:

Automate posting and checks for the highest-volume payers

Track: Underpayment variance identified and recovered quarterly

4. Patient cost estimates and financial counseling assistants

The Trend:

Patient collection rates dropping: 54.8% (2021) → 47.8% (2023) → 34.4% (2024 commercial).

34.4% Commercial patient collection rate in 2024, (down from 37.6% in 2023)

The solution: AI generates clearer estimates and guides staff through payment conversations.

• Produce accurate out-of-pocket estimates pre-visit

• Guide staff through payment plan scripts

• Explain coverage and realistic options

Q2 Target 4:

Deploy AI estimates in 1-2 high-balance service lines

Track: Point-of-service collections · Bad-debt write-offs · Charity classifications

The Connected Flow

Prior Auth → Worklists → Posting → Estimates

One continuous workflow. No manual handoffs.

In our next newsletter, we’ll give you a Q3 roadmap.

Dino Gane-Palmer