Are anesthesiologists adopting AI pre-op tools?

Yes, particularly in private-practice groups and in groups using the Anesthesia Care Team (ACT) model with CRNAs or CAAs. Adoption is faster in ASC and office-based settings than in academic hospital practice. The driver is the same as for CRNAs: documentation time recovery and same-day cancellation prevention, both with directly measurable ROI.

How does AI pre-op fit into the ACT model?

In the Anesthesia Care Team model — anesthesiologist medical direction with CRNAs or CAAs delivering — AI pre-op tools fit naturally: the CRNA or CAA runs the initial clearance review with AI assistance, the anesthesiologist reviews the AI's verdict and the provider's sign-off, and any flagged complexity escalates to the anesthesiologist's bench for additional workup. The audit trail (who reviewed what, when, with what AI output) supports the medical-direction billing requirements.

What's the liability exposure for an anesthesiologist using AI pre-op?

The same as without AI — the licensed clinician is responsible for the clinical call. AI does not add liability when used as decision-support (provider reviews and signs every output). AI may reduce liability when it catches misses that would have caused cancellations or adverse events. The risk is using AI tools that pitch themselves as autonomous decision-makers and signing charts without review — that creates exposure regardless of the AI's accuracy. Choose tools framed as decision-support, review every output, sign every chart.

What Anesthesiologists Should Know About AI Pre-Op Tools (2026)

Five questions every anesthesiologist should ask before adopting an AI pre-op tool in their practice — from a CRNA-founder who works alongside anesthesiologists every day.

I'm a CRNA, but most of the anesthesia work in this country happens under medical direction by an anesthesiologist. The AI tools your CRNA and CAA team members are starting to use need to fit your workflow, your billing, and your liability posture. Here are the five questions I'd ask.

1. Where does the verdict come from?

The AI's clearance verdict (Cleared, Cleared with Conditions, Not Cleared) needs traceable rationale. Ask: can you see the citations? Can you see which clinical rule fired? Can you tell whether the verdict comes from documented findings or from inference the AI made? Tools that show their work survive a chart review. Tools that don't are black boxes you sign your name to.

2. Does the audit trail support medical direction?

If your group bills under medical direction, the audit trail needs to capture who reviewed what, when, with what AI output. The CRNA's sign-off and your review need to be separable in the record. Ask the vendor to show the audit log on a real case. If they can't, the tool is not ready for ACT-model billing.

3. How does the tool handle PHI?

Ask for the data flow diagram. Identifiable PHI should not leave the HIPAA-covered system without explicit de-identification. If the vendor uses an external LLM API (which most do, including MyPreOp.ai), confirm that PHI is de-identified before the API call. Confirm the BAA covers every link in the chain. Tools that can't answer this question in detail are tools you should not use.

4. What's the override surface?

When you disagree with the AI's ASA classification or flag, how do you override it? Is the override one click? Is it captured in the audit? Can the override note explain why? AI tools where the human override is buried two menus deep have inverted the priority: they've made human judgment the friction and AI judgment the default. Walk.

5. How is it priced for a group?

Per-provider monthly subscription is the cleanest model for ASC and office-based groups. $59/mo per provider unlimited, with a group tier at $499/mo for up to 10 providers, is what I'd look for. Enterprise SaaS at $20K/year with per-case overage charges fits hospital procurement but adds friction in private practice. Look for a tool you can subscribe to without a sales call.

My recommendation

Try one tool on one weekend's clearances. Run it alongside your existing workflow. Compare the AI's output to what you would have caught yourself. If the tool has rationale, audit trail, PHI handling, override surface, and fair pricing — adopt it. If any of the five are weak, walk.

MyPreOp.ai is one option built around all five. See the platform overview or the buyer's framework.