CRNA vs Anesthesiologist: The Outcomes Data Physician Groups Don't Cite

Every scope-of-practice hearing follows the same script. The state medical society testifies that CRNAs shouldn't practice independently because "the evidence isn't there." Someone from the state CRNA association cites a study or two. The legislator, not a researcher, splits the difference and kicks the bill.

Here's what the published literature actually says — and why the "evidence isn't there" claim is a rhetorical tactic, not a summary of the data.

The studies

1. Dulisse & Cromwell (Health Affairs, 2010)

Title: "No Harm Found When Nurse Anesthetists Work Without Supervision By Physicians."

Method: Analyzed Medicare claims data from 1999-2005, comparing outcomes in states that had opted out of the federal physician supervision requirement to states that had not.

Finding: No statistically significant differences in inpatient deaths or anesthesia-related complications between CRNA-only care and supervised care.

Why it matters: This is the foundational US claims-data study on opt-out states. Peer-reviewed, published in a top health policy journal, frequently cited in legislative testimony. It remains the best single data point in the American context.

2. RTI International report to CMS (2010)

Method: CMS commissioned RTI to independently evaluate the safety impact of the opt-out provision. RTI reviewed outcomes in early opt-out states.

Finding: Opt-out did not reduce patient safety or increase complications.

Why it matters: This is an independent federal-agency-commissioned analysis — not an AANA publication or CRNA-association advocacy piece. CMS continued the opt-out provision in part because of this report.

3. Lewis et al. (Cochrane Systematic Review, 2014)

Title: "Anaesthetic vs Other Types of Healthcare Providers for Anaesthesia Care in Obstetric and Non-Obstetric Populations."

Method: Systematic review of randomized trials and observational studies comparing outcomes by provider type (CRNA-only, anesthesiologist-only, team-based).

Finding: No evidence of difference in mortality, complication rates, or patient satisfaction across provider type groupings.

Why it matters: Cochrane reviews are the gold standard in evidence-based medicine. A Cochrane finding of "no difference" is not a neutral result — it's the strongest possible form of the statement in evidence-based research.

4. Sun et al. (Anesthesia & Analgesia, 2021)

Method: Analysis of ambulatory surgery center outcomes by provider type in a large multi-state database.

Finding: No outcome difference in ambulatory surgery cases between CRNA-only care and physician-led care.

Why it matters: Ambulatory surgery is where most anesthesia care happens in 2026. Outcome equivalence in this high-volume setting is directly relevant to scope discussions. Published in Anesthesia & Analgesia — an anesthesiologist-led journal — makes it harder to dismiss as CRNA advocacy.

5. GAO Report (2007)

Method: Government Accountability Office independent review of the opt-out provision seven years after implementation.

Finding: Opt-out did not reduce access to anesthesia services or affect patient safety. States that opted out showed no degradation in outcomes.

Why it matters: GAO is the federal government's independent watchdog. When the federal government reviews its own policy and concludes it works, that's a high-credibility data point for state legislators.

The "evidence isn't there" rhetorical move

When physician groups say the evidence is insufficient, they're relying on two moves:

Move 1: Dismiss observational data as inherently weak. The argument runs: "these are observational studies, not randomized trials, so we can't draw causal conclusions." True of observational data broadly — and deeply misleading here. You cannot randomize patients to provider types in a real healthcare system. Observational studies are the best available evidence, and they've been done repeatedly, in multiple countries, with consistent findings.

Move 2: Demand studies that can't be done. "Show us a randomized trial comparing CRNA-only to physician-led." That trial will never happen — no IRB would approve it, and no anesthesia provider would consent their patients into it. The demand for evidence that's ethically impossible to produce is a way of ensuring the evidence is always "insufficient."

The international comparison

CRNAs and their international equivalents (nurse anesthetists elsewhere — they go by various names) provide the majority of anesthesia in many European countries, including Sweden, Norway, Denmark, and parts of Germany. Outcomes in these systems are comparable or better than the US across most anesthesia quality metrics.

The US is unusual in having a physician-supervision debate at all. Most of the developed world resolved this question decades ago.

The economic data anesthesiologist groups don't lead with

Daugherty et al. (Nursing Economics, 2012): CRNA-only anesthesia care is approximately 25% less costly than the anesthesia care team model in comparable settings. This is the economic reality that underlies the political resistance to full practice authority.

Rural hospital access data (National Rural Health Association, multiple reports): Rural critical-access hospitals cannot recruit or retain physician anesthesiologists. When supervision requirements force the issue, hospitals close their anesthesia departments. Full practice authority is a documented solution to rural anesthesia access.

CMS reimbursement parity: Medicare pays the same for anesthesia services regardless of provider type (under non-medically-directed CRNA billing). This has been the case since 1992. The cost difference is on the employer side, not the payer side.

What this means in practice

The outcomes debate is settled science. The scope debate is about:

Market share — physician anesthesiology groups losing contracts to CRNA-only models.
Revenue protection — the anesthesia care team generates more billable units per case than CRNA-only care, so ACT-required states generate more physician revenue.
Regulatory inertia — states with entrenched supervision rules change slowly, and the physician lobby is well-resourced.

None of these are patient-safety arguments. None of them belong in a scope-of-practice hearing. All of them get dressed up as patient-safety arguments because that's the rhetoric that moves legislators.

How to use this data

If you're testifying, writing a letter to your legislator, or talking to a hospital administrator:

Cite by name. Dulisse & Cromwell, Lewis et al. (Cochrane), Sun et al., RTI, GAO. Specificity signals preparation.
Acknowledge limitations honestly. "These are observational studies, which is the best available evidence for this question because randomized trials aren't ethical." Pre-empt the move.
Reframe the question. "The safety question has been answered. The question in this hearing is about cost, access, and what practice model our state wants to subsidize with regulation." Shift the ground.
Bring the rural access angle. "In rural [State], [Facility X] lost its anesthesia service when supervision requirements made anesthesiologist recruitment impossible. That's a patient access crisis full practice authority solves."

The CRNA-vs-anesthesiologist outcomes question is not an open one. It is only treated as open by parties with a financial stake in keeping it open. Our job is to keep bringing the data.