That Is Not My Job

A hospital lawyer told me he could not recommend changes to prevent the next case. He was right about one thing: nobody had given him that job. That is precisely the problem.

Some years ago I was working with a hospital lawyer on a malpractice case.

He was competent, thorough, and honest.

We had reviewed the case together in detail.

The clinical picture was clear.

Something had gone wrong, and the pathway by which it went wrong was identifiable.

We both understood what it was.

I turned to him and said: given what we have found, what do you recommend we change to prevent this from happening again?

He looked at me with an expression I have thought about many times since. Not defensive. Not evasive. Genuinely puzzled.

“That is not my responsibility,” he said. “I am not a doctor. It is not my place to make clinical recommendations.”

He was correct on every count.

It was not his responsibility.

He was not a doctor.

It was not his place.

And that, I realized, is one of the central problems with how American medicine handles adverse outcomes. When the case is over, nobody whose job it is to understand what went wrong has a job that includes recommending what should change. The lawyer protects the institution. The risk manager documents the event. The peer reviewer assesses the individual clinician. The M&M conference discusses the case. And then everyone returns to their lane.

Prevention is in nobody’s lane.

Each Person Does Their Job. Nobody Does the One That Matters.

This is worth sitting with. The hospital lawyer in that room was not indifferent to the outcome. He was a professional doing exactly what he had been hired to do: assess liability, manage exposure, protect the institution’s legal position. Recommending clinical practice changes was genuinely outside that mandate. If he had made such recommendations, he would have been operating outside his expertise and potentially creating additional liability for his client.

The risk manager’s job is documentation and compliance. The quality officer’s job is accreditation. The peer reviewer’s job is to assess whether the clinician met the standard of care. The malpractice insurer’s job is to settle or defend. Each of these roles is real, each is occupied by capable people, and none of them has prevention as its primary output.

The NTSB understood this problem sixty years ago. Its founding mandate was not to assign blame, not to manage liability, not to protect any party’s legal position. It was to determine probable cause and issue safety recommendations. Prevention was not a byproduct of the NTSB’s work. It was the only product. Every other output -- the investigation report, the probable cause determination, the factual reconstruction -- exists in service of the safety recommendation.

In the room where I sat with that lawyer, there was no NTSB equivalent. There was no person or body whose only job was to ask: what changes would prevent the next case?

The Lawyer’s Constraint Is a System Design

I want to be precise about what the lawyer’s response revealed, because it is easy to hear it as an excuse and miss what it actually was: an accurate description of how the system is built.

The American hospital response to an adverse outcome is organized around two functions: legal defense and regulatory compliance. Both are legitimate. Both are necessary. Neither is designed to produce prevention. Legal defense is designed to minimize liability. Regulatory compliance (the Joint Commission root cause analysis, the peer review process, the sentinel event report) is designed to satisfy an accreditor. These are different goals from prevention, and optimizing for them does not automatically produce prevention as a side effect.

In fact, optimizing for legal defense actively works against prevention in one specific way: it creates strong incentives not to produce a written document that clearly identifies what went wrong and what should change. Such a document, if not properly protected, becomes discoverable evidence. The instinct of every legal team I have ever worked with is to produce the minimum documentation necessary to satisfy the accreditor while preserving the maximum possible ambiguity about causation.

The NTSB solved this problem legislatively. Under 49 U.S.C. Section 1154(b), no part of an NTSB accident report may be admitted into evidence in a civil lawsuit arising from the accident. Congress made an explicit choice that safety learning was worth more than litigation ammunition. That protection is why NTSB investigators can write exactly what they found without worrying about what plaintiff’s attorneys will do with it.

Medicine has never made that choice. And so the lawyers remain in their lane, the investigators remain in theirs, and the written record of what went wrong remains as vague as possible. Prevention waits.

Who Should Have That Job

The answer the NTSB model provides is: a person or body whose only job is prevention, who is independent of all the parties with a financial interest in the outcome, and whose work product is a safety recommendation addressed to whoever has the power to make the relevant change.

That body does not currently exist in American obstetrics.

What should exist is an Obstetric Patient Safety Investigation Board, independent of ACOG, independent of hospital systems, independent of malpractice insurers, and independent of the Joint Commission, that investigates every maternal death and every case of birth-related brain injury, produces a public report, and issues safety recommendations with named recipients, measurable outcomes, and tracked implementation.

The hospital lawyer would not be asked to sit on that body. Neither would the risk manager or the peer reviewer. Those people would continue doing their jobs.

The investigation board would do a different job: the one nobody is doing now.

Until that body exists at the national level, ObGyn departments can build a local approximation of it. The investigation checklist at tools.obmd.com/ob-case-review walks a review team through the same phases the NTSB uses: establish the facts, review the record, assess each decision against the standard of care, examine communication, examine system factors, determine causation, and generate prevention recommendations addressed to whoever has the authority to make each specific change. The AI analysis that runs on the entered data produces a Safety Recommendation Report in NTSB format: specific actions, named responsible parties, deadlines, and measurable metrics.

That report gives the department chair what I did not have in that room: a document whose explicit purpose is to answer the question the lawyer could not answer.

What Needs to Change at the System Level

The lawyer’s response pointed to three specific system failures that no individual institution can fix alone.

The first is the absence of a mandated prevention function. Every hospital has a legal defense function. Every accredited hospital has a sentinel event reporting function. No hospital is required to have a function whose sole purpose is to produce safety recommendations from adverse outcome investigations and track their implementation. ACOG should require it. The Joint Commission should require it. Until one of them does, it will remain optional, which means most institutions will not have it.

The second is the absence of litigation protection for candid investigation findings. As long as a clearly written investigation report identifying a specific practice gap can be subpoenaed and used against the institution in court, the legal team’s instinct will be to prevent that document from existing. Congress needs to extend to medical adverse outcome investigations the same protection that exists for NTSB aviation reports. Without that protection, candor is professionally irrational.

The third is the absence of a prevention-focused mandate anywhere in the existing system. The lawyer’s job is defense. The risk manager’s job is documentation. The peer reviewer’s job is individual assessment. None of these roles has “prevent the next case” as its primary or even secondary function. That function needs to be explicitly assigned, resourced, and held accountable. Right now, it belongs to everyone in principle and to no one in practice.

Conclusion

I do not blame that lawyer.

He did exactly what he was hired to do, and he was honest with me about what he could and could not contribute.

That honesty was more valuable than a vague promise to “look into it” that would have produced nothing.

What I learned from that conversation is that the gap in American obstetric patient safety is not primarily a gap in competence or in caring.

The clinicians involved in most adverse outcomes are competent.

The lawyers are competent.

The risk managers are competent.

The gap is structural. The system has been designed to do everything except the one thing that would actually reduce the rate of preventable adverse outcomes: assign clear responsibility for prevention to a specific person or body with the independence, the authority, and the mandate to produce it.

The NTSB is not staffed by people who are smarter than hospital lawyers or risk managers or department chairs. It is structured differently. Its investigators have one job. Their findings are protected from litigation. Their recommendations are public and tracked. That structure is what produces results, not the intelligence of the people inside it.

We know how to build this. We have had the model for sixty years. The question is whether American obstetrics is willing to adopt it, or whether we will continue producing adverse outcomes at the current rate while each of the responsible parties continues doing their job and nobody does the one that matters.

The investigation tool

—> ObGyn Adverse Outcome Investigation Checklist -- NTSB methodology, AI-powered Safety Recommendation Report, free. tools.obmd.com/ob-case-review