Running an M&M Meeting That Changes Something

The morbidity and mortality meeting is the oldest quality-improvement instrument in surgery and, in most departments, the least effective. The usual form is familiar: a registrar presents six cases in forty minutes, the consultant who did the operation explains why it was unavoidable, someone says "learning points" without naming any, and the meeting ends because the list starts at half past.

That meeting is not useless — it is worse than useless, because it consumes the department's only protected hour for examining harm and returns nothing, while allowing everyone to believe the harm has been examined.

The good news is that the difference between a performative M&M and a genuinely useful one comes down to a handful of structural choices. This post covers them. It sits within our clinical governance series.

What the meeting is actually for

Three purposes, which pull in slightly different directions and should be named explicitly:

1. Learning — extracting generalisable lessons from individual cases.
2. Assurance — demonstrating that the department reviews its outcomes, which is a governance requirement almost everywhere.
3. Support — helping a colleague who has had a bad outcome, which is not the same as scrutinising them.

Most dysfunctional M&Ms have quietly collapsed these into a fourth purpose nobody will admit to: allocating blame, or its mirror image, collectively agreeing that nothing could have been done. Both prevent learning, and the second is by far the more common in surgical departments.

The failure modes

The defensive recital. The operating surgeon presents their own case and, entirely understandably, presents it as a defence. Nobody challenges a colleague in front of the department. The case is closed with "unavoidable complication".

The blame hunt. The opposite, and rarer, but far more damaging. One bad meeting in which an individual is exposed will end honest reporting in that department for years.

The list without analysis. Twelve cases, three minutes each, no depth on any. Volume substitutes for insight.

No memory. Actions agreed last month are never revisited, so the same complication is discussed again in six months as though it were new.

Wrong cases. Only deaths are reviewed, so the near-misses and the moderate-harm events — where the learning actually is — never reach the room.

Choose cases on a rule, not on instinct

If case selection is left to whoever is organising the meeting, it drifts toward the dramatic and the uncontroversial. Write down inclusion criteria and apply them:

• All deaths within 30 days of surgery
• All unplanned returns to theatre
• All unplanned ICU admissions
• All readmissions within 30 days
• All surgical site infections requiring intervention
• All never events and serious incidents
• All nerve injuries, vascular injuries and wrong-site concerns
• Any case a team member — of any grade — asks to discuss

That last criterion matters more than the rest combined. A standing, unquestioned right for anyone to put a case on the agenda is the clearest signal a department can give about whether the meeting is genuinely about learning.

Deaths alone are the wrong denominator. In elective orthopaedics, mortality is low and most of it is not modifiable; the modifiable harm sits in infection, return to theatre, nerve injury and delayed diagnosis.

Give every case the same structure

Free-form storytelling is what allows a case to become a defence. A fixed template forces analysis:

Sixteen minutes per case means three or four cases in an hour, not twelve. A department that generates more reviewable cases than that needs a longer or more frequent meeting, not a faster one.

Classify preventability explicitly

Forcing a classification prevents the meeting from dissolving into "these things happen". A simple, widely used scheme:

Pair it with a harm severity grading. The pairing is what makes the register useful — a cluster of "potentially preventable, moderate harm" cases with a common factor is exactly the signal a department needs, and it is invisible if every case is discussed in isolation.

Expect the classification itself to be contested. That argument is the meeting. A department that classifies everything as not preventable is not reviewing its practice.

Discuss systems, not surgeons

The single most important chairing skill is redirecting from the individual to the conditions. The prompts that do this reliably:

• "What made this hard to spot?"
• "Was the information available to the person making the decision?"
• "Would a different team on a different day have done the same thing?"
• "What would have to be true for this to be impossible?"

The last one is the most productive question in surgical safety, because it moves the conversation from vigilance — which decays — to design, which does not. See Root Cause Analysis for the formal version of this analysis, and note that the strongest interventions are structural rather than educational, as covered in PDSA Cycles That Actually Finish.

A related rule: the operating surgeon should not be the only person presenting their own case. Having a colleague present it, with the surgeon present to add facts, removes the reflex to defend and makes honest analysis possible.

The actions register is what makes it real

This is the mechanism that separates meetings that change something from meetings that do not, and it is almost trivially simple.

Keep a standing register. Every action has a specific description, a named owner, and a date. And — the part departments skip — the meeting opens by reviewing the register, not by presenting the first new case.

Reading last month's actions out loud at the start of every meeting does three things. It makes it socially uncomfortable to accept an action you have no intention of completing. It surfaces actions that were never realistic. And it demonstrates to everyone in the room that the meeting has consequences, which is the thing that makes people engage honestly with the next case.

"Remind the team to be careful" is not an action. "Change the pre-op checklist so the implant size is confirmed by two people before the box is opened — owner: J. Smith — by 30 September" is an action.

Psychological safety, concretely

Nobody makes a meeting psychologically safe by announcing that it is. It is built from specific, visible choices:

• Consultants present their own complications, publicly, first — the culture is set from the top or not at all
• No case is attributed by name in the slides
• The chair interrupts blame language immediately and consistently
• Trainees are asked their view before the most senior person speaks
• A colleague who has had a bad outcome is contacted before the meeting, not ambushed in it
• Anyone may add a case without justifying why

The failure of this is expensive and quiet. A department where reporting has dried up looks, from the dashboard, exactly like a department that has become safe.

If a case genuinely raises a concern about an individual's practice, that is a separate process — a supportive conversation, and where necessary the formal route. Using the M&M meeting to conduct it destroys the meeting.

One jurisdictional caution

The legal status of what is said at M&M varies, and it matters.

In many US states, statutory peer-review privilege protects the content of qualifying quality-review proceedings from discovery in civil litigation — the protection is a creature of state law, with meaningful differences in scope and in what is required to qualify for it. Several other jurisdictions have statutory quality-assurance protections of varying strength.

In the UK, there is no general equivalent privilege: M&M minutes are ordinarily disclosable.

The practical consequence is the same either way, and it is not "say less". It is write factually. Record what happened, the contributory factors identified, and the actions agreed. Do not record speculation about blame, characterisations of colleagues, or conclusions about liability — those help nobody, and they read very badly when quoted back years later.

Accurate as of July 2026. Peer-review and quality-assurance protections are jurisdiction-specific and vary substantially by US state — confirm the position where you practise.

If you are the trainee presenting

Presenting at M&M is an examinable, interviewable skill and a genuinely useful one.

• Prepare the timeline properly. Most presentations fall apart because the sequence is unclear.
• Separate fact from interpretation. Present what happened, then what you think it means, and signpost the transition.
• Use the contributory-factors framework rather than a narrative. It shows structured thinking.
• Bring the literature where relevant — a complication rate from a real source changes the discussion.
• Propose an action. Even if it is rejected, proposing one demonstrates you understand what the meeting is for.
• Never criticise a colleague in the slides. Describe the decision and its context, not the person.

Done well, this is also portfolio evidence and strong interview material — far better than a completed audit that changed nothing.

The summary

Pick cases by written rule, including near-misses and anything a team member asks to discuss. Give every case the same structure and enough time to analyse rather than recite. Force an explicit preventability classification and expect it to be argued about. Chair toward systems and away from individuals. Keep an actions register with named owners and dates, and open every meeting by reading it.

The test of whether your M&M works is not how many cases it covers. It is whether anyone can name something the department does differently because of it.