Adverse events are always possible when dealing with patients who have traumatic injuries. And there’s probably no event more adverse than death. Your trauma performance improvement program must quickly and effectively identify these events, determine if there were any opportunities that might make them less likely in similar future patients, and implement a remedy if appropriate.

But as usual, it’s complicated. Mortality can generally be broken down into anticipated vs unanticipated. But it’s not as cut and dried as you might expect. What is an unanticipated mortality? How many of each do trauma programs have in a typical year? How do you take into account withdrawal of care? How does your PI program handle and document it? And what about opportunities for improvement? Let’s dig in to this complex topic.

What Is An Unanticipated Mortality (UM)?

There are several possible definitions. The most quantitative way is to run the numbers. Was your patient mathematically expected to survive, yet did not? You can use your trauma registry to make a first-pass scan of your patients to look for candidates.

The simplest method is to perform a query that identifies all patients with a TRISS probability of survival greater than 50% who died. TRISS predicted that this patient had a better than even prospect of survival, but for whatever reason it didn’t work out that way. You should only use this technique as a last resort, or to double-check how well your program normally detects them. If you do use it, understand that it overestimates the number of UMs. This occurs for two reasons:

• TRISS probability of survival scoring was not really designed for this use, and gives a significant number of false positives
• It does not take withdrawal of care into consideration, so many elderly patients will incorrectly fall into this category

The other definition is more accurate, but requires good subjective clinical judgement. Ask an experienced provider this question, “given the severity of injuries and patient demographics and comorbidities, would a similar patient have survived?” As I mentioned, sound and unbiased clinical judgement is needed for this one. An experienced trauma surgeon who is impartial can render this initial judgement.

Whichever technique you use, though, there is much more to do. This just starts the ball rolling so that you can use the full capabilities of your trauma program to investigate.

How Often Do Unanticipated Mortalities Occur?

Obviously, any number is too many. But trauma care is complex, as are our patients. There are many, many opportunities for things to go awry. Good clinical care tries to identify, anticipate, and avoid them as much as possible. But a few will always slip through.

I tend to visualize the mortalities seen at trauma centers as a set of exponential ratios. For every one unanticipated mortality, I expect about ten mortalities with opportunity for improvement and a hundred mortalities without opportunity. So a rough ratio looks like this:

1 UM : 10 OFI : 100 no OFI

Remember, this is a generalization. Obviously, we all strive to have no UMs. Some programs may identify more OFIs because they are counting trivial issues like not taking a temperature in the trauma bay. But in general, this number holds. If you find that your numbers are out of bounds, dig deeper to identify the reason.

What About Withdrawal Of Care?

Just because the patient is old or has severe comorbidities does not change the determination. These factors only increase the likelihood of death, but have minimal impact on UM vs non-UM. The fact that an adverse event occurred just makes it more likely for the family to withdraw care. Could that event have been avoided?

The best way to determine a preliminary classification of UM is to get that experienced, unbiased trauma surgeon back in the room. Forget that you already know the outcome. Given an average patient in that condition, should they have been reasonably likely to leave the hospital?

How To Handle An Unanticipated Mortality

So you’ve identified an unanticipated mortality. What next? First, collect and save as much information as possible! The chart documentation will be there forever. But people’s recollections fade with time, and the medical record captures very little of the minute to minute management and thought processes that were taking place. Talk to everyone involved and take detailed notes. Assemble the key documentation from the medical record.

UMs universally require discussion at tertiary PI review, the multidisciplinary trauma PI committee. You cannot close this at primary or secondary review, ever! Trauma center reviewers take particular interest in these charts because they demonstrate the strengths (or weaknesses) of your PI program. They are guaranteed to go through every UM chart with a fine-toothed comb.

And since time is of the essence, make sure that the case is added to the agenda of the very next PI committee meeting. No delays! Make sure that all participants have materials in advance to review, and that everyone has done there homework prior to the meeting. There is no excuse for people trying to browse through the record at the actual meeting. Consider it a light version of a Root Cause Analysis, because it actually is.

Present the case, and have all involved participants comment on their (or their colleagues’) parts. Then open a general discussion about the root causes of the death, and exactly what could have been done to prevent it. Remember, the purpose of PI is to ensure that “this event is unlikely to happen to another patient because…” The group will need to come to a consensus that the mortality was truly unanticipated and classify it as such. It will also need to identify specific opportunities for improvement that require correction so the event can’t happen again. Finally, action plans need to be formulated for each opportunity, with specific people responsible for getting them done.

But wait, you’re not done yet. The PI program now needs to ensure that each and every one of these action plans are carried out and complete. Documentation needs to be detailed and easy to follow. This starts with the minutes from the PI committee meeting, which need to clearly delineate the discussion and list the action items and persons responsible for them.

Documentation of the work to resolve the action items can be maintained in several ways. It can be a set of addenda to the PI committee minutes, or it can be free-standing documentation that is linked to the PI minutes. Be sure that everything is easy to follow, and that it includes any and all supporting documentation. This might be workgroup minutes, or new policies, or practice guidelines that were implemented as a consequence of the event and PI committee discussion.

Are There Always Opportunities For Improvement?

One would think that there is always something that could have been done, right? Over my career, I have seen exactly three unanticipated mortalities without any opportunities for improvement. How can this happen, you say?

Here’s a theoretical example:

A young male in excellent health is involved in a moderate speed motor vehicle crash. He sustains an isolated closed tibia/fibula fracture. He is admitted to the trauma service in the afternoon and is given a 30mg dose of Lovenox per your protocol. He undergoes ORIF the next morning, and twice-daily Lovenox resumes postop. He has an uneventful recovery and is working well with therapies the following day. He is deemed ready for discharge on postop day 2. As he is getting dressed to go home, he clutches his chest and falls to the floor in PEA arrest. He cannot be revived.

Huh? What happened here? The two top possibilities are a freak massive pulmonary embolism, or a freak arrhythmia due to a pre-existing conduction system defect. Is this an unanticipated mortality? By either definition above, yes!

Now, what can you do to make this less likely to occur in future patients? Nothing! They received DVT prophylaxis per your practice guideline, and no other therapy or investigation was indicated. There was no indication of heart disease in the patient’s history. Should you have done an EKG or an echo? Are you going to start doing that on every young healthy patient after this? No!

So all things considered, the next time this type of patient comes in, you will do exactly the same things. And they will do fine 99.999% of the time. This is a perfect example of one of the rare cases with an unanticipated mortality without opportunity for improvement.

But the reality is that nearly every UM you encounter at your program will have opportunities for improvement. Be sure to identify every one of them, look at every aspect of care to find the opportunities for improvement, discuss in your multidisciplinary trauma PI committee, and implement an aggressive action plan to keep it from happening again. And remember, work not documented is work not done!