I’ve spent the last several posts reviewing the sparse data that we have on the impact of subdural hematoma management in elderly patients. With this information, I had hoped to arrive at some answers as to what to do when certain common patient presentations are encountered.

Unfortunately, the data is not very good, and is structured to raise false hopes. Overall, it looks like about 30-40% of selected patients die in the postoperative period. And the percentage of patients who are discharged at their pre-injury level of independence is in the low single digits. In fairness, one paper did show an improvement from severely disabled to moderately disabled or recovered, although the authors obfuscated how many actually made it to the good recovery group.

The biggest problem with all of the literature we have is that the patients were selected for surgery based on the opinion of the neurosurgeon. This means that many patients who they felt would do poorly with operation were excluded. It is extremely likely that the inclusion of these patients would have dragged down the already poor numbers that were reported. But in fairness, we might have also found a few surprise saves among those patients; I guess we’ll never know.

So let me give you my take on the scenarios that I presented so many days ago. Remember, these are my opinions and are not meant to be gospel. Other trauma professionals will need to interpret the information themselves and make their own decisions.

Scenario 1 – An elderly female falls and sustains a modest subdural hematoma with no shift and a normal exam. Follow your established practice guidelines unless some of the factors in the following scenarios are present.

Scenario 2 – Same as above, but the patient presents 8 hours after the fall. The clock started ticking when the fall occurred. Since your practice guideline recommends monitoring for 6 hours and then a followup CT of the head, the initial CT is the followup scan. The patient could then be discharged if there are no alarming findings on initial CT and the neuro exam is normal.

Scenario 3 – Same as scenario 1 but the patient has advanced dementia. These patients were generally excluded from the studies, and they are not expected to do well. Frankly, they will likely be much worse after an operation and will require an even higher level of post-discharge care (if they make it that far) and more involvement of family. It is critically important that the trauma professionals have a frank talk with the family to make sure they understand the overwhelming likelihood that their loved one will never be as good as they were before the injury. Surviving an operation does not mean going back to their usual living situation. The family absolutely needs this information to make the best choice for their loved one.

Scenario 4 – Same as scenario 1 but the patient has a well documented “do not actively resuscitate” order in place. The patient and their family need the same talk as above so they can appreciate all of the risks and the few, if any, benefits of surgery. Only then can they make an informed if they want to consider temporarily rescinding their DNAR order to allow surgery.

Scenario 5 – Same as scenario 1 but the patient is 95 years old. The data showed that patients in their 80s tended to do even more poorly than younger patients. There were very few nonagenarians in the literature, but it can be expected they would do even worse than the octos. They and their families need the same depressing talk so they can make the right decision.

Bottom line: Communication is key. And good data is even more key, although we have too little of it. For now, all we can do is paint a somewhat depressing picture of generally poor outcomes in highly selected patients. Hopefully we’ll have better data some day and can slice and dice things a little better. This may eventually allow us to offer surgery to those patients who will actually benefit from it the most.

In my last post, I discussed a paper that examined the fate of very well-selected elderly patients with traumatic subdural hematoma. Today, I’ll focus on one that was just published that took all comers, kind of. Hopefully, this will give us a better idea of what outcomes to expect after emergency craniotomies.

This work was conducted at the University of New Mexico, and was yet another retrospective review. They took all comers with age > 65 and acute subdural hematoma. However, criteria for proceeding to surgery were based on neurosurgeon discretion. Only 62 patients were identified during a 5 year period, and the Glasgow Outcome Scale score was the primary outcome.

Here are the factoids:

• 60% of patients were taking preoperative anticoagulant or antiplatelet drugs
• Perioperative mortality was 39%, and this increased to 44% at three months
• Of the remaining 38 survivors, 4 were in a vegetative state,  26 were severely disabled, 6 had moderate disability, and 2 had a good recovery
• By 6 months, 20 of the patients in the severely disabled category improved to either moderate disability or good recovery

The authors conclude that, although mortality was high, a significant number of patients (31%) made a meaningful recovery by 6 months. These were patients who had achieved a GOS score of 4 or 5.

Bottom line: Once again, read closely. If you look at the numbers at discharge, 39% were dead and 3% were recovered. The rest ranged from a vegetative state to varying degrees of disability and independence. 

Over the first three months  postop, the severe disability number shrank, with 5 dying, 3 moving to moderate disability, and 12 making a recovery. This continued to improve somewhat over the next 3 months, but the authors don’t clearly state how many were actually in the recovered group.

So the final numbers that we can tease apart show a 44% mortality and at least 25% recovered. These sound pretty good, right?

Unfortunately, the retrospective design and small numbers are heavily influenced by the selection process. Remember, the patients who received an operation were more likely to survive if the neurosurgeon was skilled in selecting his or her patients vs declaring them as having a “nonsurvivable injury.” We still don’t know the answer to our questions for all comers, but it’s probably quite a bit worse than these numbers. I would imagine that every one of those not operated upon died, and including them would skew these numbers tremendously towards nonsurvival.

So what’s a trauma professional to do? In my next post, I’ll try to bring it all together in a way that we can apply to our own patients.

Reference: Mortality and functional outcome in surgically evacuated acute subdural hematoma in elderly patients. World Neurosurg 126:e1234-e1241, 2019.

In previous posts, I proposed several scenarios with elderly patients presenting with subdural hematomas and discussed the use of practice guidelines to help direct their care. The principal conundrum has been in knowing who will do well vs who will not.

Today, I’ll review a paper that examined functional outcome / salvageability in patients with subdural hematomas. It is from a Swiss group that retrospectively reviewed their experience over a six year period. Interestingly, they had specific criteria in place (fifteen years ago) that would limit craniotomy to study patients with:

• A Karnofsky Performance Scale score of 80 or more and living independently. This scale evaluates the ability to carry out activities of daily living using a score of 0 to 100. Scores > 80 indicate that there may be some symptoms of disease, but daily activities can be carried out with some effort or less.
• No known dementia
• No comorbidities that had a survival time of less than 12 months.
• Desire to proceed with surgery and consent to do so.

Patients with fixed, dilated pupils were excluded. Here are the factoids:

• 42 patients older than 65 years presented during the study period, and 37 met inclusion criteria
• 81% of patients had comorbidities and 43% were on some type of anticoagulant or platelet agent
• Median GCS was 8, so these patients had significant head injury
• One third (13) died in the perioperative period, and one quarter experienced nonlethal complications
• Anticoagulation or antiplatelet agents did not appear to affect mortality
• Final Glasgow Outcome Scale scores were favorable (4-5) in 40% and unfavorable to severely disabled (1-3) in 60%. However, these numbers were calculated using all 37 study patients, and did not exclude the 13 who died! I’m not sure how this works, exactly.

Bottom line: Read this one closely. The authors conclude that, although morbidity, mortality, and adverse outcomes are high, there is a good outcome in 41% of patients.

Really? This is why it is so important to read the whole paper. If you just browsed the abstract and its conclusion, you would have missed the fact that they only accepted independent patients with no dementia or critical comorbidities! The patient group was highly selected which biased them toward better outcomes. Furthermore, there were only 37 people in this retrospective study. 

Personally, I learned very little from this study. I cannot use it to guide me in answering the questions I posed with the original scenarios.  Tomorrow, I’ll review a more recent paper to see if we can find any more clues.

Reference: Age and salvageability: analysis of outcome of patients older than 65 years undergoing craniotomy for acute traumatic subdural hematoma. World Neurosurg 78(3/4):306-311, 2012.

I described several variations on the theme of elderly patients and subdural hematoma in my last post. All were situations in which an operation was not immediately indicated. Practice guidelines were in place to smooth the evaluation process for such patients. But do those guidelines really apply in some or all of these cases?

The real question that needs to be answered is “what is the real purpose of the guideline?”

Is it designed to standardize and streamline care? Certainly. But what is it’s real purpose? It is supposed to separate those who need additional treatment from those who do not. So in this case, it seeks to identify patients who are likely to need surgical intervention for their lesion.

In scenario 2, where the patient presents 8 hours after the fall, the “evaluation timer” started at the time of the event. If your practice guideline dictates that you obtain a repeat head CT 6 hours after arrival in the ED, isn’t your first scan at 8 hours really the same as the repeat scan? Shouldn’t you just need the one image, then send them home if they have a normal neurologic exam?

And isn’t there a point at which surgical intervention is no longer an option? That’s what makes scenarios 3-5 more difficult. Can we identify a subset of patients for whom surgery is not an option? For those who have a written “do not resuscitate” status (scenario 4) and don’t change their mind, is any followup evaluation needed at all?

For the other scenarios, we really need to know if there are subsets of patients for whom surgical intervention is inadvisable or contraindicated. Those patients should not need followup studies or even additional monitoring. One could even argue that they don’t need to be seen in an ED at all!

Lots of questions! In my next post, I’ll review some of the data on outcomes after brain surgery for traumatic injuries in elderly patients. Hopefully, we can come to some conclusions and/or recommendations for my scenarios based on this data!