One of the big unanswered questions in the management of pneumothorax is, how big is too big? At what size is a chest tube of some type mandatory?
The problem is that we just don’t have any good data. Seems like a simple problem, right? Unfortunately, it’s not. A pneumothorax is a three dimensional collection that surrounds the lung in very random ways. All we had to detect and “measure” them for decades was the lowly chest x-ray. Unfortunately, this is a 2D shadow picture that does not allow us to accurately estimate the size.
A few old papers exist that tried to quantify pneumothorax, but they are of no use now that we have chest CT. Unfortunately this new technology has drawbacks, as well. First, it’s just a stack of 2D images that our minds assemble into a 3D mental model, so it’s still difficult to quantify the air. And second, you shouldn’t be getting a chest CT just to diagnose pneumothorax. In blunt trauma, it’s really just for checking the thoracic aorta for injury.
So we’re left with the original question, and there are three answers. If there is any physiologic compromise (hypoxia, tachypnea, anxiety), then the chest should be drained. If the pneumothorax is enlarging over serial chest x-rays, then it should be drained before it causes physiologic change. And finally, if there is concern that it is so large that it will take too long to absorb, especially in older patients with comorbidities, a chest drain should be inserted. This is a somewhat soft indication, however.
Bottom line: The three reasons above are the usual answers to the question, “how big is too big?” For me, once the pneumothorax pushes the lung 1-2 cm away from the chest wall from apex to base, it’s time for a tube.
The hypothetical question I posed yesterday centered around what to do in a case where a patient is assaulted, sustaining easily survivable injuries, but then dies while being treated in the hospital due to a medical error. The police will escalate the criminal charge from simple assault to manslaughter, but the death was no longer really a direct result of the assault. Yet the assailant and police don’t really know any better. What to do?
There are many possible answers to this question, and it depends on who is being asked:
- The police / prosecuting attorney will say that it makes little difference. The assailant caused the victim to be injured and admitted to the hospital. Medical errors do occur in any hospital, and the assailant placed the victim in the position where this could occur. They will proceed with prosecuting the assailant on the higher charge.
- The hospital attorney will say that only the family may be informed of the error and resultant death. It is a privacy violation (in the US) to directly report any specific patient information to the police unless allowed by the family.
- The family will inform you that they are hiring an attorney to bring a civil malpractice case against you and the hospital.
- The assailant will say that you damn well better report it, and remind you that he’s facing years in prison if you don’t.
- The trauma professionals involved in the medical error will say that they should notify the police on ethical grounds so that the assailant will know that he was not responsible, and that he should not be punished as severely.
So what’s the right answer? As with any ethical questions in health care, there are only shades of gray. In the US system, the usual answer is to communicate the error to the family only. The justice system will not alter the charges based on the new information, so reporting the police is of no use and violates privacy laws.
Bottom line: In any situation as complex as the one described, proceed with great care. Seek out the advice of your mentors, the ethics committee, and the hospital attorney. One person’s idea of what is ethical may be very different from another’s, and the legal realities may render some of the arguments moot. Hasty and uninformed action without proper due process can have grave consequences for all.
What would you do in this case? And better yet, what should you do? And why might the two answers be different?
First, an important note. This is a hypothetical case. It has never happened in any hospital I’ve worked in, and I have not heard of it happening in one. I have completely fabricated it to make a point.
An elderly man is walking to the store in his neighborhood, and he is assaulted and knocked to the ground by a young man. Witnesses restrain the assailant, and police arrive to take him away to jail, while prehospital providers arrive and transport the victim to the hospital. The assailant is charged with assault and released.
The victim has a facial fracture and a very small intraparenchymal hemorrhage. He is expected to be discharged the following day after a repeat CT scan. The fracture does not need treatment. However, while being monitored in the ICU, a medical error occurs and the patient dies.
The police re-arrest the assailant and charge him with manslaughter, which has a much stiffer jail sentence.
Do you (or your hospital) have a responsibility to let the police know that the new charge is not justified? Is there a potential opening for a civil suit against you (or your hospital)? Can you do anything given current privacy laws?
Tweet out your answer or leave comments below. I’m interested in comments from my legal colleagues, too. What would you do?
The management of CSF leaks after trauma remains somewhat controversial. The literature is sparse, and generally consists of observational studies. However, some general guidelines are supported by large numbers of retrospectively reviewed patients.
- Ensure that the patient actually has a CSF leak. In most patients, this is obvious because they have clear fluid leaking from ear or nose that was not present preinjury. Here are the options when the diagnosis is less obvious (i.e. serosanguinous drainage):
- High resolution images of the temporal bones and skull base. If an obvious breach is noted, especially if fluid is seen in the adjacent sinuses, then a CSF leak is extremely likely.
- Glucose testing. CSF glucose is low compared to serum glucose.
- Beta 2 transferrin assay. This marker is very specific to CSF. However, the test is expensive and results may take several days to a few weeks to receive. Most leaks will have closed before the results are available, making this a poor test.
- Place the patient at bed rest with the head elevated. The basic concept is to decrease intracranial pressure, which in turn should decrease the rate of leakage. This same technique is used for management of mild ICP increases after head injury.
- Consider prophylactic antibiotics carefully. The clinician must balance the likelihood of meningitis with the possibility of selecting resistant bacteria. If the likelihood of contamination is low and the patient is immunocompetent, antibiotics may not be needed.
- Ear drops are probably not necessary. They may confuse the picture when gauging resolution of the CSF leak.
- Wait. Most tramatic leaks will close spontaneously within 7-10 days. If it does not, a neurosurgeon or ENT surgeon should be consulted to consider surgical closure.
- Brodie HA, Thompson TC. Management of complications from 820 temporal bone fractures. Am J Otol, 1997;18:188-197.
- Brodie HA. Prophylactic antibiotics for posttraumatic cerebrospinal fluid fistulas. Arch Otolaryngol Head, Neck Surg. 123:749-752.
In my last post, I discussed the Passy-Muir valve, which allows patients who have a tracheostomy tube in place, and are not on a ventilator, to talk. But what about patients who are still vent dependent? It’s very frustrating for both patient and trauma professionals when we can’t communicate with each other.
Pulmodyne, Inc. makes the Blom tracheostomy tube system, which solves this problem. This device has a large fenestration in the back of the tube with a special bubble valve (see below), coupled with an inner cannula that has a 1-way flap valve. This allows controlled release of air into the pharynx, enabling speech while on the ventilator.
A multicenter study looked at voice production and intelligibility of speech in a group of 23 ventilated patients with a trach tube in place. Although not entirely clear in the paper, it appears that all were changed to Blom trach tubes for the study (2 had one in place at the beginning of the study). Overall, voice production and intelligibility were good. Most were able to begin audible speech within about 6 minutes of initial application. One deconditioned patient took longer. The video below shows an example of the speech that is achievable.
Bottom line: This novel product allows a subset of trauma patients to speak while still on the ventilator. It is most appropriate for those who do not have significant head injury, especially those with facial trauma requiring airway protection with a tracheostomy.
I have no financial interest in Pulmodyne, Inc.