Two days ago, I discussed getting the traditional chest x-ray routinely after chest tube insertion. The answer was yes, it is important even if it appears to be functioning correctly. But yesterday, I also showed you how the chest x-ray can lie.
Remember this image?
Looks perfect! But it’s a 2-D view and you don’t know where the tube is in the anterior-posterior axis. It turns out to be in the patient’s subcutaneous tissues of his back, near his scapula!
So what if this is a trauma activation patient and you are getting ready to send your patient for a chest CT shortly? Should you follow the usual dogma and still get a conventional chest x-ray prior to leaving the trauma bay?
The answer is no! Typically, your trauma activation patient should have rapid access to the CT scanner, so you won’t have to wait very long. And the additional 3-D information is very helpful in making sure the tube is placed exactly where you want it.
Bottom line: If you are planning on obtaining a chest CT anyway in your trauma patient, don’t bother with a conventional chest x-ray first to check chest tube position. But DON’T order a chest CT for this reason alone! Remember, the chest CT is only for detecting aortic injury in blunt trauma. It should not be used for diagnosing fractures, hemothorax, or pneumothorax. Or chest tube position!
A blunt trauma activation patient presents with a pneumothorax seen on the initial chest x-ray, obtained in your trauma bay. You professionally insert a large chest tube, and all appears to go well. You shoot a followup chest x-ray and this is what you get:
What do you think of the tube position? Looks great, right?
But if you look carefully, you can see the lung outline in the middle of the right side of the chest. Big-time pneumothorax despite what looks like a perfectly placed tube. There are several possible explanations, and many of you sent me your guesses:
- The tube is in the lung. This rarely happens to normal lungs. Sure, you can probably do it to an ARDS lung, but otherwise it’s not very likely.
- The tube is in the fissure. This does happen on occasion, but not often. And many times it works anyway.
- The tube is occluded or kinked. A PA or AP chest x-ray will show the kink, although bent tubes frequently work anyway. If a hemothorax is present, it is possible that a clot is plugging the tube. Clearing a plugged tube will be the subject of another post.
- It’s not really a chest tube. Hopefully, this would have been detected when it was placed, but it isn’t always. The chest x-ray above looks great, right? Unfortunately, it’s a 2 dimensional representation of a 3-D object. Where is that tube in the z-axis?
In this case the correct answer is the last one. This is one time when I would actually recommend a lateral chest x-ray. Have a look at the result. You can clearly see the tube snaking around into the soft tissues of the back.
Bottom line: Remember that a perfect x-ray doesn’t necessarily mean a perfect tube. Go through the various possibilities quickly, and make it work.
More dogma, or is it actually useful? Any time a chest tube (tube thoracostomy) is inserted, we automatically order a chest x-ray. Even the ATLS course recommends obtaining an image after placement. But anything we do “automatically” is grounds for critical analysis to see if there is a valid reason for doing it.
A South African group looked at the utility of this practice retrospectively in 1004 of their patients. They place 1042 tubes. Here are the factoids:
- Patients were included if they had at least one chest x-ray obtained after insertion
- Patients were grouped as follows: Group A (10%) had the tube inserted on clinical grounds with no pre-insertion x-ray (e.g. tension pneumothorax). Group B (19%) had a chest x-ray before and had ongoing clinical concerns after insertion. Group C (71%) had a chest-xray before and no ongoing concerns.
- 75% of injuries were penetrating (75% stab, 25% GSW), 25% were blunt
- Group A (insertion with pre-x-ray): 9% had post-insertion findings that prompted a management change (kinked, not inserted far enough)
- Group B (ongoing clinical concerns): 58% required a management change based on the post-x-ray. 33% were subcutaneous or not inserted far enough (!!)
- Group C (no ongoing clinical concerns): 32 of 710 (5%) required a management change, usually because the tube was too deep
The authors concluded that if there are no clinical concerns (tube functioning, no clinical symptoms) after insertion, then a chest x-ray is not necessary.
Bottom line: But I disagree with the authors! Even with no obvious clinical concerns, the tube may not be functioning for a variety of reasons. Hopefully, this fact would then be discovered the next day when another x-ray is obtained. But this delays the usual progression toward removing the tube promptly by at least one day. It increases hospital stay, as well as the likelihood of infection or other hospital-associated complication. A chest x-ray is cheap compared to a day in the hospital, which would potentially happen in 5% of these patients. I recommend that we continue to obtain a simple one-view chest x-ray after tube insertion.
Tomorrow: Look at the chest x-ray. Is it a good chest tube?
The next day: What if you placed the chest tube in your resuscitation room and are planning to go to CT for additional imaging? Is it worthwhile getting a chest x-ray, or should you just check the tube with the CT scan?
Previous studies have shown that higher hospital costs are associated with longer length of stay (LOS). This makes sense, because the longer a patient stays in the hospital, the more that is “done” for them, and more daily charges are incurred. Obvious savings can occur if we look globally at services, medications, etc while the patient is in the hospital.
But does the admission service make a difference in LOS or cost? It shouldn’t if care is fairly uniform. A group of orthopedic surgeons at Vanderbilt in Nashville looked at a large group of isolated hip fracture patients (n=614) to see if LOS (used as a surrogate for cost) was significantly different. They also tried to control for a host of factors that could affect time in the hospital between the two groups.
Here are the factoids:
- About half of the patients were admitted to the orthopedics service, and half to medicine
- Median length of stay was way different! 4.5 days on Ortho vs 7 days on Medicine
- Readmission rates were also significantly higher on Medicine, 30% vs 23%
- After controlling for factors such as medical comorbidities, age, smoking and alcohol, ASA score, obesity, and others, a regression model showed that patients were still likely to stay about 50% longer if admitted to a medicine service.
Bottom line: Obviously, this is the experience of a single institution. But the difference in length of stay, and hence costs, was striking. As the US moves toward a bundled payment system, this will become a major problem. The initial LOS is more costly on the medicine service, and readmission for the same problem will not be reimbursed. Why the difference? Coordination of care between two services? Lack of familiarity with surgical nuances? This study did not look at that.
But it does point out the need to more closely integrate the care of the elderly in particular, and patients with a broad range of needs in general. An integrated team with orthopedic surgeons and skilled geriatricians is in order. And a set of protocols for standard preop evaluation and postop management is mandatory.
Does Admission to Medicine or Orthopaedics Impact a Geriatric Hip Patient’s Hospital Length of Stay? J Orthopedic Surg 30(2):95-99, 2016.
Splinting is an important part of the trauma resuscitation process. No patient should leave your trauma resuscitation room without splinting of all major fractures. It reduces pain, bleeding, and soft tissue injury, and can keep a closed fracture from becoming an open one.
But what about imaging? Can’t the splint degrade x-rays and hamper interpretation of the fracture images? Especially those pre-formed aluminum ones with the holes in them? It’s metal, after all.
Some of my orthopedic colleagues insist that the splint be removed in the x-ray department before obtaining images. And who ends up doing it? The poor radiographic tech, who has no training in fracture immobilization and can’t provide additional pain control on their own.
But does it really make a difference? Judge for yourself. Here are some knee images with one of these splints on:
Amazingly, this thin aluminum shows up only faintly. There is minimal impact on interpretation of the tibial plateau. And on the lateral view, the splint is well posterior to bones.
On the tib-fib above, the holes are a little distracting on the AP view, but still allow for good images to be obtained.
Bottom line: In general, splints should not be removed during the imaging process for acute trauma. For most fractures, the images obtained are more than adequate to define the injury and formulate a treatment plan. If the fracture pattern is complex, it may be helpful to temporarily remove it, but this should only be done by a physician who can ensure the fracture site is handled properly. In some cases, CT scan may be more helpful and does not require splint removal. And in all cases, the splint should also be replaced immediately at the end of the study.