Tag Archives: hemothorax

Chest Tubes: Size Doesn’t Matter – Part 2

A few days ago, I wrote about a paper that seemed to suggest that using a smaller chest tube (28-32 Fr) vs larger ones (36-40 Fr). The results suggested that their function was very similar. I emphasized that I thought the result was intriguing, because I’m of the opinion that bigger is better for getting clotted blood out. However, I am amenable to changing my mind based on newer, better data.

But I did caution readers that I would like to see more data. One study should never change your practice! Then I see a lot of chatter on Twitter about another study from 2016 that looks at even smaller tubes, with people saying they will now switch to pigtail catheters (12 Fr)!!

First, not a logical progression of thinking there. And second, let’s take an actual look at the paper. It’s from an emergency medicine group in Fukui, Japan, which retrospectively reviewed their 7 year experience with using a small (20-22 Fr) vs large (28 Fr) tubes. They identified a total of 124 chest tube insertions to compare, 68 small and 56 large.

Now let’s look at the factoids:

  • Demographics, mechanism, and ISS were the same between groups
  • Duration of insertion and initial drainage were also the same between groups
  • Complication rates were similar, with 1 empyema and 2 retained hemothoraces in each group
  • Additional tubes were place in 2 patients with small tubes vs 4 with large tubes
  • Thoracotomy was performed in 2 patients with small tubes vs 1 with a large tube

Based on all of this, the authors concluded that there was no difference in drainage efficacy, complications, or need for additional invasive procedures.

Wait a minute!! Again, if you only read the abstract, you might be led to start using ever smaller chest tubes. But read the entire paper! There are many problems with this paper, including:

  • It’s a very small, retrospective review. This automatically means that the statistical power is suspect.
  • Why did they only document 124 insertions over 7 years?? That’s about one every 3 weeks! Either a lot of data are missing or they are not very busy. But Fukui Prefectural Hospital has over 1000 beds! So it’s the former, not the latter.
  • The retrospective nature means it is not possible to determine why a particular tube size was chosen. Roll of the dice? This fact alone introduces a huge potential for selection bias. Was a smaller tube selected because the hemothorax looked smaller? Probably! The fact that 4 patients with larger tubes had another one placed suggests that they were being used for larger collections. And patients with higher ISS tended to get bigger tubes.

Bottom line: Don’t change your practice based on this paper. And certainly don’t choose to use even smaller pigtails. And of course, always critically read any paper that you like to make sure you are not cherry picking the ones you choose to believe. IMHO, it’s still best to use big (36 Fr) or bigger (40 Fr).

Reference: Small tube thoracostomy (20-22 Fr) in emergent management of chest trauma. Injury 48:1884-1887, 2016.

Retained Hemothorax Part 4: The Practice Guideline

Over the last three days, I reviewed some data on lytics at the request of some of my readers. Then I looked at a paper describing one institution’s experience dealing with retained hemothorax, including the use of VATS. But there really isn’t much out there on how to roll all this together.

Until now. The trauma group at Vanderbilt has a paper in press describing their experience with a home-grown practice guideline for managing retained hemothorax.  Here’s what it looks like:

I know it’s small, so just click it to download a pdf copy. I’ve simplified the flow a little as well.

All stable patients with hemothorax admitted to the trauma service were included over a 2.5 year period. The practice guideline was implemented midway through this study period. Before implementation, patients were treated at the discretion of the surgeon. Afterwards, the practice guideline was followed.

Here are the factoids:

  • There were an equal number of patients pre- and post-guideline implementation (326 vs 316)
  • An equal proportion of each group required an initial intervention, generally a chest tube (69% vs 65%)
  • The number of patients requiring an additional intervention (chest tube, VATS, lytics, etc) decreased significantly from 15% to 9%
  • Empyema rate was unchanged at 2.5%
  • Use of VATS decreased significantly from 8% to 3%
  • Use of catheter guided drainage increased significantly from 0.6% to 3%
  • Hospital length of stay was the same, ranging from 4 to 11 days (much shorter than the lytics studies!)

Bottom line: This is how design of practice guidelines is supposed to work. Identify a problem, typically a clinical issue with a large amount of provider care variability. Look at the literature. In general, find it of little help. Design a practical guideline that covers the major issues. Implement, monitor, and analyze. Tweak as necessary based on lessons learned. If you wait for the definitive study to guide you, you’ll be waiting for a long time.

This study did not significantly change outcomes like hospital stay or complications. But it did decrease the number of more invasive procedures and decreased variability of care, with the attendant benefits from both of these. It also dictates more selective (and intelligent) use of additional tubes, catheters, and lytics. 

I like this so much that I plan to adopt it at my center!

Download the practice guideline here.

Related posts:

Posts in this series:

Reference: Use of an evidence-based algorithm for patients with traumatic hemothorax reduces need for additional interventions. J Trauma, in press, December 14, 2016.

Retained Hemothorax Part 3: VATS

I’ve written about the use of lytics to treat retained hemothorax over the past few days. Although it sounds like a good idea, we just don’t know that it works very well. And they certainly don’t work fast. Lengths of stay were on the order of two weeks in both studies reviewed.

The alternative is video assisted thoracoscopic surgery (VATS). So let’s take a look at what we know about it. This procedure is basically laparoscopy of the chest. A camera is inserted, and other ports are added to allow insertion of instruments to suck, peel, and scrape out the hemothorax.

A prospective, multi-center study was performed over a 2 year period starting in 2009. Twenty centers participated, contributing data on 328 patients with retained hemothorax. This was defined as CT confirmation of retained blood and clot after chest tube placement, with evidence of pleural thickening.

Here are the factoids:

  • 41% of patients had antibiotics given for chest tube placement (this is interesting given the lack of consensus regarding their effectiveness!)
  • A third of patients were initially managed with observation, and most of them (82%) did not need any further procedures (83 of 101 patients)
  • Observation was more successful in patients who were older, had smaller hemothoraces (<300cc), smaller chest tubes (!!, <34 Fr), blunt trauma, and peri-procedure antibiotics (?)
  • An additional chest tube was inserted in 19% of patients, image guided drain placement in 5%, and lytics in 5%. Half to two-thirds of these patients required additional management.
  • VATS was used in 34% of patients. One third of them required additional management including another chest tube, another VATS, or even thoracotomy.
  • Thoracotomy was most likely required if there was a diaphragm injury or large hemothorax (<900cc)
  • Empyema and pneumonia were common (27% and 20%, respectively)

Bottom line: There’s a lot of data in this paper. Most notably, many patients resolve their hemothorax without any additional management. But if they don’t, additional tubes, guided drain placement, and lytics work only a third of the time and contribute to additional time in the hospital. Even VATS and thoracotomy require additional maneuvers 20-30% of the time. And infectious complications are common. This is a tough problem!

Tomorrow, I’ll try to roll it all together and suggest an algorithm to try to optimize both outcomes and cost.

Posts in this series:

Reference: Management of post-traumatic retained hemothorax: A prospective, observational, multicenter AAST study. 72(1):11-24, 2012.

Retained Hemothorax Part 1: Lytics

Hemothorax is a common complication of chest trauma, occurring in about one third of cases. It is commonly treated with a chest tube, which usually takes care of the problem. But in a few cases some blood remains, which can result in an entrapped lung or empyema.

There are several management options. Historically, these patients underwent thoracotomy to peel out the fibrinous collection stuck to lung and chest wall. This has given way to the more humane VATS procedure (video assisted thoracoscopic surgery) which accomplishes the same thing using a scope. In some cases, another tube can be inserted, sometimes under CT guidance, to try to drain the blood.

So what about lytics? It’s fibrin, right? So why not just dissolve it with tissue plasminogen activator (tPA)? There have been very few studies published over the years. The most recent was in 2014. I’ll review it today, and another tomorrow. Finally, I’ll give you my thoughts on the best way to deal with retained hemothorax.

Here are the factoids:

  • This was a single center, retrospective review of data from 1.5 years beginning in 2009
  • A total of seven patients were identified, and most had hemothorax due to rib fractures. Three presented immediately after their injury, 4 were delayed.
  • Median time from injury to chest tube placement was 11 days
  • Median time the chest tube was in place was 13 days, with an average hospital stay of 14 days
  • Patients received 1 to 5 treatments, averaging 24mg per dose
  • There was one death in the group, unrelated to TPA treatment
  • No patient “required” VATS, but one underwent thoracotomy, which turned out to be for a malignancy

Bottom line: The authors conclude that tPA use for busting retained hemothorax is both safe and effective. Really? With only seven patients? The biggest problem with this study is that it uses old, retrospective data. We have no idea why these patients were selected for tPA in this 5-year old cohort of patients. Why did it take so long to put in chest tubes? Why did the chest tubes stay in so long? Maybe this is why they were in the hospital so long?

Plus, tPA is expensive. A 100mg vial runs about $6000. Does repeatedly using an expensive drug and keeping a patient in the hospital an extra week or so make financial sense? So it better work damn well, and this small series doesn’t demonstrate that.

Tomorrow, I’ll look at the next most recent paper on the topic, from way back in 2004.

Posts in this series:

Reference: Evaluation of chest tube administration of tissue plasminogen activator to treat retained hemothorax. Am J Surg 267(6):960-963, 2014.

Serial Hemoglobin To Monitor Chest Tube Output? Huh?

I’ve already written about the (f)utility of serially monitor hemoglobin (Hgb) or hematocrit (Hct) levels when managing solid organ injury nonoperatively. What about if you are concerned with bloody output from a chest tube drainage system? Could it be of any use there?

Seems like a reasonable idea, right? Wrong. As always, think it through and do the math! Here are the questions you need to ask yourself:

  • What is the Hgb or Hct of the fluid coming out of the chest tube? At worst, it will be the same as the patient, assuming that pure, whole blood is coming out. But this is seldom the case. The fluid is usually described as “serosanguinous”, which is not very exact, but tells you that it is thinner than blood. And if it looks more like Kool-Aid, the concentration is very low indeed.
  • What is the volume in the container? Most collection systems will collect a maximum of 1 to 1.5 L of juice.
  • How fast is it coming out? These things almost never fill right in front of your eyes. It’s a slow process, with less than a few hundred ccs per shift.

Here’s a few hundred ccs of thin drainage in a collection system. Probably decrease in Hgb value – < 0.1, which is far less than the range of lab error.

Bottom line: So now do the math. Let’s say the fluid has half the hematocrit / hemoglobin of whole blood. Losing one unit (500cc) of whole blood will generally drop your Hgb by about 1 gm, or your Hct by about 3%. If the blood is half-strength like I am proposing (and the usual drainage is typically much thinner), it will take twice as much (one liter) loss to drop the lab values by that much. This will probably come close to filling up the average collection system. If it takes a day or two or more to fill up, you are not going to see much change in their lab values. And most of the time, the blood in the system is thin like Kool-Aid, so your patient is really losing very little actual blood.

So measuring serial hemoglobin / hematocrit as you watch a hemothorax drain doesn’t make sense. Unless the output is pure blood and the system is filling up in front of your eyes, of course. In that case, a trip to the OR to fix the problem might be a better idea than doing a blood draw and sitting around waiting for the result to come back.

Related post: