So once your routine cardiopulmonary bedside ultrasound examination has revealed the presence of a significant pleural effusion, should you drain it?
I guess the first real question is actually what is a significant pleural effusion? As you can imagine, the N=1 principle pops up again, and there is no single threshold answer (e.g.: >750ml drained and <749ml ignored), naturally. So here are the main factors that impact physiologically:
1. underlying global respiratory function – eg how much reserve you have. With healthy lungs, a young adult can usually tolerate complete atelectasis of one lung or even more, which would take 2-3 liters of effusion at least. However, the more compromised gas exchange is, the less atelectasis can be tolerated. This also encompasses the level of required ventilatory support, including the level of PEEP. Basically, the worse your lungs are, the more significant the effusion becomes.
2. thoracic compliance – which combines chest wall and abdominal pressure, since the diaphragm is thin and easily displaced. The greater the compliance, the less atelectasis will occur as the diaphragm and chest wall will shift to accommodate some of the fluid. This is very important especially in surgical patients (or anyone with a tense abdomen).
…so there is no number. Of course, if you take it to the extremes, I think everyone would agree that if you have 200ml, it’s probably not worth doing, and if you have 2,000ml, it is. But in the grey zone of, say 500-1,000 ml, you have to make a call given the patient in front of you.
Now what does the literature say?
First of all, there is no large study looking at the impact of effusions in generally ill patients (ED/ICU). There are some small, specific studies showing worse prognosis in some pathologies (eg Legionella pneumonia) but nothing that can really be extrapolated to general ED/ICU patients.
What about in ARDS? Talmor et al, (Surgery 1998, v 123) showed benefit in survival in a small study of ARDS patients.
I think the best overall study was Vignon et al (CCM 2005, v 33) who demonstrated that bedside ultrasound was much more sensitive in diagnosing pleural effusion (CXRs read as normal had as much as 1,000 ml effusions, as well as diagnoses of effusions which were actually consolidations ), but also that it was quite accurate in quantifying (inter-pleural distance >45-50mm at the base correlates with effusions >800 ml).
Usta et al (Interactive CV and thor surg, 2010 v 10) had similar results in spontaneously breathing post op patients with 45 mm (interpleural distance between diaphragm and base of lung) corresponding to about 700ml, although their range was quite wide, and they were using a somewhat impractical sitting position. They actually came up with a formula (16xdistance in mm) to estimate effusion volume.
Sikora et al (ISRN Emergency Medicine 2012) did a good review on the whole thing as well.
My empiric observations: I must have put in at least a thousand pigtails or (back in the day) CVCs in pleural effusions over the last 14 years, and my therapeutic goal has always been to drain effusions at least 750 to 1,000 mls. I have drained 350-500 ml in patients in severe respiratory failure a few times, some with apparent benefit, others without. I don’t recall regretting putting one in, but I do regret a few cases of not draining, only to have recurrent respiratory failure post-extubation. Grossly, I would say 750-1,000 or more in a patient with respiratory failure warrants drainage, but 1,000 mls in a CHF patient on nasal prongs can wait for the outcome of diuretic therapy. Of course we are talking about therapeutic and not diagnostic drainage. Under ultrasound guidance with an experienced operator, pneumothoraces should be exceedingly rare.
Bottom Line: start by examining your patient, physically and by bedside ultrasound. Understand his degree of respiratory failure and estimate the size of the effusion and the compliance of his chest. Then you have to make the call (that’s why we’re MDs) as to whether this particular amount of pleural effusion in this particular patient is worth draining.
Next week: how to do it video!
Great post Phillippe and I fully agree that a global assessment is necessary before draining an effusion. With ultrasound we’ve become really good at identifying effusions, but need to understand physiology and especially chest wall/abdominal compliance to understand the physiological implications of drainage vs conservative management.