We have previously blogged on the topic of chest tube/drain management1. In one of our most widely read posts we argued that many pneumothoraces can be managed without a chest drain. When trained as a junior doctor in Virchester the presence of a pneumothorax in a trauma patient was an absolute indication for a chest drain, since that time we have changed. We CT a lot more patients than we ever did and many of the pneumothoraces we diagnose are only seen on CT2. There is increasing evidence that these patients can be managed without a chest drain if patients are closely observed3,4.
For awake patients this is relatively straightforward, but there is a dilemma for the patient who is then going to be (or who already is) on positive pressure ventilation. Should they get a chest drain as a matter of course as some state if an occult (as in not visible on CXR and not clinically detectable) pneumothorax is detected? To my knowledge the RCT has not been done and many of the observational studies have small numbers of ventilated patients. There is some weak observational data out there 5–9 that a proportion of patients can be managed conservatively, even if ventilated, but only if very closely observed. It’s almost certain that we did this many times in the past too. I have no doubt that I used to ventilate many patients with occult pneumothoraces as we never used to CT them (so did not know). I cannot imagine that they are a new phenomena, it’s just that we now know that they are there.
This week we look at a UK paper that provides some background and evidence from TARN on how these changes are influencing practice in the UK. The paper was published in CHEST10 earlier this year and you can read the abstract below. As we always say you should go and read the full paper online and make your own mind up on the evidence.
What kind of paper is this?
This is an observational paper based on the TARN database11. TARN is a well established prospective audit system with wide coverage of the UK trauma networks.
What did they do?
The authors looked at data between 2012 and 2016 from a single UK trauma centre reviewing all patients who were reported as having a pneumothorax or haemothorax. They matched that data to hospital records to find out how many patients received chest drains, together with the characteristics of physiological derangement and anatomical injury.
What did they find?
602 patients were included in the study. The average ISS of 26 represents a cohort with significant injury.
277/602 (42%) of patients did not get a chest drain at the point of diagnosis. 255 of those never got a chest drain and continued with conservative management. In other words 10% of them needed a chest drain later. Of note, of the patients who underwent PPV the proportion managed conservatively was the same, i.e. 90%.
Failure of conservative management was a result of expanding pneumothorax or haemothorax. Interestingly there was a 10% complication rate in the patients who were not managed conservatively which is a useful reminder that chest drainage is not without complication.
The authors also looked to see if there were any other characteristics that would predict a failure of conservative management. They looked at a range of characteristics including sex, ISS, rib fractures, PPV, GCS, Distress, Size of PTX, but none affected the failure rate except a haemothorax of >2cm (presumably measured at the hilum?). If so then conservatively managing an HTX of that size might be considered a bit sporting by some.
Should this paper change practice?
There are a number of issues with this paper that we need to think about. It is a single centre study, and although of a reasonable size we should always be cautious of single centre studies. For example the number of penetrating trauma patients seen in this cohort is much lower than that we see in Virchester. However, the findings here are in keeping with other studies including a large review paper published in Annals of Emergency Medicine this month12.
We must be mindful that the patients at baseline are very different. We don’t really know the characteristics of the patients at the initial decision to place a drain or not and I suspect it will be variable between teams and decision makers. Similarly we don’t really know enough about treatment failure. I would want to know how many needed emergent intervention for treatment failure.
TARN eligibility may also miss some patients with very small, isolated pneumothoraces who might have been missed. However, they are likely to be very low risk and if reattended for further management they should still have been entered into TARN (unless they ended up in another country which is rather unlikely).
The initial decision to place a chest tube was made by the clinical team on the day. Such decisions are to some extent subjective and thus the criteria for what could or could not be considered for conservative management is unclear and almost certainly variable. This is not an RCT with specific entry criteria and so I do find it a little difficult to assure myself that I would make the same decisions as the teams in this paper.
Perhaps the most interesting aspect of this paper is the management of patients undergoing positive pressure ventilation (PPV). In that group there has been increased anxiety about leaving patients without a chest drain but there is evidence here and in other papers that these can be managed conservatively if closely monitored. If you stop and think about this then we always have done. We know that we now spot pneumotharaces on CT that we used to miss on plan chest X-ray and in the days before routine CT we clearly and obviously managed them without chest drains.
In truth the changes in practice seen in this paper probably reflect practice across the UK (anecdotal comments from colleagues and practice in Virchester anyway).
The bottom line.
This study, in keeping with others reassures me that many patients with traumatic pneumothoraces and haemothoraces can be managed conservatively even if on PPV. However, we cannot be complacent, all patients managed conservatively should be closely observed in an appropriate clinical area.
Simon Carley @EMManchester