Here at St. Emlyn’s we have been managing patients with pneumothoraces as out patients for many years. The approach in the UK has been ‘guided’ by the BritishThoracic Society guidelines.
For many PNXs we simply aspirate, and if successful we discharge the patient back to community and follow them up in clinic with serial chest radiographs.
In recent years we have started to question this approach. We have gone over to using very small (8 and 10 French) chest drains for simple pneumothoraces which are far less invasive than the old ATLS methods of finger sweep insertions. Arguably the insertion of a small Seldinger guided chest drain is a relatively minor procedure, only slightly more uncomfortable than having a pleural aspiration – but of course you then have to admit the patient as they have a chest drain in as they are connected to an underwater drain……or do you?
Arguably you don’t. There is an alternative and that is to place a Heimlich valve on the end of the chest drain to maintain a negative pressure seal. This is not something that appears in the BTS guidelines, but there are clearly people doing it.Could we then safely discharge patients with a small chest drain and a Heimlich valve in place? It’s a good question that Fraser Brims and Nick Maskell have tried to answer in Thorax.
[DDET “So what was studied here?”] This is a systematic review but not a meta-analysis. This is wise as the trials found are really quite different and pooling data in a meta-analysis would have been unwise. The focus was on papers looking at the use of Heimlich valves in the treatment of pneumothoraces. RCTs, case control and case series were included so its a mixed bag of quality and design[/DDET]
[DDET “Have they got all the data?”] The ‘method’ in a systematic review largely comes down to getting hold of the right trials so we want to see a good search strategy and attempts to go through the grey literature. This is achieved well with good evidence of a robust electronic search backed up by hand searching through other resources.[/DDET]
[DDET “Data quality”] The quality of the data relates to the trials themselves and the authors rightly assess the overall quality of trials as ‘variable’. Of the 18 trials only one is rates as very good, 2 are good, 7 are moderate and the rest poor.
Only two trials are RCTs (with a total of 80 patients in them).
So, the authors are limited by the quality of the data they have. This means that any conclusions they draw are going to be limited. [/DDET]
[DDET “Is there an overall message here?”] Tricky this. With such variability in the papers and quality the authors are rightly cautious in their conclusions. Despite this the authors do pool data in a single table, which with such variability is ‘courageous’ in my opinion. I think we have to be very cautious about such pooling. The general principle should be that you can’t take a whole load of information from different sorts of trials, add them all together and then expect to have a reliable result. Despite this I cannot help myself from looking and noticing that the success rates for the successful treatment of using a Hemlich valve only are pretty high (85.8%). With such cautions can we believe this number? No we cannot, but perhaps it is enough to interest us to do better studies in future.
They have also looked at the overall number of complications in a similar way, and therefore I have the same concerns, but having said that the incidence appears to be low. Looking across the 1235 patients in all the studies there are no life threatening complications. That’s reassuring to a degree and again enough for us to consider whether more research is worth pursuing, It’s certainly reassuring to anyone doing this already or considering it in the future.[/DDET]
[DDET “Unanswered questions”] This paper has got me thinking about a number of issues in my own practice which I don’t think this study can answer.
- 1. Are the patients in these trials ‘aspiration failures’ or were they patients primarilly treated with chest drain and HV?
- 2. The lack of a powered RCT means that we are relianbt on a lot of observational data which is likely biased.
- 3. The studies range from 1973 to 2011. Whilst the Seldinger technique was described back in 1953 chest drain insertion using the technique is a relatively recent technique in the UK.
- 4. Can we do this as an out patient therapy? The authors suggest that this should be possible in selected patients and it certainly seems as though it is possible from the data provided.
So there is much to think about here.[/DDET]
[DDET “Final thoughts”] My final thoughts are that this is something that I would really like to offer to my patients, but the feeling is that the published evidence given in this trial is insufficient to make that leap just yet. They conclude that an adequately powered RCT is required to determine reliable outcome data. I have to agree, so if there are any budding triallists out there I’d be jolly grateful if you could include an out patient treatment arm of the trial.
I would also like to thanks the Lee family who kindly gave permission for the use of this photo from their website. Not only is at a great photo but it’s also a very interesting story from a patient’s perspective about what it feels like to have a pneumothorax, a chest drain, hemlich valve, underwater seal and then surgery. Joss certainly went through the full gamut of procedures and I really think their experiences are worth a read (an amazing family).
Before you go please don’t forget to…