I read with interest the paper by Singh and coworkers on the subject of pleural drainage 1. In the respiratory field, this is a hot topic and guidelines were recently published 2. Singh and coworkers used a 16-French gauge catheter, which is broadly in agreement with the British Thoracic Society guidelines (and this is what we sometimes use in our hospital). However, the British Thoracic Society guidelines 2 do advocate a slightly smaller gauge. The move is certainly away from large bore drains, and I would strongly support this.

Our unit has also achieved good success using 12-French guage drains inserted using a Seldinger technique (Thal-Quick Chest Tubes, Cook Critical Care, Bloomington, IN, USA) in patients on a respiratory ward. They seem to be patient-friendly with little initial morbidity. However, it has been noted that blockage by fibrinous or clotted material does occur. I am surprised that the authors did not encounter this problem, even with the larger drain.

I note the comments on the use of ultrasound guidance to access the pleural space. A recent article 3 reported increased efficacy using ultrasound guidance in the hands of an interventional radiologist, performing the procedure in 'real time'. This seems a useful method but it is a luxury that is often not available. The use of ultrasound to locate the most suitable position for access, using an 'X marks the spot' methodology, has in our experience often been unreliable. I suspect that this is because the pleural access is not performed in 'real time'. I would be interested to know the authors' thoughts on this and whether any of their catheters were inserted using ultrasound guidance.

I would urge caution on behalf of Singh and coworkers in interpreting some of the data. This is a highly selected, small group of patients. In this setting, the absence of infection or blockage is probably of limited value. Our experience has revealed that there are definite, although low, rates of blockage and infection. A larger, prospective, 'real world' study is needed.

Nonetheless, research in this area is always welcome, and the study is of interest to respiratory physicians as well as intensivists.

Competing interests