BTS guidelines for the management of pleural infection
- 1Consultant Microbiologist, Princess Royal University Hospital, Farnborough Common, Orpington, Kent BR6 8ND, UK;
- Correspondence to:
Dr R J O Davies
Churchill Hospital, Headington, Oxford OX3 7LJ, UK;
We have read the BTS guidelines for the management of pleural infection1 and are concerned about the proposed antibiotic choices for the initial treatment of culture negative or pending pleural infection. Section 2.8 of the text and table 2 detail the antibiotic choices but, in our opinion, leave considerable gaps in antibacterial cover against the likely pathogens. In particular:
Amoxycillin (text) is not reliably active against Staphylococcus aureus.
Clindamycin has no activity against Gram negative aerobic organisms (especially not Haemophilus influenzae as mentioned in the text).
Benzyl penicillin (table) rarely now has activity against Staphylococcus aureus and we suggest that relying on ciprofloxacin is unwise. In addition, this combination will not cover many anaerobic bacteria.
We do not consider chloramphenicol is an appropriate agent to use in this category of patients in view of the serious side effect profile.
Third generation cephalosporins such as ceftazidime and cefotaxime have unreliable activity against many anaerobic bacteria.
Pneumococci are considerably less susceptible to ceftazidime than to other cephalosporins and penicillins2; the policy (table), however, suggests its use as a single agent.
Piperacillin (text) is no longer available in the UK except in combination with a β-lactamase inhibitor.
We suggest that the antibiotic choices in the BTS guidelines for the management of pleural infection should be changed to the following:
For community acquired pleural infection, either cefuroxime + metronidazole or co-amoxiclav or (for the penicillin/cephalosporin allergic individuals) clindamycin + ciprofloxacin, all administered intravenously. Oral treatment choices would be co-amoxiclav or clindamycin + ciprofloxacin.
For hospital acquired infection, clinicians should seek guidance from the local microbiologists but the following choices would be reasonable in the interim: piperacillin + tazobactam or cefotaxime/ceftriaxone + metronidazole or meropenem.
Relating to the initial diagnosis of pleural infections, we were also concerned that mycobacteria were mentioned only once in the article. Pleural fluid has a poor yield for diagnosis of tuberculosis and more emphasis should have been placed on the routine use of pleural biopsy for histology and culture of tuberculosis which has much higher diagnostic rates. The algorithm should include the investigation of pleural tuberculosis.3
In conclusion, we would commend the inclusion of a medical microbiologist in discussions leading to guidelines dealing with the diagnosis and treatment of infections.
↵* On behalf of the Outer South East London Microbiology Group: M Viagappan, University Hospital, Lewisham; A Mackay, Queen Elizabeth Hospital, Woolwich; A Teall, Queen Elizabeth Hospital, Woolwich; H Roberts, Princess Royal University Hospital, Bromley; J Kensit, Queen Mary’s Hospital, Sidcup; A Fowler, Queen Elizabeth Hospital, Woolwich; A Shaw, Darent Valley Hospital.
We would like to thank Dr Roberts and colleagues for their interest in the BTS guidelines on pleural infection1 and for their thoughtful letter. In our view, guidelines (particularly the first set in an area) exist partly to stimulate a debate which subsequently better informs care.
Since the advisory regimens in the guidelines were first drafted (and they are “advisory” and to be used in line with local microbiological advice), the microbiology of pleural infection has been greatly clarified (not least by the joint BTS/MRC trial in pleural infection). This same work is also identifying high risk patient groups, clarifying drain type choice, and accurately defining intrapleural treatment. We have no doubt that these new data, as well as some of the points raised by Dr Roberts and colleagues, will strengthen the next revision of the BTS guidelines. The recent data show that only 10% of community acquired infections are staphylococcal, while 50% of hospital acquired infections are due to staphylococcal disease and 66% of these are MRSA infections. Thus, a regimen with limited staphylococcal cover may be appropriate in community infection (although thorough anaerobic cover is needed here), but isolated meropenem in hospital acquired infection (suggested by Roberts and colleagues) would be ineffective for 25% of patients in this setting. Here we might currently favour vancomycin + meropenem (or similar). The BTS/MRC trial suggests that about 50% of patients with hospital acquired pleural infection are currently receiving ineffective empirical antibiotics—emphasising the importance of clarifying this issue. The suggestion of a combination of clindamycin + ciprofloxacin in community acquired infection seems a particularly elegant improvement on the regimen of clindamycin alone advocated in some US centres and mirrored in our original suggestions.
We share the view that, when the pleural infection guidelines are next updated, a microbiologist should be on the drafting panel and not only included during peer review. The drafting panel for these guidelines was a compromise between size and inclusivity in all the therapeutic areas, since it had to cover all the pleural syndromes. This, for example, led to the absence of an oncologist for the malignant effusion guideline and a physician with particular skills in cystic fibrosis for the pneumothorax guideline (again peer review was the chosen method for including these specialists). On the plus side, this led to an efficient guideline generation process. We have previously encouraged the BTS to constitute separate groups for each of the guidelines as they come up for future review for just this reason.