• Aspergillus lung disease
• asthma
• non-invasive ventilation
• palliative care
• pneumonia
• respiratory infection
• sleep apnoea
• tuberculosis
• bacterial infection
• innate immunity
• macrophage biology
• viral infection
• pleural disease
• asthma mechanisms
• COPD exacerbations
• COPD mechanisms
• cough/mechanisms/pharmacology

We read Dr Lim and Dr Woodhead's update on the British Thoracic Society (BTS) 2009/2010 community acquired pneumonia (CAP) audit with interest and noted the high inpatient mortality rate of 18.3%.1 As a contributing site, we received a useful summary of our data in comparison with the national data and local mortality rates for severe CAP (CURB65 3-5) were 21.4% versus 42.6% nationally. This provoked an examination of local severe CAP admissions between December 2009 and May 2011 (n=169) that found 25% mortality with age, gender and comorbidity distributions similar to national audit data. We suspect variations in case definitions may be important in understanding differences between local and national data.

The Thorax report focuses on adherence to local antibiotic guidelines. Nationally, 55.5% of patients received antibiotics in line with local prescribing policies (64% in severe CAP), but there was no association between adherence and mortality.1 From our 18-month severe CAP data, over 89% of our patients received antibiotics in line with local policy.

Local policies are assumed to reflect BTS guidelines, which advise intravenous co-amoxiclav plus clarithromycin as the first choice in severe CAP.2 Locally, we recommend intravenous benzylpenicillin plus clarithromycin in severe CAP as advocated by the British Society for Antimicrobial Chemotherapy.3 They suggest that it is ‘likely to apply the least bacterial ecological pressure while still being clinically effective’ and they refer to the ‘collateral damage’ of resistance and Clostridium difficile infection, which results from injudicious broad-spectrum antibiotic use.3 Locally, the spectre of C difficile was a key factor when arriving at our antibiotic policy and although this represents only one of a range of measures we have implemented, our C difficile rates have remained low since adopting this policy.

Studies have shown narrow-spectrum agents to be as effective as broader-spectrum antibiotics in severe CAP.4 A large Australian study comparing benzylpenicillin with ceftriaxone, in combination with a macrolide or doxycycline, found similar outcomes in severe CAP.4 Furthermore, the benzylpenicillin group had lower 30-day mortality (though high-risk patients were more likely to receive ceftriaxone).4 British data in this area are lacking and further high quality studies and indepth interrogation of the BTS audit are desirable.

The BTS audit was excellent but national data should be interpreted carefully, as forming case definitions is difficult. Local data should inform locally appropriate policies, as suggested by Mandell, and this may include narrow-spectrum agents, which combine efficacy with favourable complication profiles.5