Are the Australian guidelines asking too much of the Pneumonia Severity Index (PSI)?

To the Editor: We agree with Buising and colleagues1 that, in terms of predicting clinical outcomes, the Pneumonia Severity Index (PSI) developed by Fine and colleagues2 is heavily weighted towards age and pre-existing comorbidities. However, we disagree with both their proposed “solution” and the concept on which it appears to be based.

Although the current Australian antibiotic guidelines suggest that admission to the intensive care unit should be considered mainly for patients with class V community-acquired pneumonia (CAP),3 a review of the data of Fine and colleagues suggests that patients in both class IV and class V are most likely to need this type of care. In the PSI’s validation cohort of 38 039 patients, 73% of those requiring intensive care fitted these classes.2 Thus, by simply modifying the current antibiotic guidelines to include patients with CAP in either class IV or V as being at greatest risk of needing intensive-care admission, the recommendations would be accurate.

By comparison, Buising and colleagues advocate using the modified British Thoracic Society (BTS) rule, which was validated in only 244 patients.4 While this approach may have some future merit, we believe there are insufficient data to advocate its use at present. A comparison of the PSI and original BTS criteria found that PSI classes IV and V were more sensitive at predicting need for intensive-care admission.5

Secondly, we are concerned about the suggestion by Buising and colleagues that young patients with severe CAP who are not in PSI class V could have worse outcomes if they do not receive broad-spectrum antibiotics.1 This implies that severe CAP is more likely to be due to unusual or resistant pathogens. This is not supported by available evidence. Instead, early clinical consideration of the likely pathogens and the potential use of new diagnostic “point of care” tests (eg, pneumococcal and Legionella urinary antigen assays and analysis of throat swabs by polymerase chain reaction for respiratory viruses and “atypical” pathogens) are likely to be of greatest benefit in empirical antibiotic prescribing.

Although CAP is a common admission diagnosis, there are very few published Australian studies defining its aetiology, optimal treatment and clinical outcomes. We are currently undertaking a large prospective study (the Australian Community-Acquired Pneumonia Study) at six major hospitals in three states to address these issues. Results should be available in late 2005.