Community acquired pneumonia: aetiology and usefulness of severity criteria on admission

Community acquired pneumonia remains an important cause of hospital admission and carries an appreciable mortality. Criteria for the assessment of severity during admission have been developed by the British Thoracic Society (BTS). A study was performed to determine the sensitivity and specificity of a severity rule based on a modification of the BTS prognostic rules applied on admission, to compare severity as assessed by medical staff with the modified rule, and to determine the microbiological cause of community acquired pneumonia in Christchurch.

A 12 month study of all adults admitted to Christchurch Hospital with community acquired pneumonia was undertaken. Three hundred and sixteen consecutive patients with suspected community acquired pneumonia were screened for inclusion. Variables obtained from the history, examination, investigations, and initial treatment were examined for association with mortality.

Two hundred and fifty five patients met the inclusion criteria. Their mean age was 58 years (range 18-97). A microbiological diagnosis was made in 181 cases (71%), Streptococcus pneumonia (39%), Mycoplasma pneumoniae (16%), Legionella species (11%), and Haemophilus influenzae (11%) being the most commonly identified organisms. Patients had a 36-fold increased risk of death if any two of the following were present on admission: respiratory rate > or = 30/min, diastolic BP < or = 60 mm Hg, urea > 7 mmol/l, or confusion. The severity rule identified 19 of the 20 patients who died and six of eight patients admitted to the intensive care unit as having life threatening community acquired pneumonia. The sensitivity of the modified rule for predicting death was 0.95 and the specificity 0.71. In 47 cases (21%) the clinical team appeared to underestimate the severity of the illness.

The organisms responsible for community acquired pneumonia in Christchurch are similar to those reported from other centres except for Legionella species which were more common than in most studies. The modification of the BTS prognostic rules applied as a severity indicator at admission performed well and could be incorporated into management guidelines.