The UK-ESPE study: paediatric empyema in the UK

Aims The incidence of paediatric empyema has been rising in the UK. The UK-ESPE study aims to monitor the changing epidemiology of paediatric empyema and to compare empyema management across the UK. Methods Demographic, clinical and microbiological data on children with empyema undergoing pleural drainage were obtained from 19 paediatric respiratory centres from September 2006 until March 2011. Culture negative samples of pleural fluid were investigated using pneumococcal PCR. Pneumococcal PCR positive samples then underwent non-culture serotyping using an assay capable of detecting 14 serotypes/groups. Robust multivariate survival models were used to analyse length of stay (LOS). Chi-squared, Kruskal-Wallis and Fisher's test were used for other analyses. Results Data on 635 children were obtained (56% male, median age 4.3 years). Where a microbiological cause was identified (n=287), Streptococcus pneumoniae accounted for 68%, Streptococcus pyogenes 14% and Staphylococcus aureus 4%. A pneumococcal serotype was detected in 125 patients; of these serotype 1 accounted for 46% and other prominent serotypes were 3 (22%), 7 (12%) and 19A (9%). Serotypes covered by the 7 valent conjugate vaccine (PCV-7) accounted for 48% of serotypes detected in 2006/07 but had disappeared by 2010/11. Non-PCV-7 serotypes increased over the same period. When comparing severity between serotypes, 19A was associated with severe disease and the only reported death. Median tertiary LOS was 8 days (range 3-33) and median total hospital stay (THS) 11 days (range 5-43). Surgical methods (open thoracotomy or video-assisted thoracoscopy) were associated with a decrease in tertiary LOS of 41% (95% Confidence intervals 7-88%, p=0.015) and THS of 43% (95% CI 7-92%, p=0.015) when compared to non-surgical methods (thoracocentesis +/- fibrinolysis). This effect did not achieve significance when each treatment group was analysed separately. Antibiotic choice in the tertiary centre did not significantly change either LOS measure. Conclusions Pneumococcus was the most commonly identified cause of paediatric empyema in the UK. The incidence of causative serotypes has evolved over time. Continuing surveillance is required to monitor changes following the introduction of the 13 valent vaccine in 2010. Primary surgical strategies were associated with reduced hospital stay in this cohort but this needs further investigation.