Article Text

Download PDFPDF

S114 Predictors of primary treatment failure in children with empyema
  1. ES Ward-Booth1,
  2. DA Spencer2,
  3. JY Paton3,
  4. SP Rushton4,
  5. M Brodlie1,
  6. MF Thomas1
  1. 1Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK
  2. 2Paediatric Respiratory Unit, Great North Children’s Hospital, Newcastle upon Tyne, UK
  3. 3School of Medicine, Glasgow University, Glasgow, UK
  4. 4School of Biology, Newcastle University, Newcastle upon Tyne, UK


Introduction and objectives Empyema is a relatively rare but severe form of respiratory infection in childhood associated with significant morbidity. A decade long prospective observational study of paediatric empyema – the UK-ESPE study – completed recruitment in April 2016. This preliminary analysis focused on identification of predictors of primary treatment failure.

Methods Data were available from 23 UK centres on 542 patients aged 0–16 years, admitted between December 2006 to December 2015, who were diagnosed with empyema and required pleural drainage. A standard logistic regression approach was used to evaluate the influence of age, sex, pre-hospital illness length, total illness length, time from admission to pleural drainage, length of pleural drainage, length of hospital stay, primary treatment modality and co-morbidity on the risk of primary treatment failure. Variables were investigated using univariate analysis and those that were significant were included in a multivariate model. Stepwise removal was then used to generate a parsimonious final multivariate model. The primary outcome measure was treatment failure as defined by a second procedure carried out >24 hours after the primary pleural drainage procedure.

Results The study group was 56% male, median age was 4 years (IQR 2–7). Recorded pre-hospital illness length was a median of 7 days (IQR 4–10) and total hospital length of stay 12 days (IQR 9–16). 34.7% of children had a co-morbidity recorded. Baseline characteristics were similar in children who did (n = 37) or did not fail (n = 505) primary treatment. There was substantial variation in the choice of primary treatment; chest drainage alone (20.1%), chest drainage with fibrinolytics (58.4%), video-assisted thoracoscopic surgery (12.1%), decortication (8.9%) and other treatments (0.3%). Treatment failure occurred in 6.9% of children and no demographic factor was found to influence treatment failure. The only significant predictor of treatment failure was primary method of treatment (Table). The relative risk of treatment failure was highest in children who received chest drainage alone (OR = 3.04 (95% CI: 1.47–6.27)).

Conclusions In this large cohort no predictors of treatment failure other than primary treatment modality were identified and those who had chest drainage alone had the highest risk of failure.

Abstract S114 Table 1

Parsimonious multivariate model of potential predictors of treatment failure

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.