• Kate L Brown, Deborah A Ridout, Mike Shaw, Iris Dodkins, Liz C Smith, Maura A O'Callaghan, Allan P Goldman, Susan Macqueen, and John C Hartley.
• Cardiac Intensive Care Unit, Great Ormond Street Hospital for Sick Children, Institute of Child Health, London, UK.
• Pediatr Crit Care Me. 2006 Nov 1;7(6):546-50.

ObjectiveTo describe the use of a multidisciplinary approach to sepsis surveillance and evaluate impact on outcome.DesignProspective clinical study or clinical audit cycle.SettingTertiary pediatric extracorporeal membrane oxygenation (ECMO) center.PatientsPatients were 215 children supported with ECMO January 1999 to December 2004.InterventionsA multidisciplinary team met monthly to evaluate cases of bloodstream infection and mediastinitis, review trends, and update unit policies. Changes in practice were made at the end of 2001 in order to address a perceived high rate of sepsis: a) reeducation; b) introduction of electively preprimed ECMO circuits; and c) preference for neck rather than chest cannulation in cardiac patients. Prophylactic antibiotics were used from preprocedure for 24 hrs only throughout the study.Measurements And Main ResultsOver the entire study period, 39 children had 47 septic episodes, with a rate of 24.9 per 1000 ECMO days. Multiple logistic regression analyses indicated that infection was associated with duration of ECMO support (odds ratio 1.24; 95% confidence interval 1.15, 1.35 per day) and case type: Closed vs. open chest was protective in cardiac patients (odds ratio 0.08; 95% confidence interval 0.01, 0.50). Infection increased the odds of death by 2.01 (95% confidence interval 1.00, 4.05), but this effect was less important than case type and ECMO days. After policy changes were implemented, there was a reduction in sepsis from 29.3 to 20.1 episodes per 1000 ECMO days. There was reduced sepsis in respiratory patients: neonates from 28.0 to 6.6 and pediatric patients from 42.4 to 16.9 episodes per 1000 ECMO days. Despite policy changes, sepsis remained a problem in cardiac patients with open sternum: 65.1 per 1000 ECMO days.ConclusionsECMO support is a high-risk setup for nosocomial infection, in particular for cardiac patients with open sternum for whom antibiotic prophylaxis is justified. Multidisciplinary surveillance offers an excellent approach for quality improvement in this challenging field.

19 keywords...

hide…