134 PAPERS AND POSTER ABSTRACTS / Australian Critical Care 30 (2017) 109–135 Burden of disease and change in practice in critically ill infants with bronchiolitis in Australia and New Zealand 2002 to 2014 Luregn Schlapbach a,∗ , Lahn Straney d , Ben Gelbart b , Janet Alexander c , Donna Franklin c , John Beca c , Jennifer Whitty c , Subodh Ganu c , Barry Wilkins e , Anthony Slater e , Elizabeth Croston f , Simon Erickson f , Andreas Schibler, on behalf of the Australian & New Zealand Intensive Care Society (ANZICS) Centre for Outcomes & Resource Evaluation (CORE), the Australian & New Zealand Intensive Care Society (ANZICS) Paediatric Study Group a Paediatric Critical Care Research Group, Mater Research, University of Queensland, Australia b PICU, Royal Children‘s Hospital, Melbourne, Australia c University of Queensland d Monash University e PICU Children’s Hospital at Westmead f Princess Margaret Hospital Bronchiolitis represents the most common cause for non- elective admission to pediatric ICUs. There are limited large-scale data on the burden of severe bronchiolitis in critically ill infants. We aimed to characterize the incidence, respiratory support, outcomes, and costs of severe bronchiolitis in ICU. We performed an observational multicenter study of all ICUs contributing to the ANZ Paediatric Intensive Care Registry. Infants <24 months with bronchiolitis requiring admission to ICU 2002- 2014 were included. Intubation and invasive ventilation were defined as primary outcome. Population-based ICU admission rates and associated costs were estimated. During the study period, bronchiolitis was the most common cause of ICU admission and was responsible for 9,628 (27.6%) of 34,829 non-elective admissions. The estimated population-based ICU admission rate due to bronchiolitis increased by 11.76/100,000 each year (95%>CI 8.11 to 15.41). The proportion of bronchioli- tis patients receiving invasive ventilation decreased from 36.8% in 2002, to 10.8% in 2014 (adjusted OR 0.35; 0.27-0.46), whilst a dramatic increase in High-Flow Nasal Cannula therapy (HFNC) use to 72.6% was observed (p<0.001). In an adjusted model, 60.9% of the variation in invasive ventilation was not explained by case mix or temporal trends, likely reflecting underlying differences in unit practice. Annual direct hospitalization costs due to severe bronchi- olitis increased from AU$11.4 M in 2002 to AU$44.3 M in 2014. In conclusion, we observed a major change in practice in the management of severe bronchiolitis with early use of HFNC therapy. Our data suggest that thresholds to admit bronchiolitis patients to ICU have changed over the past decade with a major impact on healthcare-related costs and resource utilization in ICUs. Interventional trials addressing risk stratification and safe manage- ment of bronchiolitis outside ICUs are urgently warranted. http://dx.doi.org/10.1016/j.aucc.2017.02.067 Lactate and non-lactate components of base excess improve the discrimination of the paediatric index of mortality (PIM2) model Barry Wilkins ∗ , Victoria Ward ∗ PICU, Children’s Hospital at Westmead, Sydney, Australia Absolute base excess (BE), one of 10 variables in the PIM2 score, measures metabolic acid-base status, including lactate and other strong ions. We aimed to test if separating BE into lactate, chloride, albu- min and phosphate components, or into lactate and (BE + lactate) improved PIM performance in a single Paediatric Intensive Care Unit. Standard PIM variables including absolute BE (allowing venous samples), lactate, chloride, albumin and phosphate were analyzed in custom models, using SAS v14.1, to determine association with mortality. Univariate and multivariate analyses were performed. P values and 95% confidence intervals were calculated. Receiver operator characteristic (ROC) curves were constructed to analyze the discriminative performance of these models. All admissions (n=11,910) 2006-2015 inclusive were analyzed. All 10 PIM2 variables were significant in univariate analysis (p<0.0001) but recovery and cardiopulmonary bypass lost signif- icance in this multivariate custom model. Lactate, absolute (BE + lactate), chloride, albumin and phosphate were also significant. Allowing venous BE values increased the area under the ROC curve (AUC) from 0.9038 to 0.9098. Replacing absolute BE by lactate increased the ROC area further to 0.9109, and then adding its non- lactate components (BE + lactate) to 0.9147. Using three of the non-lactate components separately, chloride, albumin and phos- phate increased ROC area to 0.9119, 0.9170 and 0.9192 respectively. Although albumin and phosphate are not candidates as PIM vari- ables because of infrequency of first-hour measurement, replacing BE with lactate and either absolute (BE + lactate) or chloride improves discrimination in this custom model. A prospective multi- centre study is required to establish whether these variables should be included in later PIM models. http://dx.doi.org/10.1016/j.aucc.2017.02.068 Paediatric Critical Care Posters Improving clinical handover in retrieval Tina Kendrick ∗ , Jeanette Marchant NETS NSW, Westmead, Australia Clinical handover was identified as a root cause of major health- care incidents across NSW and thus mandated for change. NETS NSW is the dedicated retrieval service for newborns and children in NSW and the ACT. Handover is fundamental for decision-making in retrieval where direct clinical contact is not possible. A patient retrieval involves multiple handovers. It is essential information is exchanged in a coordinated and systematic way to minimise risk of error from poor communication to enhance safe, timely patient transfer. Literature on optimising retrieval handover is sparse. Aims were to standardise and improve clinical communication during handover. A four phase study was conducted: data gathering from focus groups and observations of handovers; tool development; implementation of tools evaluation. Focus groups revealed retrieval clinicians experienced inadequate handovers from referring cli- nicians during the initial call to NETS when retrieval decisions are made. Inexperienced and junior clinicians called NETS with limited patient information, without a structured handover tool, and no guide for information NETS required. When conferencing