The prognostic ability of early Braden Q Scores in critically ill children

Abstract

Pressure ulcer (PU) risk assessment tools are an important component of good nursing care; however, it is essential that these tools offer a good sensitivity and specificity, in addition to clinical utility in the population being assessed. The aim of this study was to examine how the lowest Braden Q score recorded in the first 24 h of paediatric intensive care unit admission related to the risk of PU development in an English PICU (paediatric intensive care unit). A retrospective cohort study was undertaken over 12 months in a single PICU in the North West of England. 891 critically ill children with a Braden Q score were evaluated. The lowest Braden Q score within the first 24 h of PICU admission was matched to reported PU development and grade. The Braden Q score was found to perform well in children aged 3 weeks to 8 years without congenital heart disease (CHD), which is the population it was validated on. At a cut off score of ≤16 it yielded a sensitivity of 100% specificity of 73.1%, positive predictive value (PPV) 2.56 and a negative predictive value (NPV) of 100 and an area under the curve (AUC) of 0.87(0.75-0.98). When used in other age groups and when it included children with CHD, it performed less well with lower AUC and wider confidence intervals, but it performed moderately well in the group of term to 14 years with a sensitivity of 75% specificity of 72.6%, PPV 1.5 and a NPV of 99.8 and AUC of 0.74 (0.49-0.98). Our results in a heterogeneous UK PICU population found the Braden Q score performed well in the specific population it was validated for (PICU children aged 3 weeks to 8 years without CHD), however, it performed moderately well in the more heterogonous PICU population of term to 14 years including children with CHD.