G147 Reviewing childhood deaths via electronic health records and wider data linkage

Abstract

AimsThe Child Death Reviews Working Group Report (Scottish Government, 2014) recommended that a national review system is implemented across Scotland. NHS Scotland does not currently have one system that captures data of all children who have died. The aim of the present study is to review childhood deaths, within one health board in Scotland, through analysing data from Electronic Health Records (EHR) and wider sources.MethodsA mixed-methods case series analyses approach was adopted within a constant comparison framework. EMIS is an EHR system implemented within this board three years ago. It captures health data of all patients 0–5 years and patients 6 years and over referred to Specialist Children’s Services. EMIS data of deceased patients (0–19 years old) within one Scottish health board, who died between July 2015 and September 2017, were analysed alongside data from wider sources including the Specialist Child Protection Service (SCPS).ResultsBetween July 2015 and September 2017, 156 children (m=87, f=69), aged between 0 and 19 years old were recorded as deceased within EMIS. 57.7% of the children died before their first birthday and 89.7% were known to at least one NHS service. None were known to the SCPS. Preliminary analyses suggest that an age ×environment interaction exists, with the highest death rates observed in infants who resided within decile 1: postcode areas of highest multiple deprivation (SIMD, 2016).ConclusionsThe study represents this health board’s first step in establishing a systematic and robust analytical process to review child deaths via EHR and wider data linkage. The findings demonstrate the potential that this approach offers in building a comprehensive understanding of why a child has died, what risk factors were known and how this knowledge can better protect children in the future. The analyses also suggest that gene ×environment vulnerabilities may exist and that particular age-groups require greater support to protect their survival outcomes. Limitations of the study are acknowledged and recommendations for future research proposed.