Amy Downing1, Richard Wilson, Matthew Cooke. 1. Public Health and Epidemiology, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK. a.downing@bham.ac.uk
Abstract
OBJECTIVE: The objectives of this study are to determine whether it is possible to link ambulance service and Emergency Department (ED) data for assault patients, to look at the potential advantages of this linkage and to investigate the quality of coding in the two data sets. DATA AND METHODS: Data from West Midlands Ambulance Service and seven EDs in the urban West Midlands were linked using probabilistic linkage. The linked data were analysed to investigate demography, priority category, diagnosis, conscious level, disposal and assault coding. PRINCIPAL FINDINGS: 84.2% of the ambulance records were linked to an ED record. Only 40.7% of the linked records were coded as assault in the ED data and only 46.7% of ED assault cases brought by ambulance could be linked. 77.6% of all assault injuries were to the head, face and neck. Only 1.0% of patients presented with coma. 12.0% of all assault patients and 53.5% of the highest priority cases were admitted. CONCLUSIONS: Data linkage is possible and can increase the amount of information available. Data quality problems were identified in both datasets, which has implications for the monitoring and prevention of assaults. The use of a common identifier would aid the following of patient pathways.
OBJECTIVE: The objectives of this study are to determine whether it is possible to link ambulance service and Emergency Department (ED) data for assault patients, to look at the potential advantages of this linkage and to investigate the quality of coding in the two data sets. DATA AND METHODS: Data from West Midlands Ambulance Service and seven EDs in the urban West Midlands were linked using probabilistic linkage. The linked data were analysed to investigate demography, priority category, diagnosis, conscious level, disposal and assault coding. PRINCIPAL FINDINGS: 84.2% of the ambulance records were linked to an ED record. Only 40.7% of the linked records were coded as assault in the ED data and only 46.7% of ED assault cases brought by ambulance could be linked. 77.6% of all assault injuries were to the head, face and neck. Only 1.0% of patients presented with coma. 12.0% of all assault patients and 53.5% of the highest priority cases were admitted. CONCLUSIONS: Data linkage is possible and can increase the amount of information available. Data quality problems were identified in both datasets, which has implications for the monitoring and prevention of assaults. The use of a common identifier would aid the following of patient pathways.
Authors: Peter Miller; Kerri Coomber; James Smith; Michael Livingston; Matthew Stevens; Steven Guthridge; Robin Room; Cassandra J C Wright; Daile Rung; Sarah Clifford; Ryan Baldwin; Sumon Das; Yin Paradies; Debbie Scott; Kalinda E Griffiths; Clare Farmer; Richelle Mayshak; Bronwyn Silver; Sam Moore; Jordan Mack; Vincent Mithen; Danielle Dyall; J Ward; John Boffa; Tanya Chikritzhs Journal: BMJ Open Date: 2022-04-01 Impact factor: 2.692