• Emily Bebbington, Joanna Miles, Amber Young, van BaarMargriet EMEDutch Burn Repository R3, Association of Dutch Burn Centres, Maasstad Hospital, Maasstadweg 21, 3079 DZ Rotterdam, the Netherlands., Nicole Bernal, Ragnvald Ljones Brekke, Lotte van Dammen, Moustafa Elmasry, Yoshiaki Inoue, Kara A McMullen, Lia Paton, Oliver C Thamm, Lincoln M Tracy, Nukhba Zia, Yvonne Singer, and Ken Dunn.
• Centre for Mental Health and Society, Bangor University, Wrexham Academic Unit, Technology Park, Wrexham LL13 7YP, UK. Electronic address: e.bebbington@bangor.ac.uk.
• Burns. 2024 May 1; 50 (4): 850865850-865.

IntroductionPooling and comparing data from the existing global network of burn registers represents a powerful, yet untapped, opportunity to improve burn prevention and care. There have been no studies investigating whether registers are sufficiently similar to allow data comparisons. It is also not known what differences exist that could bias analyses. Understanding this information is essential prior to any future data sharing. The aim of this project was to compare the variables collected in countrywide and intercountry burn registers to understand their similarities and differences.MethodsRegister custodians were invited to participate and share their data dictionaries. Inclusion and exclusion criteria were compared to understand each register population. Descriptive statistics were calculated for the number of unique variables. Variables were classified into themes. Definition, method, timing of measurement, and response options were compared for a sample of register concepts.Results13 burn registries participated in the study. Inclusion criteria varied between registers. Median number of variables per register was 94 (range 28 - 890), of which 24% (range 4.8 - 100%) were required to be collected. Six themes (patient information, admission details, injury, inpatient, outpatient, other) and 41 subthemes were identified. Register concepts of age and timing of injury show similarities in data collection. Intent, mechanism, inhalational injury, infection, and patient death show greater variation in measurement.ConclusionsWe found some commonalities between registers and some differences. Commonalities would assist in any future efforts to pool and compare data between registers. Differences between registers could introduce selection and measurement bias, which needs to be addressed in any strategy aiming to facilitate burn register data sharing. We recommend the development of common data elements used in an international minimum data set for burn injuries, including standard definitions and methods of measurement, as the next step in achieving burn register data sharing.Copyright © 2024 The Authors. Published by Elsevier Ltd.. All rights reserved.

17 keywords...

hide…