Tracy LM, Dyson K, Mercier LL, Cleland H, McInnes JA, Cameron PA, Singer Y, Edgar DW, Darton A, Gabbe BJ. Injury 2019; ePub(ePub): ePub.
Affiliation
School of Public Health and Preventive Medicine, Monash University, 553 St Kilda Road, Melbourne 3004, Australia; Health Data Research UK, Swansea University Medical School, Swansea University, Singleton Park, Swansea, Wales SA28PP, United Kingdom.
INTRODUCTION: The negative impact of inhalation injuries on in-hospital outcomes for burn patients is well known, but the burns community is yet to form a consensus on diagnostic criteria and clinical definitions. The diagnosis of inhalation injuries is consequently highly subjective. This study aimed to assess the variation in the rate of documented inhalation injury for adult patients in Australian and New Zealand burn units.
METHODS: Data for sequential admissions collected from eight adult burn centres across Australia and New Zealand between July 2009 and June 2016 were extracted from the Burns Registry of Australia and New Zealand (BRANZ). Inhalation injury was classified in two ways: (i) a field in the BRANZ data dictionary, and (ii) through a series of International Classification of Disease 10th Revision Australian Modification (ICD-10-AM) codes. Variation in inhalation injury prevalence was assessed using descriptive statistics, funnel plots, logistic regression, and predicted probabilities.
RESULTS: There were 11,206 admissions to BRANZ sites over the study period. Inhalation injury prevalence was the highest at Site D (13.1% for the BRANZ field and 11.8% for the ICD-10-AM codes), but there was significant variation between the contributing sites and the inhalation injury classification methods.
CONCLUSION: There is significant variation in the prevalence of documented inhalation injury among Australian and New Zealand burns units. The variation in the prevalence of documented inhalation injury across Australian and New Zealand sites reinforces the need for a consensus definition in the diagnosis of these injuries. Further work is required to improve data quality and reconcile the differences between clinical and ICD-10-AM coding prevalence before changes in clinical practice can be recommended from these data.