Patient characteristics, incidence, technique, outcomes and early prediction of tracheostomy in the state of Victoria, Australia

Background: Tracheostomy is a relatively common procedure in Intensive Care Unit (ICU) patients. Aims: To study the patient characteristics, incidence, technique, outcomes and prediction of tracheostomy in the State of Victoria, Australia. Methods: We used data from the Australian and New Zealand Intensive Care Society Adult Patient Database (ANZICS APD) and the Victorian Admitted Episode Dataset (VAED) to identify and match patients who had received a tracheostomy from 2004 to 2014. Results: Between 1st January 2004 and 30th June 2014, 9750 patients received a tracheostomy with 7670 available for matching and 6010 (78.4%) successfully matched. Of the matched tracheostomy patients, median age was 61 years, median APACHE IIIJ score was 66 and overall hospital mortality was 21%. The incidence of tracheostomy almost halved over the decade with more than half of tracheostomies (53.5%) being percutaneous. Hospital mortality of patients receiving a tracheostomy decreased from 26.5% in 2004 to 16.5% in 2014 by an average decrease of 6%/year. No robust model could be developed to predict tracheostomy. Conclusion: The incidence of tracheostomy and the adjusted mortality rate of patients who received a tracheostomy have significantly decreased over a decade. Day of admission information could not be used to predict subsequent tracheostomy.