Global predictors of tracheostomy-related pressure injury in the COVID-19 era: A study of secondary data

To determine the incidence and risk factors of tracheostomy-related pressure injuries (TRPI) and examine the COVID-19 pandemic's impact on TRPI incidence. Secondary analysis of Global Tracheostomy Collaborative database and a multi-center hospital system’s electronic medical records. 27 hospitals, primarily in the United States, United Kingdom, and Australasia. 6,400 adults and 2,405 pediatric patients hospitalized with tracheostomy between 1 January 2019 and 31 December 2021. TRPI as a binary outcome, reported as odds ratios. TRPI incidence was 4.69 % in adults and 5.65 % in children. For adults, associated risks were female sex (OR: 0.64), severe obesity (OR: 2.62), ICU admission (OR: 2.05), cuffed tracheostomy (OR: 1.49), fenestrated tracheostomy (OR: 15.37), percutaneous insertion (OR: 2.03) and COVID-19 infection (OR: 1.66). For children, associated risks were diabetes mellitus (OR: 4.31) and ICU admission (OR: 2.68). TRPI odds increased rapidly in the first 60 days of stay. Age was positively associated with TRPI in adults (OR: 1.014) and children (OR: 1.060). Black patients had higher TRPI incidence than white patients; no moderating effects of race were found. Hospital cluster effects (adults ICC: 0.227; children ICC: 0.138) indicated unmeasured hospital-level factors played a significant role. Increasing age and length of stay up to 60 days are TRPI risk factors. Other risks for adults were female sex, severe obesity, cuffed/fenestrated tracheostomy, percutaneous insertion, and COVID-19; for children, diabetes mellitus and FlexTend devices were risks. Admission during the COVID-19 pandemic had contrasting effects for adults and children. Additional research is needed on unmeasured hospital-level factors. These findings can guide targeted interventions to reduce TRPI incidence and inform tracheostomy care during public health crises. Hospital benchmarking of tracheostomy-related pressure injuries is needed.