Paediatric tracheostomy: Can we predict time to decannulation?

Abstract

Introduction

Paediatric tracheostomy is associated with a significant burden of care on health systems and primary caregivers. Decannulation is regularly part of the discussion during insertion. Being fully aware of factors that affect successful decannulation and time to decannulation is important for consenting primary caregivers, but there is a paucity of comprehensive evidence.

Aim

Our objective was to try and establish important predictive criteria.

Methods

A retrospective review of all tracheostomy patients in a national tertiary paediatric referral centre over 20 years with data on factors drawn from existing literature and additional factors significant for paediatric otolaryngologists was performed. Univariable and multivariable analysis for decannulation success and cox regression analysis on time to decannulation was performed. We included all patients that were cared for at our facility and excluded only patients that had incomplete data available.

Results

153 patients were included for analysis. Corrected age at insertion of less than 6 weeks and sole tube feeding were found to be significant predictors of successful decannulation on univariable analysis, with maintenance of significance of feeding route on multivariable analysis. There was no significant effect on decannulation success with presence of comorbidities, long term ventilation, gestation, emergency status, EXIT procedure or requirement for intervention pre-decannulation on univariable analysis. There was no significant difference in time to decannulation with sole upper airway obstruction or long-term ventilation, mixed oral and tube feeding route, cardiac, respiratory, or genetic comorbidities, or requirement for intervention pre-decannulation. Significantly shorter time to successful decannulation was seen with age at insertion of greater than 12 months, while a significantly longer time was seen with sole tube feeding route, age of less than six weeks at insertion, indication for insertion of a combination of upper airway obstruction and long-term ventilation, neurological co-morbidities and two or more comorbidities on Cox regression analysis.

Conclusion

Our analysis demonstrated feeding route to be a significant predictor of decannulation success on multivariable analysis. We also demonstrated multiple predictors of time to decannulation. These findings should aid in the discussion around decannulation with primary caregivers.