Discussion
The analysis of this study highlights the relationship between obstructive lung pathology and acute tracheostomy complications. Obstructive pulmonary pathologies, specifically COPD, OSA, and asthma, were more likely to endure acute post-tracheostomy complications with a moderate statistical significance. While the statistical significance was moderate, it does carry clinical implications. COPD is an independent predictor of the need for tracheostomy,20and thus, any increased risk for post-tracheostomy complications impacts this vulnerable population with an already tenuous respiratory status. Tracheostomy has also been shown to improve the cardiovascular endpoints of patients with OSA,21 but it is not without risk. Considering that the chronic obstructive pulmonary patients are more likely to require mechanical ventilation and therefore tracheostomy than patients without these diseases,19 a higher complication rate is a risk that should be considered when making decisions regarding tracheostomy. Ultimately, the benefit from early tracheostomy to avoid prolonged intubation, and its complications, likely outweighs the potential risk of acute post-tracheostomy complications. Therefore, electing for tracheostomy while focusing on the prevention of common complications is warranted.
In this cohort, the overall complication rate was 9% and complication types were consistent with prior studies with bleeding being the most common (33%) followed by tracheitis (15%) and dislodgement (10%). The most common early complication of tracheostomy is bleeding, and prevention of clot descent through the distal airward is essential.22 Tracheostomy infections are also a common complication and maintaining a high clinical suspicion for them is important.23 Additionally, counseling patients on recognizing early signs of major complications as well as close follow up and monitoring is crucial.
One of the secondary outcomes examined in this study was BMI. BMI was not a significant predictor of acute tracheostomy complications in this patient cohort, as shown through the bivariate analysis and multivariable regression models. This is in direct contrast to existing studies that show that obesity, measured by BMI, carries a higher risk of tracheostomy complications.13,14 This could be due to the timeline of complications assessed in different studies. Complications such as unplanned readmissions occur further from surgery while this study specifically looked at the first 14 days of the post-operative period. Therefore, even though this study demonstrated no significant correlation with acute tracheostomy complications, clinicians should be aware of this existing association if a patient has a high BMI, neck circumference, or skin-to-tracheostomy site distance.
The other secondary outcome measure was complication rate among patients with chemotherapy and radiation treatment in the early post-tracheostomy window (post-operative day 0-6). Radiation was found to have a moderate association in multivariable analysis, and chemotherapy and radiation were highly correlated with one another. The increased risk of early complication likely occurred due to poor wound healing considering chemotherapy and radiation are well described causes of poor wound health irrespective of the surgical procedure performed. Chemotherapy frequently slows wound healing, as some widely used chemotherapeutic drugs directly impact cell cycle progression and new tissue deposition. Radiation increases inflammation in local tissues, promoting cell death and slowing wound healing.24 In head and neck cancer patients particularly, many have received neck radiation prior to surgery, impacting the tissues around the tracheostomy site. Slow healing wounds have a higher incidence of infection, recurrent bleeding, and more serious invasive processes including osteomyelitis.25 Considering infection and bleeding are two of the most common complications in the acute post-tracheostomy setting, close monitoring of the tracheostomy site is warranted. It is important to note the limitations of this study. Data collection and medical record review may have been subject to sampling bias, as the cohort was not randomized. Additionally, as a retrospective analysis, there may be bias when interpreting results.