The Authors Reply: Interpreting Reductions in Unpanned Extubation Rates: Attribution and Denominator Effects

Dear Editor:

Thank you to Chapa-Rodriguez and Narechania for their thoughtful letter to the Editor entitled “Interpreting reductions in unplanned extubation rates: attribution and denominator effects.” The points raised are valid and well stated. We are grateful for the opportunity to reply here.

We utilized unplanned extubations (UEs) per 100 ventilator days to align with the Solutions for Patient Safety Definition.^1–2 Certainly, as the denominator decreases, the risk of an event occurring likewise can decrease. However, the impact on the rate is unpredictable. We noted that there were concurrent ICU liberation initiatives ongoing that helped decrease ventilator days. These likely contributed in some way to the results seen in our quality improvement project, especially our work in analgosedation and extubation readiness.^3–5 This indeed reflects real-world quality improvement and is distinctly different from the sterile design of a randomized controlled trial.

Although we did not adjust for census or case-mix, we feel a strength of our work is the sustainability of our improvement over a 2-year period. Seasonal variation is a major contributor to case-mix in the pediatric ICU. As an example, viral respiratory infections such as the respiratory syncytial virus tend to predominate in the colder months. ⁶ Sustained improvement over 2 years with the associated seasonal variations suggests success across a wide range of illnesses, acuity, and census.

Additionally, the COVID-19 pandemic spanned early 2020 through early 2023. There was a dramatic change in case-mix as well as a drop in admission rates during this time period.^7–9 Meanwhile, our UE/100 ventilator days rate remained high. Since the end of the pandemic and coinciding with our improvement in UE/100 ventilator days, we have seen a return to pre-pandemic norms by some metrics. ¹⁰ Assuming our local trends followed, it is reasonable to suggest that our outcomes coincided with a likely increase in overall illness severity and census numbers.

Although we agree that it is difficult to interpret which bundle elements had the greatest impact on the primary outcome, we propose that it is less about the specific interventions shown in detail. The emerging field of implementation science emphasizes tailoring the approach to the local context with a focus on barriers and facilitators for implementation, or improvement in the specific situation. ¹¹ Rather than the interventions themselves, we contend that the strength was in responding to the uniqueness of our unit, led by a consistent and passionate respiratory therapist champion. We feel that this approach has the best external validity, with specific interventions left to the discretion of local stakeholders and experts who can understand the local context. We hope that our experience helps others build a case for investing in future respiratory therapist leaders in driving quality improvement.