Wide Variation in Unplanned Extubation Rates Related to Differences in Operational Definitions.

OBJECTIVES

Unplanned extubation (UE) rate is a patient safety metric for which there are varied and inconsistently interpreted definitions. We aimed to test the sensitivity of UE rates to the application of different operational definitions.

METHODS

We analyzed neonatal intensive care unit (NICU) quality improvement data on UE events defined inclusively as "any extubation that was not performed electively, or not previously intended for that time." Unplanned extubations were classified as involving an endotracheal tube (ETT) that was either objectively "dislodged" or "removed" without proof of prior dislodgement. We used descriptive statistics to explore how UE rates vary when applying alternate UE definitions.

RESULTS

For 33 months, 241 UEs were documented, 70% involving dislodged tubes and 30% ETTs removed by staff. Among dislodged ETTs, only 9% were found completely externalized, whereas 77% were at an adequate depth but in the esophagus. Thirteen percent of events occurred outside the NICU and 13% were initially unreported. The overall UE rate was 4.9/100 ventilator days. If the least inclusive definition was used (i.e., counting only "self-extubations" by patients, requiring reintubation, and occurring within the NICU), 83% of UEs would have been excluded.

CONCLUSIONS

Most UEs in our NICU population involved staff either removing ETTs from the trachea or partly removing them after internal dislodgement. In settings where ETTs removed by staff are not counted, UE rates may be substantially lower and associated risks underestimated. An inclusive, patient-centric operational definition along with a standardized classification would allow benchmarking, while enabling targeted approaches to minimize locally predominant causes of UEs.