Purpose: We previously developed a bedside model (I-TRACH), which used commonly obtained data at the time of intubation to predict the duration of mechanical ventilation (MV). We now sought to validate this in a prospective trial.
Methods: A prospective, observational study of 225 consecutive adult medical intensive care unit patients requiring MV. Utilizing the original 6 variables used in the I-TRACH model (Intubation in the ICU, Tachycardia [heart rate > 110], Renal dysfunction [blood urea nitrogen > 25], Acidemia [pH < 7.25], Creatinine [>2.0 or >50% increase from baseline values], and decreased HCO3 [<20]), we (1) confirmed that these were still predictive of length of MV by multivariate analysis and (2) assessed the correlation between the number of criteria met and the subsequent duration of MV. In addition, we compared the performance of I-TRACH to Acute Physiology Age Chronic Health Evaluation-II and III, Sequential Organ Failure Assessment, and Acute Physiology Score as predictors of length of MV.
Results: Mean age was 62.6 ± 18.7 years, with a mean duration of MV of 5.8 ± 5.7 days. The number of I-TRACH criteria met directly correlated with the duration of MV. Individuals with ≥4 criteria were significantly more likely to require MV >7 and >14 days. Similarly, those who remained on ventilators for both >7 and >14 days met significantly more I-TRACH criteria than those requiring shorter durations of MV (1.7 ± 1.3 vs 2.8 ± 1.3 vs 3.8 ± 1.3 criteria, P < .001). I-TRACH performed better than all other models used to predict the duration of MV.
Conclusion: Similar to our previous retrospective study, these findings validate I-TRACH in determining the subsequent need for MV >7 and >14 days at the time of intubation.
Keywords: I-TRACH; mechanical ventilatory support; medical intensive care unit; predictive model; prolonged mechanical ventilation.