Can outcomes of intensive care unit patients undergoing tracheostomy be predicted?

Abstract

Objective: To determine whether outcomes (mortality and need for intensive care unit [ICU] readmission) of patients undergoing tracheostomy in the ICU can be predicted by common clinical or historical criteria.

Methods: We conducted a retrospective review of data from the medical record and Project Impact database in a 24-bed medical-surgical ICU in a 500-bed university hospital. In 2004 through 2006, 60 adult patients underwent tracheostomy as part of their ICU management. We classified each patient as either not readmitted, readmitted, died on floor (after ICU discharge), died on first ICU admission, or combined readmitted/died-on-the-floor. Patients who died on the regular floor were significantly heavier than patients discharged without need for readmission (P = .03). Patients with a history of sepsis and those with a history of neurological disease had a tendency toward worse outcomes, but these did not reach statistical significance.

Conclusions: These findings suggest that it is difficult to predict outcomes of patients who undergo tracheostomy in the ICU. Larger and prospective studies may help elucidate this matter.