Correlation of Critical Score Models with Outcomes after Tracheostomy in Critically Ill Patients

Objectives: Understand the utility in using the validated outcome prediction models (Sequential Organ Failure Assessment [SOFA] and Logistic Organ Dysfunction System [LODS]) as additional information in making surgical decisions in the specific population of critically ill patients undergoing tracheotomy. Methods: Retrospective chart review of 94 consecutive critically ill patients at a tertiary care Veterans Administration hospital who underwent isolated tracheotomy between January 2006 andJune 2011. Data were collected from the medical charts. Outcome measurements included decannulation, ventilatory status after tracheotomy, and date of death. SOFA and LODS scores were calculated and Students T test was used for statistical analysis. Results: Mortality was expectedly high. Twenty-two point 3 percent of patients died while still inpatients, and 53.2% died within 180 days of tracheotomy. The average time to ventilator independence was 20.1 days, and 28.7% of patients were decannulated. The LODS score was more powerful, showing a significant association with ventilator status 14 days after tracheotomy, death while an inpatient, and death within 90 days of tracheotomy (all P < 0.05). The SOFA score only showed a significant association with death while an inpatient (P < 0.05). Conclusions: This work is the first reported use of the validated LODS and SOFA outcome prediction models in the subset of critically ill patients undergoing tracheotomy. While this patient population is known to have a high mortality rate, the use of objective measurements such as LODS can be useful to help guide clinical decision making, counsel surrogate decision makers, and allocate resources.