Understanding extubation "failures" in an ICU

Peter J. Pronovost, Mollie W. Jenckes, May To, Todd Dorman, Pamela A. Lipsett, Eric B. Bass

Research output: Contribution to journalArticlepeer-review


Failed extubation leads to increased in-hospital mortality, morbidity, costs, and complications. Failed extubation is also a source of liability for hospitals, as it leads to increased patient co-morbidities, and fatalities. Compared to patients who were successfully extubated, patients who "failed" were seven times more likely to die, 31 times more likely to have increased length of stay, and six times more likely to transfer to a long term care facility on discharge. We developed a one page record to be completed by the nursing staff in the ICU at the time of extubation. We asked about frequency of suctioning, secretions, level of consciousness and espiration/ventilation variables. We asked that they complete a record prior to each extubation. For patients experiencing "failed" extubations, we also completed a chart review to collect demographic and severity of illness variables. As this is one of our institution's Performance Improvement (PI) measures, we checked all data with the PI data staff to identify any additional events. We collected information on 245 patients prior to extubation. Of these, 14 had "failures". We identified several patient haracteristics with increased risk for "failed" extubation. These included: Oxygen saturation (96%+, <=95%; OR=4.56, CI 1.13-18.12); Level of consciousness (Alert, sedated/agitated; OR 4.5, CI 1.16, 16.91); secretions1 (thick, thin; OR=2.55, CI 0.74, 9.23); secretions2 (scant; moderate/copious; OR 2.14, CI 0.63, 7.26); cough (strong, weak; OR 1.95, CI 0.58, 6.65). We also identified variables that appeared to protect patients: Suction (>4 hours, every 2-4 hours; OR 0.11, CI 0.01, 0.81); tidal volume (400+, 100-390; OR=).29, CI 0.01, 2.91). We have identified several patient characteristics associated with increased risk of "failed" extubation, including patient alertness, cough, secretions and O2 saturation. We also noted some variables that appear to protect patients from failure. We hypothesized that he latter could be from increased surveillance by nursing staff. We have found that documenting the care process at the point of service delivery provides valuable data to guide physician clinical decision making in an area of common ICU practice.

Original languageEnglish (US)
Pages (from-to)A105
JournalCritical care medicine
Issue number12 SUPPL.
StatePublished - Jan 1 1999

ASJC Scopus subject areas

  • Critical Care and Intensive Care Medicine


Dive into the research topics of 'Understanding extubation "failures" in an ICU'. Together they form a unique fingerprint.

Cite this