Study highlights regional variation in mechanical ventilation discontinuation strategies

Invasive mechanical ventilation discontinuation practices regarding patient screening, spontaneous breathing trial conduct, ventilator modes, protocols and clinician roles vary significantly by region, researchers reported in a new study.

“Cumulative exposure to invasive mechanical ventilation has been associated with exposure to harmful co-interventions, increased morbidity and long-term functional and cognitive sequelae. Thus, limiting the duration of invasive mechanical ventilation has been identified as a key research priority for critical care,” Karen E. A. Burns, MD, associate professor of medicine at the University of Toronto and critical care medicine specialist at St. Michael’s Hospital, Toronto, said during a presentation at the virtual Society for Critical Care Medicine’s Critical Care Congress. “Randomized trials, to date, have evaluated the use of screening protocols to identify patients for a spontaneous breathing trial, alternative spontaneous breathing trial techniques and spontaneous breathing trial durations, and strategies to liberate patients who failed the initial spontaneous breathing trial. Notwithstanding, discontinuation involves diverse clinicians whose work profiles differ across settings and may be influenced by factors related to patients, institutions and care processes.”

Burns and colleagues conducted a study, with a primary objective to characterize regional variation in discontinuation practices regarding use of daily screening, spontaneous breathing trial (SBT) conducted, preferred ventilator modes used before discontinuation attempts, use of written protocols to guide care during weaning, and sedation and mobilization practices. The researchers also aimed to examine the association between the initial invasive mechanical ventilation discontinuation strategy and clinical outcomes, including duration of mechanical ventilation, ICU and hospital mortality, length of stay, reintubation, readmission to ICU, and the proportion of patients on mechanical ventilation and in the ICU at day 28; the association between clinical outcomes and initial SBT outcome and initial SBT timing; and factors associated with use of selected discontinuation strategies and initial SBT failure.

The study included 1,868 patients (mean age, 60 years; 63% men) who received invasive mechanical ventilation for at least 24 hours at 142 ICUs in Canada, India, the U.K., Europe, Australia/New Zealand and the U.S. Researchers collected data on at least 10 discontinuation events and all deaths before mechanical ventilation discontinuation attempt. Patients were followed until successful extubation/disconnection in those with a tracheostomy, death, ICU discharge/transfer or until day 60 for ventilator-dependent patients.

SBT was defined as a focused assessment of a patient’s capacity to breathe spontaneously with any type of technique with the aim of discontinuing invasive mechanical ventilation. Sedation levels were characterized by sedation agitation scale scores: low (1 to 2), mid-sedation (3 to 5) and higher (6 to 7).

Regional variations

Overall, 364 (19.5%) patients died before a discontinuation event. Additionally, 424 (22.7%) patients underwent a direct extubation, 150 (8%) underwent a direct tracheostomy and 930 (49.8%) underwent an initial SBT. Among those who underwent an initial SBT, 169 (18.2%) failed the initial SBT attempt, according to Burns.

The researchers reported significant regional variation in the conduct of daily screening, with screening once daily to undergo an SBT in more than half of the ICUs in Canada, India and the U.S. compared with less than half of ICUs in Europe, U.K. and Australia/New Zealand (P < .001). Screening twice daily (range, 3.7% to 19.2%) or more than twice daily (range, 0% to 26.9%) was less frequent across all regions. Results also showed significant regional variation in the use of written directives for daily screening for the conduct of SBTs and for early mobilization as well as differences in the personnel most responsible for conducting SBTs and adjusting ventilator settings (P < .001 for all).

Pressure support plus positive end-expiratory pressure and T-piece were the most commonly used SBT techniques across regions, used in 49.1% and 25% of patients, respectively, according to the results. Pressure support plus PEEP was more frequently used in Canada, the U.K., Australia/New Zealand and the U.S. Pressure support was also the most used ventilator mode across regions (74.5%). In 48.4% of patients, pressure support was most used before an initial SBT (P < .001). T-piece SBTs were more frequently used in India and Europe, Burns said.

Clinical outcomes

Burns reported a “wide variety” of different outcomes with respect to the discontinuation strategy utilized and clinical outcomes.

Patients who underwent a direct tracheostomy had less favorable clinical outcomes, with 34% still in the ICU by day 28, 25.3% still on mechanical ventilation by day 28, mortality of 15.3% and an average of 20 days spent in the ICU and 35 days spent in the hospital, according to results presented.

Patients with an initial SBT spent more time on mechanical ventilation (4.1 days vs. 2.9 days; P < .001), more time in the ICU (8.1 days vs. 6.7 days; P < .001) and had higher ICU mortality (10.3% vs. 4.7%; P = .006) compared with patients with a direct extubation, according to the results.

Patients with an initial SBT failure spent more time on mechanical ventilation (6.1 days vs. 3.5 days; P < .001), more time in the ICU (10.6 days vs. 7.7 days; P < .001), had higher ICU mortality (17.2% vs. 8.8%; P = .03), and at day 28 were more likely to be in the ICU (15.7% vs. 8%; P = .008) and on mechanical ventilation (11.7% vs. 4.1%; P = .001) compared with patients with an initial SBT success, according to the results.

Patients who underwent late initial SBT spent more time on mechanical ventilation (6.1 days vs. 2.1 days; P = .001), more time in the ICU (10.8 days vs. 5.9 days; P < .001), more time in the hospital (22.8 days vs. 14.3 days; P < .001) and at day 28 were more likely to be in the ICU (12.6% vs. 6.3%; P = .01) and on mechanical ventilation (7.7% vs. 3.3%; P = .02) compared with patients who underwent early SBT, according to the results.

Patient factors associated with conducting an initial SBT included older age (P = .02) and the presence of comorbidities including hypertension (P = .03) and malignancy (P = .01). ICUs in regions outside of the U.S. were significantly less likely to conduct an initial SBT compared with a direct extubation. Factors associated with an initial SBT failure vs. an initial SBT success included a higher Sequential Organ Failure Assessment (SOFA)-respiratory score (P = .005) and a lower sedation agitation scale score (P = .02). Lack of a ventilator weaning protocol (P = .03) was also associated with initial SBT failure, according to the results.

In an exploratory analysis, the researchers observed an association between sedation level and probability of SBT failure. Patients with low sedation levels were 2.4 times (95% CI, 1.3-4.5; P = .006) more likely to fail an initial SBT compared with patients with mid-range sedation levels.