Providers Consistently Delay Extubation after Successful Spontaneous Breathing Trials: A Retrospective Cohort Study

Abstract

Rationale

Many patients who pass a spontaneous breathing trial (SBT) are not extubated, leaving them at risk for life-threatening ventilator-associated complications. Determining barriers to timely extubation may facilitate shorter overall durations of mechanical ventilation.

Objectives

We sought to identify the patient-related and patient-independent barriers to timely extubation within 6 hours of passing an SBT.

Methods

We analyzed electronic health record data from adult patients on mechanical ventilation who had been admitted to a medical or cardiac intensive care unit at an academic, tertiary-care center between January 1, 2015, and December 31, 2023. We utilized a mixed-effects, multivariate logistic regression model to evaluate the association between timely extubation within 6 hours of first passing an SBT and 15 potential reasons to delay extubation, accounting for clustering at the attending-physician level.

Results

Among 3,240 patients, 62.3% underwent timely extubation within 6 hours of first passing an SBT. Patients with delayed extubation experienced a median of 2.0 (interquartile range = 1.0–3.8) additional days on mechanical ventilation after passing an SBT, for a total of 3,930 days. This delay accounts for 32% of the total time on mechanical ventilation for the study population. Patients were less likely to have timely extubation if they underwent a procedure in the 24 hours after passing an SBT (average marginal effect [AME], −17.7%; 95% confidence interval [CI] = −29.2 to −6.3); had lower levels of consciousness (AME, −16.5%; 95% CI = −23.2 to −10.0); were on low-dose vasopressors (AME, −12.9%; 95% CI = −19.2 to −4.8) or high-dose vasopressors (AME, −12.2%; 95% CI = −17.8 to −8.1); or had copious secretions (AME, −8.7%; 95% CI = −13.3 to −4.1). However, 55% of patients with delayed extubation experienced none of these top five potential barriers. There was minimal physician variability in the decision to extubate after a successful SBT (median odds ratio: 1.10; 95% CI = 1.02 to 1.78). The most frequent reason documented by respiratory therapists for not extubating a patient after passing an SBT was attending-physician preference (43%).

Conclusions

One in 3 patients remain on mechanical ventilation after passing their first SBT, with over half lacking an identifiable barrier to extubation. Future work should be pursued to address this, including the consideration of unitwide interventions to increase timely extubation attempts among patients without contraindications.

Spontaneous breathing trials (SBTs) are a guideline-recommended bedside tool to predict respiratory readiness for extubation (1). Although there are justifiable reasons to continue mechanical ventilation, passing an SBT has good predictive value for extubation success (2–6). In recent SBT trials that protocolized extubation automatically on passing an SBT, less than 15% of patients required reintubation (7, 8). However, extubation is a complex process, and clinicians often delay extubation after patients pass an SBT (6, 9). In clinical trials where clinician discretion drives extubation decisions, the time to successful extubation is consistently delayed by 1 to 3 days (10, 11).

Data are limited regarding why extubation delays occur after patients pass an SBT. Clinicians anecdotally cite patient-related factors, such as secretions and altered mental status, and patient-independent factors, such as physician preferences (6, 12, 13). A recent analysis identified several patient-related barriers after passing an SBT, including oversedation, sepsis, and neurologic diagnoses (9). Still, our understanding of why so many patients remain on mechanical ventilation after passing an SBT is poor, and the relative importance of patient-related and patient-independent barriers to extubation is unknown. Understanding the barriers is critical, because delaying extubation prolongs mechanical ventilation and puts patients at risk for ventilator-associated complications, including pneumonia, ventilator-induced lung injury, and prolonged sedation (14).

To address this gap, we analyzed patient- and provider-level electronic health record (EHR) data collected during the SBT and periextubation period. We sought to understand the potential barriers to timely extubation by quantifying the factors associated with patients who remain on mechanical ventilation after passing an SBT.

Methods

We performed a retrospective, observational cohort study of adult patients on mechanical ventilation who were admitted to a general medical or cardiac intensive care unit (ICU) at a 1,100-bed tertiary-care academic institution between January 1, 2015, and December 31, 2023. This study was approved by the University of Michigan Medical School Institutional Review Board (Study No. HUM00144238), and informed consent was waived.

We aimed to study delays in extubation among patients with medical indications for invasive mechanical ventilation. Exclusion criteria included being surgical ICU patients; being patients who neither were extubated nor ever passed an SBT; having a code status that did not allow for reintubation; being outside hospital transfers actively receiving mechanical ventilation; being on home mechanical ventilation; having a tracheostomy at the time of first passing an SBT; and requiring extracorporeal membrane oxygenation. Patients with a code status that did not allow for reintubation were excluded on the basis of the assumption that this may impact the threshold for extubation. For patients with multiple admissions requiring mechanical ventilation during the study period, only the first admission was included.

Extubation Practices

SBTs are typically performed once a day in the study ICUs, according to a respiratory therapist–driven protocol that includes ventilator settings (positive end-expiratory pressure, ⩽8; fraction of inspired oxygen, ≤50%; no high-frequency oscillatory ventilation; and mechanical ventilation for >8 h), hemodynamics (stable vasopressor requirements and no extracorporeal membrane oxygenation), and mental status (patient has respiratory effort and no neuromuscular blockade). Patients are screened each morning and SBTs are initiated typically between 7:00 and 11:00 a.m., without physician input. The respiratory therapist documents specific contraindications for any patient who does not undergo the daily SBT. The determination of passing an SBT is based on respiratory, hemodynamic, mental, and overall clinical status during the SBT, judged by the respiratory therapist. The physician in charge makes the decision whether to extubate a patient, and a physician-written order must be placed before the respiratory therapist can extubate. SBTs that occur outside of the morning respiratory therapist protocol are more likely to be physician driven.

Outcome Measures

The primary outcome was any timely extubation attempt within 6 hours of first passing an SBT. Patients who met one of the following two criteria were considered to have first passed an SBT: 1) The word “pass” was written in nursing or respiratory therapist SBT flowsheet documentation without prior documentation of a passed SBT, and 2) a patient was extubated without any prior EHR documentation of a passed SBT during their admission. Extubations without SBT documentation were included to account for variation in institutional documentation practices. If an extubation occurs immediately after a patient passes an SBT, often only the extubation and not the passing of the SBT is documented. All patients who did not have a timely extubation attempt within 6 hours of first passing an SBT were considered to have a delayed extubation.

There is no clear precedent for the ideal timing between passing an SBT and extubation. Recent randomized clinical trials protocolized patients to a 1-hour rest back on full mechanical ventilation before extubation (7, 8), whereas prior observational studies considered timely SBTs to be those conducted in the same calendar day (9) or before the next SBT (6). We sought a feasible, real-world surrogate for the extubation protocols in the recent SBT randomized clinical trials and a priori chose a 6-hour timeframe for timely extubation.

Reintubation rates were also calculated. Reintubation was defined as a patient requiring invasive mechanical ventilation within 72 hours of extubation, which is consistent with prior definitions of extubation success (7, 15, 16). The need for noninvasive mechanical ventilation was not considered in the definition of reintubation.

Potential Barriers to Timely Extubation

We extracted demographic, medication, charge, flowsheet, and physician clinical note data to evaluate 15 potential barriers to timely extubation the first time a patient passed an SBT (see Table E1 in the data supplement). These 15 reasons were selected on the basis of prior extubation literature and expert opinion (6, 9, 12, 13, 17). Patient-related potential reasons for delaying extubation included having a major procedure in the upcoming or preceding 24 hours, hemodynamic instability requiring low- or high-dose vasopressor use (18), copious secretions, older age, higher body mass index, female sex, lower level of consciousness (Richmond Agitation Sedation Score of ⩽−3), a diagnosis of coronavirus disease (COVID-19), greater total number of days on mechanical ventilation before first passing an SBT, and general medical (vs. cardiac) ICU service. A major procedure is one that requires the patient to move to an operating room or medical procedure suite, or requires a bedside gastrointestinal procedure. It does not include bedside bronchoscopy. Potential patient-independent reasons to delay extubation included SBT timing during the afternoon and nightshift, the attending physician’s first day on clinical service, and the specific attending physician being responsible for the patient’s care. The attending physician was incorporated in the model to quantify attending physician variability in the decision to extubate. Among patients with delayed extubation despite passing an SBT, we also extracted the reason for ongoing mechanical ventilation in the subset of patients for whom this was documented by the respiratory therapist.

Provider Identification

The attending physician associated with each extubation opportunity was identified on the basis of clinical note documentation around the time of the SBT, including history and physical notes, progress notes, death notes, and discharge summary notes and is further described elsewhere (see Appendix E1 in the data supplement). In the ICUs examined, attending physicians provide care for a minimum 24-hour period, which is typically extended into 1- and 2-week blocks. One attending physician is primarily responsible for the patient’s care each day, including coverage from home overnight.

Statistical Analyses

We fit a mixed-effects, multivariate logistic regression model to evaluate the association between timely extubation within 6 hours of first passing an SBT and 15 patient-related and patient-independent factors, incorporating the specific attending physician in the model as a mixed effect (see Table E1). Because this analysis also focused on quantifying attending physician variability, we excluded patients who were cared for by attending physicians with fewer than 10 patients who passed an SBT during the study period in the regression models. Although this analysis can quantify the association between potential barriers and timely extubation, it was unable to causally link any individual barrier with the decision not to extubate a given patient.

After fitting the model, we estimated risk- and reliability-adjusted extubation rates by attending physician. Extubation rates were estimated for a 61-year-old male patient with a body mass index of 27.3 in the medical ICU (without copious secretions, recent or future procedures, vasopressor use, lower level of consciousness, or a COVID-19 diagnosis) who underwent a spontaneous breathing trial in the morning on the attending physician’s first day on service, 1.8 days after intubation. These parameters represent a typical patient, defined on the basis of median (continuous) or common (categorical) values in the dataset.

To determine the variability in extubation practices between attending physicians, we also estimated the attending median odds ratio for timely extubation. This measure represents the median increase in odds of timely extubation if the patient receives care from an attending physician with a higher, compared with a lower, timely extubation rate. The resulting median odds ratio can be utilized to directly compare the strength of the provider-level association with timely extubation and the strengths of the associations between the fixed covariates and timely extubation.

Outcomes were defined a priori. We used Pearson chi-square analysis, Fisher’s exact test, Mood’s median test, and Welch’s t test for statistical testing, with P < 0.05 as the threshold for statistical significance. All analyses were conducted in R (Version 4.2.1).

Results

Study Participants

The study included 3,240 medical and cardiac ICU patients receiving a total of 12,368 days of invasive mechanical ventilation between their first intubation and first extubation or discharge (Table 1). In this cohort, 62.3% (n = 2,017) were extubated within 6 hours after first passing an SBT (Figure 1). Among the 1,223 patients (37.7%) with delayed extubation, 24% (n = 297) were extubated within 24 hours; 52% (n = 639), within 48 hours; and 67% (n = 818), within 72 hours. The reintubation rate was the same between patients with timely extubation within 6 hours and those who were extubated 6–48 hours after passing an SBT (11.7% vs. 12.4%; P = 0.63).

Table 1.

Clinical characteristics of 3,240 patients with timely or delayed extubation after first passing a spontaneous breathing trial

Characteristic	Timely Extubation (n = 2,017)	Delayed Extubation (n = 1,223)
Patient-related factors
Age, yr, median (IQR)	61.0 (48.0–70.0)	63.0 (51.0–71.0)
Female sex, % (Reference)	46.7	44.0
Race and ethnicity: Hispanic, %	2.5	2.8
Non-Hispanic White	80.9	80.7
Non-Hispanic Black	12.6	13.1
Other	4.0	3.3
BMI, median (IQR)	27.2 (22.9–32.6)	27.3 (22.7–32.5)
COVID-19 diagnosis, %	5.1	7.6
Major procedure within 24 h before SBT, %*	14.8	6.0
Major procedure within 24 h after SBT, %*	1.9	2.7
Copious secretions before SBT, %	12.2	16.7
Lower level of consciousness (RASS ⩽−3), %	4.6	9.3
Norepinephrine equivalents at time of SBT (none), %	82.4	74.0
0 to ⩽0.1 μg/kg/min	12.2	18.2
>0.1 μg/kg/min	5.5	7.8
Duration of time from intubation to first passed SBT, d, median (IQR)	1.6 (0.9–2.9)	2.2 (1.3–3.9)
ICU service, %
General medical	82.9	86.8
Cardiac	17.1	13.2
Patient-independent factors
Attending physician day on service, %
First day	10.0	11.0
Intermediate day (not first or last)	81.1	79.7
Last day	8.9	9.2
SBT timing, in military time, %
Morning (07:00–12:59)	65.4	73.6
Afternoon (13:00–18:59)	26.5	23.5
Night shift (19:00–06:59)	8.1	2.9

Definition of abbreviations: BMI = body mass index; COVID-19 = coronavirus disease; ICU = intensive care unit; IQR = interquartile range; RASS = Richmond Agitation and Sedation Scale; Reference = reference category; SBT = spontaneous breathing trial.

^*Major procedures include procedures that require the patient to move to an operating room or medical procedure suite or that require a bedside gastrointestinal procedure. They do not include bedside bronchoscopy.

Figure 1.

Extubation and reintubation outcomes for patients in the study cohort. A total of 4,245 potentially eligible patients were intubated during the study period, of which 3,240 met all inclusion and exclusion criteria. Approximately two-thirds (62%) of patients experienced timely extubation within 6 hours of first passing a spontaneous breathing trial (SBT), and one-third (38%) experienced delayed extubation. ECMO = extracorporeal membrane oxygenation; IQR = interquartile range.

Patients with delayed extubation experienced a median of 2.0 (interquartile range [IQR] = 1.0–3.8] additional days of mechanical ventilation after first passing an SBT (Figure 2). This represented a total of 3,930 additional days on the ventilator, accounting for 32% of the total number of days on mechanical ventilation among the study population.

Figure 2.

Distribution of time between first passing a spontaneous breathing trial (SBT) and extubation, for patients experiencing delayed extubation (n = 1,223). Time to extubation for 127 patients whose extubation occurs more than 7 days after passing an SBT is not shown.

Patient-Related Factors in Extubation

Regarding patient-related factors, the top five barriers associated with timely extubation were having an upcoming procedure (average marginal effect [AME], −17.7%; 95% confidence interval [CI] = −29.2 to −6.3), a lower level of consciousness (AME, −16.5%; 95% CI = −23.2 to −10.0), requiring low-dose vasopressors (AME, −12.9%; 95% CI = −19.2 to −4.8) or high-dose vasopressors (AME, −12.2%; 95% CI = −17.8 to −8.1), and having copious secretions (AME, −8.7%; 95% CI = −13.3 to −4.1; Table 2). Older patients were also less likely to experience timely extubation (AME, −2.0% per decade; 95% CI = −3.0 to −1.0), as were those who had been on mechanical ventilation longer before passing the SBT (AME, −2.5% per day on mechanical ventilation; 95% CI = −3.2 to −1.8). Patients with a recent prior procedure were more likely to undergo timely extubation (AME, 18.1%; 95% CI = 11.9 to 24.3). A patient’s body mass index, COVID-19 diagnosis, and sex were not associated with extubation.

Table 2.

Results of mixed-effects logistic regression model to predict timely extubation attempts

Characteristic	% Change in Probability of Timely Extubation (95% CI)	OR (95% CI)	P Value
Top five patient-related barriers
Future major procedure in 24 h*	−17.7 (−29.2 to −6.3)	0.44 (0.26 to 0.75)	0.002
Lower level of consciousness (RASS ⩽−3)	−16.5 (−23.2 to −10.0)	0.47 (0.34 to 0.64)	<0.001
Norepinephrine equivalents at time of SBT (Reference: none)
0 to ⩽0.1 μg/kg/min	−12.9 (−17.8 to −8.1)	0.56 (0.45 to 0.70)	<0.001
>0.1 μg/kg/min	−12.2 (−19.2 to −4.8)	0.59 (0.43 to 0.80)	0.001
Copious secretions	−8.7 (−13.3 to −4.1)	0.67 (0.54 to 0.83)	<0.001
Other patient-related factors
Major procedure in past 24 hours*	18.1 (11.9 to 24.3)	2.30 (1.72 to 3.08)	<0.001
Cardiac ICU (Reference: medical)	6.5 (1.6 to 11.4)	1.36 (1.07–1.72)	0.01
Time from intubation to first passed SBT, per d	−2.5 (−3.2 to −1.8)	0.89 (0.86 to 0.92)	<0.001
Age, per decade	−2.0 (−3.0 to −1.0)	0.91 (0.87 to 0.96)	<0.001
Female sex	1.7 (−1.6 to 5.0)	1.08 (0.93 to 1.26)	0.30
BMI, per 5 kg/m2	−0.1 (−1.1 to 0.9)	1.00 (0.95 to 1.04)	0.91
COVID-19 diagnosis	−1.8 (−8.6 to 5.0)	0.92 (0.67 to 1.26)	0.60
Patient-independent factors
Attending physician, median odds ratio (95% CI)	—	1.10 (1.02–1.78)	—
Attending physician service day (Reference: first day)
Middle day, not first or last	2.4 (−3.0 to 7.8)	1.12 (0.88 to 1.43)	0.38
Last day	0.2 (−7.3 to 7.7)	1.01 (0.72 to 1.42)	0.96
SBT timing, in military time, median (IQR)
Morning, 07:00–12:59 (Reference)	—	—	—
Afternoon, 13:00–18:59	4.8 (1.0 to 8.6)	1.24 (1.04 to 1.48)	0.01
Night shift, 19:00–06:59	23.5 (17.7 to 29.3)	3.58 (2.37 to 5.41)	<0.001

Definition of abbreviations: BMI = body mass index; CI = confidence interval; ICU = intensive care unit; OR = odds ratio; RASS = Richmond Agitation and Sedation Scale, Reference = reference category; SBT = spontaneous breathing trial.

Change in extubation probability was determined by calculating the average marginal effect of each parameter in a multivariate logistic regression model. The model includes 3,131 patients who received care from attending physicians, with at least 10 patient extubation opportunities during the study period.

^*Major procedures include those that require the patient to move to an operating room or medical procedure suite, or that require a bedside gastrointestinal procedure. It does not include bedside bronchoscopy.

Over half of the 1,223 patients with delayed extubation had none of the top five barriers to timely extubation (55%; n = 670; Figure 3). The only strong predictor of delayed extubation for an additional 13% of these patients was using low-dose vasopressors (n = 164). Another 5% of patients with delayed extubation experienced low-dose vasopressor use and at least one additional barrier to timely extubation (n = 59).

Figure 3.

Prevalence of the top five barriers to timely extubation among patients experiencing delayed extubation. Data from 1,196 of 1,223 patients who were not extubated after first passing a spontaneous breathing trial (SBT) are included. Categories representing fewer than 10 patients (e.g., 2% of patients with delayed extubation) are excluded from the figure. $ indicates patients with copious secretions and an altered level of consciousness (n = 18; 1%); * indicates patients with an upcoming procedure in the next 24 hours (n = 16; 1%); + indicates patients on low-dose vasopressors with an altered level of consciousness (n = 14; 1%); ^ indicates patients on high-dose vasopressors with copious secretions (n = 12; 1%). Pressor = vasopressor.

Patient-Independent Factors in Extubation

Among patient-independent factors, patients who first passed an SBT in the afternoon and overnight were more likely to experience timely extubation than those who passed an SBT in the morning (AME, 4.8% and 23.5%, respectively; Table 2). The day when an attending physician was on service (first, intermediate, or last), had no association with timely extubation. There was also minimal variability in extubation practices among the 49 attending physicians who had at least 10 extubation opportunities, after adjusting for other patient-related and patient-independent factors. The median odds ratio between attending physicians for timely extubation was 1.10 (95% CI = 1.02 to 1.78), which represents the median increase in odds of timely extubation if the patient receives care from an attending physician with a higher, compared with a lower, timely extubation rate. The risk- and reliability-adjusted timely extubation rates among attending physicians for a typical patient ranged from 59% to 62%. These 49 attending physician included 71% trained in pulmonary and critical care medicine, 27% trained in cardiology, and 2% trained in emergency medicine critical care. The median number of extubation opportunities per attending physician in the primary analysis was 49 (IQR = 23–91).

Documented Reasons to Delay Extubation

Among the 349 patients with a respiratory therapist–documented reason for continuing mechanical ventilation after first passing an SBT, the reasons listed were attending physician’s preference (43.3%), airway protection (39.0%), procedure scheduled (6.0%), difficult airway (3.2%), and other (20.1%).

Discussion

More than one-third of adult medical ICU patients in this study remain mechanically ventilated 6 hours after they first passed an SBT. Patients with an upcoming procedure, lower level of consciousness, low or high vasopressor use, or copious secretions were less likely to experience timely extubation. However, each of these factors were present in a small number of patients, and the majority of patients with delayed extubation had none of these top five barriers in their clinical documentation. Despite this, there was minimal attending physician variability in the decision to extubate after patients passed an SBT, with all attending physicians consistently delaying extubation for a large proportion of patients who had no identifiable reasons for remaining on mechanical ventilation.

The high proportion of patients in our study with delayed extubation after passing an SBT (38%) is in line with prior work (6, 9). In a multicenter evaluation of SBT practices, 45% of patients across nine sites who passed an SBT were not extubated before the next SBT was performed (6). Three of the top five predictors of delaying extubation in our cohort—mental status and/or sedation concerns, secretions, and procedural needs—were also among the most common reasons to delay extubation that were documented by physicians in that study, occurring 28%, 9%, and 7% of the time, respectively. In a cohort of medical ICU patients, Patel and colleagues similarly found that 23% of patients were not extubated on the day they passed an SBT (9). That assessment of barriers and facilitators to extubation focused on patients’ baseline characteristics, such as their presenting illnesses and indications for intubation. Mental status was assessed intermittently while patients were on mechanical ventilation, and, as in our study, patients with a lower level of consciousness were less likely to undergo extubation after passing an SBT.

Many of the top five barriers to timely extubation identified in our study indicate a potentially dangerous extubation and provide reasonable justification for ongoing mechanical ventilation. For example, extubating patients on higher vasopressor doses is associated with greater rates of reintubation (17), which itself is associated with higher mortality and ventilator-associated pneumonia rates (2, 19). However, 55% of patients with delayed extubation did not have any of the top five barriers to timely extubation. An additional 13% of patients with delayed extubation were only on low-dose vasopressors, at a maximum dose of 0.10 μg/kg/min norepinephrine equivalents. However, extubating patients at low vasopressor doses is associated with lower mortality and no difference in reintubation rates, compared with patients who do not require vasopressor support (17). There is considerable variability among surveyed clinicians regarding their comfort extubating patients on low-dose vasopressors, with over one-third considering it to be a relative contraindication (20). Low-dose vasopressor use was a strong predictor of delayed extubation in our study, which is not warranted and may contribute to unnecessary increases in a patient’s duration of mechanical ventilation.

Half of the patients who were not extubated within 6 hours were extubated within 48 hours after passing an SBT. Reintubation rates were very similar among these groups (11.7% vs. 12.4%). Clinicians are notoriously inaccurate at predicting extubation success (12); therefore, it is possible that many of the patients with extubation delays of 24 to 48 hours could have been successfully extubated immediately after passing an SBT without worse outcomes. This hypothesis is supported by several prior studies (21–24). In these studies, patients who underwent early extubation to noninvasive ventilation before passing an SBT, while still requiring high levels of ventilatory support, experienced shorter durations of invasive mechanical ventilation without an increase in extubation failure. Further work is needed to determine what clinical interventions are currently performed during the period between first passing an SBT and extubation, as well as to assess how critical these are to extubation success.

Extubation rates were surprisingly similar across all attending physicians, despite diverse training backgrounds. We hypothesize that unit and institutional norms influence extubation practices and that unitwide multidisciplinary interventions may be needed to increase extubation rates after patients pass an SBT. ICUs should consider implementing routines that facilitate extubation as the default practice after patients without contraindications pass an SBT. One such intervention is to add a formal multidisciplinary discussion of any barriers to extubation on passing an SBT to a bedside daily review of the ABCDEF ICU liberation bundle (25). By empowering team members to routinely articulate extubation concerns during rounds, the severity of these concerns as true barriers to extubation could be assessed while all team members are at the bedside with the patient. Furthermore, it would require teams to identify specific reasons for ongoing mechanical ventilation and encourage timely extubation to be the anticipated action if an SBT is passed among patients in whom barriers are not identified.

Notably, afternoon and overnight SBTs were associated with timely extubation compared with morning SBTs. In the study units, morning SBTs are typically respiratory therapist-driven, adhering to a daily SBT protocol. In contrast, afternoon and overnight SBTs are more likely to be physician driven on the basis of respiratory stability. Higher extubation rates among afternoon and overnight SBTs thus likely reflects these systematic differences, including higher pre-SBT expectations for extubation success and greater physician buy-in.

Our study has several limitations. Studies that utilize EHR data are retrospective by design and limited to secondary analysis of data collected for clinical purposes. Although we identified a unique set of fifteen potential barriers to timely extubation, there still may be additional factors that remain unidentified. Future work could benefit from a mixed-methods approach to assessing these unidentified extubation barriers. For example, the results of a preextubation cuff leak test are not consistently documented at our institution. Although over 95% of trial patients have a cuff leak present when assessed (26), the lack of a cuff leak may still be a strong predictor of not attempting extubation among these rare patients. However, this detailed EHR data analysis allowed for a unique, highly granular assessment of extubation barriers after passing an SBT. This study is also limited by being conducted at a single, academic, tertiary-care center. Although we included patients from two separate ICUs cared for by attending physicians with diverse training, extubation practices are likely driven in part by institution-specific cultural norms. Thus, extending this analysis to additional institutions may demonstrate additional variability in extubation practices and strengthen the study’s generalizability. Finally, the documented reason for ongoing intubation is reported in only one-quarter of patients who remain on mechanical ventilation after first passing an SBT and should be viewed as an exploratory analysis.

Conclusion

One in 3 patients remain on mechanical ventilation on first passing an SBT. All providers delay extubation at high rates, including among patients without identifiable barriers to extubation. The decision not to extubate a patient is multifactorial, associated with both patient-related and patient-independent factors. Future work should consider multidisciplinary unitwide interventions to increase timely extubation attempts and potentially shorten the duration of mechanical ventilation for patients who pass an SBT.