Secondary Logo

Journal Logo

Anaesthesia

Prolonged time to extubation after general anaesthesia is associated with early escalation of care

A retrospective observational study

Vannucci, Andrea; Riordan, Isabella Rossi; Prifti, Kevin; Sebastiani, Anne; Helsten, Daniel L.; Lander, Daniel P.; Kallogjeri, Dorina; Cavallone, Laura

Author Information
European Journal of Anaesthesiology: May 2021 - Volume 38 - Issue 5 - p 494-504
doi: 10.1097/EJA.0000000000001316

Abstract

Introduction

Prolonged time to extubation after general anaesthesia has been defined as the time from the end of surgery (EOS) to airway extubation of at least 15 min.1 This event is significant from an operating room management standpoint because it is associated with inefficient use of personnel, increased patient waiting time, inefficient operating room workflow,2,3 decreased surgeon satisfaction1,4 and increased costs.5,6 Although prolonged time to extubation is an important metric of anaesthesia care efficiency,5 it is presently unknown whether it is also associated with unplanned escalation of care.

Therefore, we studied a large adult surgical population – excluding cardiothoracic, otolaryngology and neurosurgery cases – to assess the frequency of prolonged time to extubation and its possible association with escalation of care before discharge from the postanaesthesia care unit (PACU), such as increased use of reversal agents for opioids or benzodiazepines, immediate re-intubation or need for ventilatory support.

The secondary goal was to identify factors associated with prolonged time to extubation, as prior studies provided differing results. Bayman et al.3 showed recently that prolonged extubation is rarely dependent on the performance of individual anaesthesiologists. Other investigations have shown that time to extubation is affected by the choice of anaesthetic agent1,7,8 and by other variables such as surgical specialty, length of case, intra-operative prone position and use of a tracheal tube vs. an extraglottic device.2,7

The successful identification of factors associated with prolonged time to extubation could assist clinicians in identifying patient at risk who may benefit from individualised approaches and allow schedulers to anticipate operating room turnover times more accurately.

Methods

The study protocol was approved by the Institutional Review Board (IRB) at Washington University in St. Louis (ID 201502107), by IRB Chair Designee Jennifer Weinman, CIP, MA, on 3 March 2015. The IRB granted a waiver of written informed consent based on minimal risk to patients, as measures to prevent Health Insurance Portability and Accountability Act (a US Act to protect patients’ rights and privacy) violations were implemented in accordance with institutional practices. This article adheres to STROBE guidelines.9

The investigation was a retrospective cohort study of procedures carried out under general anaesthesia in a large US academic teaching institution. In our Anaesthesia department, one attending anaesthesiologist supervises two residents or up to four certified registered nurse anaesthetists (CRNAs); the most common attending/CRNA ratio was 1 : 3. Standard monitoring following American Society of Anesthesiologists (ASA) guidelines (https://www.asahq.org/standards-and-guidelines/standards-for-basic-anesthetic-monitoring), there is availability of qualitative train-of-four monitors in each anaesthetic location, and quantitative monitoring was progressively introduced over the course of the 5-year period. Depth of anaesthesia monitors (BIS, Medtronic, Minneapolis, Minnesota, USA) were available but their use was not mandated).

Clinical protocols are available for all common procedures and made accessible for consultation at each anaesthesia station via the Department of Anesthesiology intranet.

Cases of interest were identified by interrogating the archived anaesthesia electronic medical records (EMR) from 1 January 2010 to 31 December 2014. This period was chosen because it immediately followed the adoption of an anaesthetic EMR across all Barnes-Jewish Hospital anaesthesia locations, including the PACU. All first encounters of adult patients receiving general anaesthesia and mechanical ventilation with a newly positioned airway [either an extraglottic airway (EGA) or an endotracheal tube] were included in the study dataset. Patients undergoing neurosurgery, ear, nose, throat (ENT) or cardiothoracic procedures were excluded because of the specific evaluation criteria and extubation strategies that may affect the time to extubation independently of other patient-related or anaesthesia-related factors in these cases.

Exclusion criteria were age less than 18 years, subsequent surgical encounters of patients already included at their first encounter, patients who entered the operating room with an airway device already in place, patients discharged from the operating room without undergoing extubation or directly transferred to an ICU, and patients with no documented extubation time in the record. In addition, we removed ‘implausible or exceptional cases’ defined as those where extubation preceded the EOS by more than 10 min, cases with surgery lasting more than 1000 min, as well as cases where more than 8 mg of neostigmine were administered.

The electronic anaesthesia record used in the operating rooms and in the PACU during the study period was MetaVision (iMD-soft, Dedham, Massachusetts, USA). Operating room nurses documented relevant information in the Barnes-Jewish Hospital peri-operative record (Surgical Information Systems, Alpharetta, Georgia, USA).

Data collection included: pre-operative, intra-operative and immediate postoperative information [including medical record number (MRN) and other case identifiers, patient comorbidities, date of surgery, staff assigned to the case, anaesthetic start and end times, anaesthetic and surgical procedures, patient position, type and size of airway devices, administered drugs and ventilation modalities] documented by clinicians; and operating room ‘wheel-in’ and ‘wheel-out’ time, start of surgery and EOS documented by the operating room nurses in their EMR. Data extracted from the two databases were merged in a combined dataset by using MRN, date of surgery and case number as identifiers.

Cases were divided into three groups based on the interval between EOS and airway extubation: Group 1 – regular time to extubation (≤15 min from EOS); Group 2 – prolonged extubation, (≥16 and ≤60 min from EOS); Group 3 – very prolonged extubation (≥61 min from EOS). Administration of naloxone and flumazenil was documented in the intra-operative or PACU EMR and retrieved by querying the electronic database. Immediate re-intubation was defined as airway re-intubation following a previous extubation in the operating room and occurring before transfer of care to the PACU team.

Assisted ventilation was also reported in the EMR by PACU nurses. At the time of the study, it included invasive and noninvasive mechanical ventilation (mask continuous positive airways pressure, bilevel positive airway pressure), but not high-flow nasal cannula oxygen therapy.

In this study, ‘nonprotective ventilation’ was calculated as minutes of anaesthesia time when either tidal volume was at least 12 ml kg−1 of ideal body weight and/or peak airway pressure was at least 35 cmH2O.10–12 ‘Regular working hours’ were considered 07:00 until 17:00 h.

Neostigmine doses were stratified as per prescribing information into three categories: normal to moderate doses, up to 30 μg kg−1; normal to maximal dose, 31 to 70 μg kg−1; supramaximal at least 70 μg kg−1 (https://www.accessdata.fda.gov/drugsatfda_docs/label/2013/204078s000lbl.pdf).

Statistical analysis

Standard descriptive statistics were used to describe the study population and distribution of pre-operative and intra-operative characteristics as well as the occurrence of prolonged and very prolonged extubation. Rates were calculated by dividing the number of patients with regular, prolonged or very prolonged extubation by the total number of cases over the number of years included. Odds ratios (OR) were used to compare proportions of administration of reversal agents, immediate re-intubation and assisted ventilation in the three study groups. Logistic regression analysis was used to identify variables associated with prolonged and very prolonged extubation. We tested the independence of observations through a generalised estimating equation approach and since that assumption was valid, we elected to perform a univariable logistic regression analysis.

A backwards stepwise logistic regression with variables identified as significantly associated with each type of extubation in the univariable analysis was used for multivariable analysis. An α level of 0.01 set the criteria for variables to stay in the model. Multicollinearity between variables included in the multivariable analysis was excluded using the variance inflation factor.13

Adjusted ORs and 95% CI were used to report the results of the final models. Cases with missing values were excluded from analysis, and no data imputation was performed. To mitigate the risk of an ‘exclusion bias’, we compared the characteristics of patients who were excluded from the analysis because of missing extubation time with those who were included. To further assess the relationship between variables and outcome in the prolonged extubation model, we ran sensitivity analyses (stepwise backwards regressions) under alternative assumptions in two subpopulations, (a) and (b), of Group 2. Subgroup (a) excluded patients who had not received neuromuscular blocking agents (NMBAs) during the case; subgroup (b) excluded patients who were pronated during surgery.

All statistical tests were two sided and evaluated at the α level of 0.05. SAS 9.4 (SAS Institute Inc, Cary, North Carolina, USA) was used for statistical analysis of the data.

Results

Figure 1 shows the flow diagram of the study. The final analysis included 86 123 patients. Of those, 77 628 patients (90.1%) were extubated within 15 min from EOS (Group 1) with a median [IQR] extubation time of 5 [2 to 9] min. In Group 2, 8138 patients (9.5%) were extubated between 16 and 60 min from EOS with a median extubation time of 20 [17 to 25] min. In Group 3, a total of 357 patients (0.4%) were extubated more than 60 min from EOS with a median extubation time of 123 [123 to 183] min.

F1
Fig. 1:
Flow diagram of the study. EOS, end of surgery.

The distribution of patient characteristics in each extubation group is shown in Table 1. Baseline characteristics were similar in the three study groups.

Table 1 - Patient-related variables
Extubation <16 min from EOS Extubation 16 to 60 min from EOS Extubation ≥61 min from EOS Extubation 16 to 60 min from EOS Extubation ≥61 min from EOS
N (%) 77 628 (90.1) 8138 (9.5) 357 (0.4) OR (95% CI) OR (95% CI)
Age (years) 52.4 ± 16.1 52.2 ± 16.7 53.6 ± 16.4 0.999 0.998 1.001 1.005 0.998 1.011
BMI (kg m−2) 30.0 ± 8.0 30.4 ± 8.9 29.7 ± 7.9 1.005 1.003 1.008 0.995 0.981 1.008
Sex
 Male 33 650 (43.7) 4134 (51.3) 164 (46.3) REF REF REF REF REF REF
 Female 43 438 (56.3) 3933 (48.7) 190 (53.7) 0.737 0.704 0.772 0.897 0.728 1.107
Emergency
 No 74 124 (95.5) 7548 (92.8) 329 (92.2) REF REF REF REF REF REF
 Yes 3504 (4.5) 590 (7.2) 28 (7.8) 1.654 1.511 1.81 1.801 1.222 2.652
ASA physical status
 1 or 2 47 010 (60.6) 4302 (52.9) 170 (47.6) REF REF REF REF REF REF
 ≥3 30 616 (39.4) 3835 (47.1) 187 (52.4) 1.369 1.308 1.433 1.689 1.372 2.08
Mallampati classification
 I or II 61 952 (80.3) 6130 (76.6) 266 (76.0) REF REF REF REF REF REF
 III or IV 15 246 (19.7) 1877 (23.4) 84 (24.0) 1.244 1.178 1.314 1.283 1.004 1.641
COPD
 No 74 051 (95.4) 7673 (94.3) 334 (93.6) REF REF REF REF REF REF
 Yes 3577 (4.6) 465 (5.7) 23 (6.4) 1.255 1.136 1.386 1.426 0.933 2.178
Asthma
 No 70 194 (90.4) 7386 (90.8) 329 (92.2) REF REF REF REF REF REF
 Yes 7434 (9.6) 752 (9.2) 28 (7.8) 0.962 0.889 1.041 0.804 0.546 1.183
Alcoholism
 No 39 217 (50.5) 4430 (54.4) 201 (56.3) REF REF REF REF REF REF
 Yes 38 411 (49.5) 3708 (45.6) 156 (43.7) 0.855 0.816 0.895 0.792 0.643 0.977
Diabetes
 No 65 872 (84.9) 6833 (84.0) 311 (87.1) REF REF REF REF REF REF
 Yes 11 756 (12.1) 1305 (16.0) 46 (12.9) 1.07 1.005 1.139 0.829 0.608 1.13
Congestive heart failure
 No 61 969 (79.8) 6456 (79.3) 306 (85.7) REF REF REF REF REF REF
 Yes 15 659 (20.2) 1682 (20.7) 51 (14.3) 1.031 0.975 1.091 0.66 0.49 0.888
Hypertension
 No 59 002 (76.0) 6323 (77.7) 267 (74.8) REF REF REF REF REF REF
 Yes 18 626 (24.0) 1815 (22.3) 90 (25.2) 0.909 0.861 0.96 1.068 0.84 1.357
Coronary artery disease
 No 71 259 (91.8) 7443 (91.5) 320 (89.6) REF REF REF REF REF REF
 Yes 6369 (8.2) 695 (8.5) 37 (10.4) 1.045 0.963 1.134 1.294 0.92 1.821
Obstructive sleep apnoea
 No 60 913 (78.5) 6381 (78.4) 207 (83.2) REF REF REF REF REF REF
 Yes 16 715 (21.5) 1757 (21.6) 60 (16.8) 1.003 0.949 1.061 0.736 0.558 0.972
End-stage renal disease
 No 75 626 (97.4) 7942 (97.6) 348 (97.5) REF REF REF REF REF REF
 Yes 2002 (2.6) 196 (2.4) 9 (2.5) 0.932 0.804 1.082 0.977 0.503 1.897
Values are mean ± SD or number (%). ASA, American Society of Anesthesiologists; COPD, chronic obstructive pulmonary disease; EOS, end of surgery. Bold values indicate statistical significance.

Immediate re-intubation occurred in 91 of 86 123 cases (0.12%, 95% CI 0.10 to 0.15). The distribution of characteristics used to evaluate escalation of care was described and compared among the three extubation groups (Table 2). Prolonged and very prolonged extubation were significantly associated with immediate re-intubation, assisted ventilation in the PACU, and administration of reversal agents for opioids and benzodiazepines when compared with Group 1.

Table 2 - Escalation of care in the peri-operative period
Extubation <16 min from EOS Extubation 16 to 60 min from EOS Extubation ≥61 min from EOS Extubation 16 to 60 min from EOS Extubation ≥61 min from EOS
N % N % N % OR (95% CI) OR (95% CI)
Naloxone
 No 77 321 99.6 7806 95.9 343 96.1 REF REF REF REF REF REF
 Yes 307 0.4 332 4.1 14 3.9 10.72 9.16 12.54 10.28 5.96 17.75
Flumazenil
 No 77 607 99.97 8086 99.4 350 98 REF REF REF REF REF REF
 Yes 21 0.03 52 0.6 7 2 23.77 14.31 39.47 73.91 31.22 174.98
Immediate re-intubation
 No 77 568 99.9 8117 99.7 347 97.2 REF REF REF REF REF REF
 Yes 60 0.1 21 0.3 10 2.8 3.35 2.04 5.51 37.26 18.92 73.37
PACU respiratory support
 No 76 695 99.8 8002 99.3 343 98 REF REF REF REF REF REF
 Yes 140 0.2 53 0.7 7 2 3.64 2.65 4.99 11.23 5.23 24.14
Immediate re-intubation, re-intubation before transfer to PACU. EOS, end of surgery; PACU, postanaesthesia care unit. Bold values indicate statistical significance.

The results of the univariable analysis of individual anaesthesia-related and procedure-related factors associated with prolonged and very prolonged extubation are shown in Tables 3 and 4.

Table 3 - Anaesthesia-related variables
Extubation <16 min from EOS Extubation 16 to 60 min from EOS Extubation ≥60 min from EOS Extubation 16 to 60 min from EOS Extubation ≥61 min from EOS OR
Numb. % Numb. % Numb. %
77 628 90.1 8138 9.5 357 0.4 OR (95% CI) OR (95% CI)
Anaesthesia staff
 Solo attending 1124 1.4 102 1.3 9 2.5 REF REF REF REF REF REF
 CRNA 59 970 77.4 5439 67.6 262 73.4 1.01 0.82 1.24 0.55 0.28 1.06
 Resident 16 408 21.2 2526 31.1 86 24.1 1.70 1.38 2.09 0.66 0.33 1.30
Airway type
 EGA 16 713 21.5 354 4.4 10 2.8 REF REF REF REF REF REF
 Tube 60 902 78.5 7781 95.6 347 97.2 6.03 5.41 6.72 9.52 5.08 17.86
Intra-operative albuterol treatment
 No 75 622 97.4 7699 94.6 343 96.1 REF REF REF REF REF REF
 Yes 2006 2.6 439 5.4 14 3.9 2.15 1.93 2.39 1.54 0.90 2.63
Benzodiazepine 77 628 100 8138 100
 No 10 141 13.1 1338 16.4 49 13.7 REF REF REF REF REF REF
 Yes 67 487 86.9 6800 83.6 308 86.3 0.76 0.72 0.81 0.95 0.70 1.28
Propofol 77 628 100 8138 100
 No 3047 3.9 227 2.8 7 2 REF REF REF REF REF REF
 Yes 74 581 96.1 7911 97.2 350 98 1.42 1.24 1.63 2.04 0.97 4.32
Etomidate
 No 76 390 98.4 7993 98.2 350 98 REF REF REF REF REF REF
 Yes 1238 1.6 145 1.8 7 2 1.12 0.94 1.33 1.23 0.58 2.61
Ketamine 77 628 100 8138 100
 No 76 562 98.6 7872 96.7 341 95.5 REF REF REF REF REF REF
 Yes 1066 1.4 266 3.3 16 4.5 2.43 2.12 2.78 3.37 2.03 5.58
TIVA
 No 77 236 99.5 8033 98.7 340 95.2 REF REF REF REF REF REF
 Yes 392 0.5 105 1.3 17 4.8 2.58 2.08 3.20 9.85 5.99 16.20
Isoflurane
 No 75 635 97.4 7804 95.9 338 94.7 REF REF REF REF REF REF
 Yes 1993 2.6 224 4.1 19 5.3 1.62 1.44 1.83 2.13 1.34 3.39
Desflurane
 No 33 629 43.3 2984 36.7 117 32.8 REF REF REF REF REF REF
 Yes 43 999 56.7 5154 63.3 240 67.2 1.32 1.26 1.38 1.57 1.26 1.96
Sevoflurane
 No 25 239 32.5 3111 38.2 140 39.2 REF REF REF REF REF REF
 Yes 52 389 67.5 5027 61.8 217 60.8 0.78 0.74 0.826 0.75 0.60 0.92
NMBA
 No NMBA 17 194 22.2 479 5.9 15 4.2 REF REF REF REF REF REF
 Sux only 8103 10.4 982 12.1 22 6.2 4.35 3.89 4.87 3.11 1.61 6.00
 Sux + ND 14 250 18.4 1881 23.1 100 28 4.74 4.28 5.25 8.04 4.67 13.84
 ND only 38 081 49.1 4796 58.9 220 61.6 4.52 4.11 4.97 6.62 3.92 11.17
Benzylisoquinoline NMBA
 No 74 060 95.4 7739 95.1 343 96.1 REF REF REF REF REF REF
 Yes 3568 4.6 399 4.9 14 3.9 1.07 0.96 1.19 0.85 0.50 1.48
Steroidal NMBA
 No 28 711 37 1822 22.4 49 13.7 REF REF REF REF REF REF
 Yes 48 917 63 6316 77.6 308 86.3 2.04 1.93 2.15 3.69 2.73 4.99
TOF documented
 No 37 478 48.3 2492 30.6 84 23.5 REF REF REF REF REF REF
 Yes 40 150 51.7 5646 69.4 273 76.5 2.12 2.01 2.22 3.03 2.37 3.88
Neostigmine
 No 38 116 49.1 2594 31.9 81 22.7 REF REF REF REF REF REF
 Yes 39 512 50.9 5544 68.1 276 77.3 2.06 1.96 2.17 3.29 2.56 4.21
Neostigmine total dose (μg kg−1)
 0 38 474 49.6 2656 32.6 82 23 REF REF REF REF REF REF
 1 to 30 7635 9.8 1005 12.4 35 9.8 1.91 1.77 2.06 2.15 1.45 3.20
 31 to 70 29 199 37.6 4064 49.9 209 58.5 2.02 1.92 2.12 3.36 2.60 4.34
 ≥71 2320 3.0 413 5.1 31 8.7 2.58 2.31 2.88 6.27 4.14 9.50
Nonprotective ventilation (min)
 0 1661 2.1 133 1.6 2 0.6 REF REF REF REF REF REF
 1 to 30 823 1.1 28 0.3 1 0.3 0.43 0.28 0.64 1.01 0.09 11.15
 31 to 60 9899 12.8 344 4.2 3 0.8 0.43 0.35 0.53 0.25 0.042 1.51
 61 to 90 12 560 16.2 839 10.3 13 3.6 0.83 0.69 1.01 0.86 0.19 3.81
 91 to 120 11 567 14.9 838 10.3 23 6.4 0.905 0.75 1.09 1.65 0.39 7.01
 >120 41 118 53 5956 73.2 315 88.2 1.809 1.51 2.16 6.36 1.58 25.57
Resusc. drugs 77628 100 8138 100
 No 39203 50.5 3172 39 120 33.6 REF REF REF REF REF REF
 Yes 38425 49.5 4966 61 237 66.4 1.60 1.52 1.67 2.02 1.62 2.51
Opioids REF REF REF REF
 No 1874 2.4 147 1.8 4 1.1 REF REF REF REF REF REF
 Yes 75 754 97.6 7991 98.2 353 98.9 1.34 1.14 1.59 2.18 0.81 5.85
Morphine equiv. (μg kg−1) 4.0 ± 16.8 5.0 ± 15.8 10.7 ± 82.3 1.002 1.001 1.003 1.003 1.001 1.004
CRNA, certified registered nurse anaesthetist; EGA, extraglottic airway device; EOS, end of surgery; ND, nondepolarising; NMBA, neuromuscular blocking agents; Sux, succinylcholine; TIVA, total intravenous anaesthesia; Benzylisoquinoline NMBA, cisatracurium, atracurium; Steroid NMBA, rocuronium, vecuronium; TOF, train-of-four monitoring; Nonprotective ventilation, tidal volume more than 12 ml kg−1 and/or peak inspiratory pressure more than 35 cmH2O; Resusc. Drugs: epinephrine, norepinephrine, phenylephrine or atropine. Bold values indicate statistical significance.

Table 4 - Procedure-related variables
Extubation <16 min from EOS Extubation 16 to 60 min from EOS Extubation ≥61 min from EOS Extubation 16 to 60 min from EOS OR (95% CI) Extubation ≥61 min from EOS OR (95% CI)
N (%) 77 628 (90.1) 8138 (9.5) 357 (0.4) OR (95% CI) OR (95% CI)
Surgery time (h) 1.9 ± 1.5 2.3 ± 1.7 1.7 ± 2.0 1.147 1.133 1.162 0.905 0.834 0.981
Fluids (ml kg−1 h−1) 21.7 ± 29.9 24.6 ± 44.6 101.3 ± 195.2 1.002 1.001 1.002 1.007 1.006 1.008
Anaesthesia start
 07.00 to 15.59 72 266 93.1 7381 90.7 327 91.6 REF REF REF REF REF REF
 16.00 to 06.59 5362 6.9 757 9.3 30 8.4 0.72 0.69 0.78 0.81 0.56 1.18
Extubation phase
 Daytime 66 023 85.1 6313 77.6 249 69.8 REF REF REF REF REF REF
 After-hours 11605 14.9 1825 22.4 108 30.2 1.65 1.56 1.74 2.47 1.97 3.10
Transfusions
 No 74 397 95.8 7305 89.8 296 82.9 REF REF REF REF REF REF
 Yes 3231 4.2 833 10.2 61 17.1 2.63 2.43 2.85 4.75 3.59 6.27
Steep trendelenburg
 No 73 913 95.2 7738 95.1 330 92.4 REF REF REF REF REF REF
 Yes 3715 4.8 400 4.9 27 7.6 1.03 0.93 1.14 1.63 1.10 2.41
Reverse trendelenburg
 No 70 653 91 7287 89.5 320 89.6 REF REF REF REF REF REF
 Yes 6975 9 851 10.5 37 10.4 1.18 1.10 1.28 1.17 0.83 1.65
Prone position
 No 73 601 94.8 6982 85.8 318 89.1 REF REF REF REF REF REF
 Yes 4027 5.2 1156 14.2 39 10.9 3.03 2.82 3.25 2.24 1.61 3.13
Surgical unit
 Other 184 0.2 19 0.2 2 0.6 REF REF REF REF REF REF
 ACCS 4282 5.5 609 7.5 23 6.4 1.38 0.85 2.23 0.49 0.1162 2.11
 Colorectal 3214 4.2 434 5.4 37 10.4 1.31 0.81 2.12 1.06 0.25 4.43
 Gynaecology 10 300 13.3 774 9.5 28 7.8 0.73 0.45 1.17 0.25 0.06 1.06
 Oncology surgery 5762 7.4 208 2.6 13 3.6 0.35 0.21 0.57 0.21 0.05 0.93
 Hepatobiliary 2435 3.1 267 3.3 13 3.6 1.06 0.65 1.73 0.49 0.11 2.19
 Min. Invasive Surg. 4622 6 249 3.1 4 1.1 0.52 0.32 0.85 0.08 0.02 0.44
 Ophthalmology 3092 4 180 2.2 5 1.4 0.56 0.34 0.96 0.15 0.03 0.77
 Orthopaedic 19 356 25 3172 39.1 113 31.7 1.59 0.99 2.55 0.54 0.13 2.19
 Plastic 5689 7.3 561 6.9 8 2.2 0.96 0.59 1.54 0.13 0.03 0.61
 Transplant 2443 3.2 278 3.4 26 7.3 1.10 0.68 1.80 0.98 0.23 4.16
 Urology 11 789 15.2 942 11.6 62 17.4 0.77 0.48 1.25 0.48 0.12 1.99
 Vascular 4307 5.6 425 5.2 23 6.4 0.96 0.59 1.55 0.49 0.12 2.10
Values are mean ± SD or number, %. Fluids, crystalloid and colloid; Transfusions, packed red blood cells and/or fresh frozen plasma; ACCS, acute and critical care surgery; EOS, end of surgery; Daytime, 07:00 to 16:59 h; After-hours, 17:00 to 06:59 h. Bold values indicate statistical significance.

The multivariable models for prolonged and very prolonged extubation are reported in Table 5 and Supplementary Table 6, https://links.lww.com/EJA/A359.

Table 5 - Multivariable model for prolonged extubation
Patient-related variables OR (95% CI)
 Age 0.99 0.99 1.00
 BMI 1.008 1.005 1.011
 Sex: female vs. male 0.75 0.71 0.79
 ASA physical status ≥3 vs. 1 or 2 1.11 1.05 1.17
 Mallampati class. 3 or 4 vs. 1 or 2 1.32 1.24 1.4
 Chronic alcohol use 0.91 0.86 0.95
Procedure-related variables
 Prone position: yes vs. no 2.31 2.15 2.49
 Surgery time 0.95 0.93 0.97
 Outpatient vs. inpatient facility 0.63 0.59 0.68
Anaesthesia-related variables
 Team (CRNA) vs. solo attending 1.02 0.82 1.27
 Team (resident) vs. solo attending 1.43 1.15 1.77
 Extubation after 17:00 vs. 07:00 to 17:00 h 1.19 1.11 1.26
 Inhalational anaesthesia vs. TIVA 0.54 0.42 0.68
 Succinylcholine only vs. no NMBA 1.05 0.86 1.28
 Succ. + benzylisoquinoline vs. no NMBA 0.77 0.58 1.01
 Succ. + steroidal agent vs. no NMBA 0.84 0.69 1.03
 Benzylisoquinoline only 0.68 0.54 0.87
 Steroidal agent only 0.819 0.672 0.997
 Fluids total (ml kg−1 h−1) 1.006 1.005 1.007
 Transfusions: yes vs. no 1.71 1.57 1.88
 Airway type: tracheal tube vs. EGA 3.15 2.55 3.91
 Nonprotective ventilation <30 min 0.38 0.24 0.61
 Nonprotective ventilation ≥30 and <60 min 0.41 0.33 0.52
 Nonprotective ventilation ≥60 and <90 min 0.71 0.58 0.88
 Nonprotective ventilation ≥90 and <120 min 0.75 0.61 0.918
 Nonprotective ventilation ≥120 min 1.34 1.11 1.63
 Benzodiazepines: yes vs. no 0.89 0.83 0.95
 Intra-operative vasopressors: yes vs. no 1.10 1.04 1.16
 Neostigmine <30 μg kg−1 vs. none 1.13 1.03 1.25
 Neostigmine ≥30 μg kg−1 and <70 μg kg−1 vs. none 1.26 1.17 1.35
 Neostigmine ≥70 μg kg−1 vs. none 1.68 1.48 1.92
Multivariable regression model reporting independent factors associated with extubation occurring between 16 and 60 min after the end of surgery. ASA, American Society of Anesthesiologists; CRNA, certified registered nurse anaesthetist; EGA, extraglottic airway; NMBA, neuromuscular muscle blocking agent; TIVA, total intravenous anaesthesia. Bold values indicate statistical significance.

As reported in the Methods section, we ran sensitivity analyses (stepwise backwards regressions) under alternative assumptions in two subpopulations, (a) and (b), of Group 2. Subgroup (a) excluded patients who had not received NMBAs during the case; subgroup (b) excluded patients who were pronated during their surgery.

In subgroup (a), newly identified factors associated with lower odds of prolonged time to extubation included laparoscopic vs. open procedure (OR 0.72, 95% CI 0.68 to 0.78); start of anaesthesia during regular working hours (OR 0.84, 95% CI 0.75 to 0.94) and use of desflurane (OR 0.85, 95% CI 0.81 to 0.90). Factors with higher odds of prolonged time to extubation included intra-operative administration of albuterol (OR 1.79, 95% CI 1.61 to 2.01), ketamine (OR 1.51, 95% CI 1.29 to 1.77) and use of isoflurane (OR 1.19, 95% CI 1.04 to 1.35).

The association of neostigmine use with prolonged extubation varied with the administered dose: doses up to 30 μg kg−1 were not significantly associated with prolonged extubation (OR 1.07, 95% CI 0.98 to 1.16). The OR for doses of neostigmine of between 31 and 70 μg kg−1 was 1.23 (95% CI 1.15 to 1.31) and for doses more than 70 μg kg−1, OR was 1.68 (95% CI 1.48 to 1.91). In subgroup (a), ASA physical status, use of total intravenous anaesthesia and airway type were not independent predictors of prolonged extubation.

The sensitivity analysis in subgroup (b) confirmed the association with prolonged extubation of all the variables that were found in the general model, with the exception of one level of the ‘NMBAs variable’: isolated use of steroidal NMBAs in subgroup (b) was not significantly associated with reduced odds of prolonged extubation (OR 0.82, 95% CI 0.67 to 1.02).

Characteristics of patients who were excluded from the analysis because of missing extubation time were compared with those who were included. They appeared to be similar for demographic and anthropometric variables, but different for the time of day surgery was performed (all excluded cases started after 17:00 h), and for having a higher percentage of ASA ≥ 3 patients (50.6 vs. 40.2%), urgent procedures (9.8 vs. 4.8%) and residents undertaking the case (33.5 vs. 22.1%).

Finally, in the total population, we observed several differences in time to extubation in relation to the following variables:

  • (1) Airway type: 4 [1 to 7] min in patients who received an EGA vs. 6 [3 to 11] min in those who were intubated.
  • (2) Intra-operative patient position: 10 [5 to 15] min in those who were prone during surgery vs. 6 [2 to 10] min in those who were not.
  • (3) Avoidance or use of different NMBAs:
    • (a) 4 [1 to 7] min when no muscle relaxants were used.
    • (b) 5 [3 to 11] min when benzylisoquinolines were used.
    • (c) 6 [3 to 11] min when steroidal agents were used.
    • (d) 6 [3 to 10] min when succinylcholine and a benzylisoquinoline were used together.
    • (e) 7 [3 to 11] for combined use of succinylcholine and a steroidal agent.
    • (f) 7 [3 to 11] min when succinylcholine was the only NMBA.

Discussion

In contrast to prior investigations focusing on the managerial implications of delayed extubation,2,14 this study sought to assess the rate of prolonged and very prolonged extubation and whether unplanned delayed extubations were associated with escalation of care.

The secondary aim was to identify what patient-related, anaesthesia-related and surgery-related factors were associated with prolonged time to extubation.

In our population, prolonged and very prolonged extubation together occurred in about 10% of the cases, which is a clinically relevant percentage, although a lower figure than reported in several recent studies (ranging from 15 to 20%).1,3,14,15 We hypothesise that this difference may be due mainly to the exclusion of cardiothoracic, ENT and neurosurgery cases and, to a lesser extent, to our choice to define prolonged time to extubation as an interval at least 16 min from EOS, instead of an interval at least 15 min from EOS as other authors have done.1

Importantly, this study showed a significant association between prolonged time to extubation and instances of predefined escalation of care, such as airway re-intubation, ventilatory support and administration of reversal agents for opioids and benzodiazepine. Although very prolonged extubation is rare, prolonged extubation is a common event, with significant impact on patient care.

In this study, the rate of airway re-intubation and use of ventilatory support was progressively higher in patients with prolonged and very prolonged extubation compared with patients extubated within 15 min of EOS.

Re-intubation before PACU discharge occurred in approximately one in 1000 cases (0.1%). Other single-centre studies15,16 and a survey of a national dataset17 showed comparable rates of early re-intubation (0.19, 0.14 and 0.06%, respectively), although definitions of ‘early re-intubation’ varied. Escalation of care requiring re-intubation is uncommon in the postoperative period, but, while necessary to avoid preventable disastrous outcomes,18 these rescue interventions are still associated with increasing costs, longer patient stay and high mortality.19,20

A recent study single-centre retrospective study observed that ‘64% of anaesthesia-attributable cardiac arrests were caused by airway complications that occurred primarily with induction, emergence or in the PACU’.21

The patient safety relevance of postoperative respiratory complications has driven the development of predictive scores22–24 for these events. Although these models have not previously incorporated unanticipated delayed extubation, our results indicate that prolonged time to extubation may be a simple and useful marker of patients at risk of postoperative respiratory complications. Indeed, we observed a strong overlap of factors that were associated with prolonged time to extubation in our study and with early postoperative respiratory complications in previous reports. In the early 1990s, researchers found that up to 69% of re-intubations in the PACU were related to anaesthetic management (excessive sedation, high doses and/or residual effects of NMBAs, and inappropriate fluid management).15,25 Since then, significant changes in surgical techniques, acuity of patients undergoing elective surgery and pharmacological and clinical practices have altered the epidemiology of postoperative escalation of care. Recent studies have identified several additional factors associated with early re-intubation, including male sex, a higher ASA physical status, pre-existing respiratory or heart disease, longer case duration and the presence of a resident undertaking the case.16,17,26

Our study was not designed to identify individual factors associated with early escalation of care other than prolonged time to extubation. However, the following factors that we found to be associated with prolonged time to extubation, such as age, male sex, ASA status at least 3, presence of a trainee on the case, length of surgery, use of certain NMBAs and neostigmine, and fluid infused, had previously been associated with postoperative respiratory complications.15–17,27

Moving on to compare our results with prior studies focusing on the time interval from EOS to extubation,1,3,7 we confirmed some previously associated factors (intra-operative use of prone position, increased length of surgery and use of tracheal tubes vs. extraglottic airways) and identified additional associations (Table 5 and Supplementary Table 6, https://links.lww.com/EJA/A359). On the contrary, most of these factors are patient-related and/or procedure-related and remain outside the control of the anaesthesia team, and several of the anaesthesia-related factors cannot be anticipated reliably at the moment of case scheduling.

We recognise that our study has several limitations. First, 29 746 cases were excluded due to implausible or missing ‘EOS’ and ‘extubation time’ data. We realise that this could be a potential source of bias, but we verified that patient characteristics in the excluded group were similar to the study population and assumed that this relevant information was missing at random. At our institution, the ‘EOS’ and ‘extubation time’ time points are manually entered in the medical record by operating room nurses and anaesthesia clinicians. These events do not have direct billing implications (differently from operating room wheel-in and wheel-out time, or anaesthesia start and end time) and we have observed that clinicians often document those events later than they occur, after having attended the immediate needs of patient care. Therefore, it is possible that occasionally busy clinicians documented those times imprecisely, or even missed recording them, without an intentional selection bias.

Second, the partition of time to extubation into three periods, while based on our review of the literature, has an element of subjectivity. The definition of ‘regular extubation’ derives from the seminal work of Dexter et al.1 Similarly, the definition of ‘prolonged extubation’ (Group 2) is based on the results reported by Bayman et al.3 showing that almost all the ‘prolonged extubation’ cases they studied were extubated within 1 h from EOS. Extrapolating from that same work, we considered that extubation occurring more than 1 h after EOS (our criterion to define Group 3 or cases of ‘very prolonged extubation’) would probably identify egregious situations or clinical complications.

Third, this study was conducted before sugammadex became regularly available in the United States and at our institution (in year 2015). We acknowledge that we could have obtained different results had sugammadex been available to treat incomplete NMBA reversal and residual postoperative weakness.28 However, we believe that the information provided by this study is current and important in many contexts. Worldwide, sugammadex is used in only a fraction of patients who received muscle relaxation. In many settings, sugammadex is either not available or its use is limited by financial considerations. According to a recent study of O’Reilly-Shah et al.,29 ‘of 11 863 anaesthesia provider respondents in 183 countries, 5510 (46%) reported sugammadex was available… A majority of these providers (72%) reported selective usage of sugammadex. Most (56%) had some form of extrinsic restriction on sugammadex access, primarily due to cost…’ The same research group in another study30 demonstrated that unrestricted use of sugammadex did not reduce the overall incidence of mechanical ventilation in PACU (2.3% before vs. 1.9% after), while a significant decrease occurred only in patients in whom mechanical ventilation was due to residual neuromuscular blockade (from 0.63 to 0.20%). Based on this review of the literature, we believe that the association that we newly propose between prolonged time to extubation and early escalations of care is important and clinically relevant.

Fourth, our study setting was a large teaching institution where many trainees receive professional training. Therefore, our results may not translate to clinical settings where trainees are not involved.

Fifth, since we could not retrospectively distinguish between patients who had a planned vs. unplanned ICU admission, we excluded all patients directly admitted to ICU from operating room. Therefore, we could not identify and study cases of unplanned delayed extubation that were directly admitted to ICU, and that could otherwise have been included in Group 3.

Sixth, we are unable to explain why nonprotective ventilation seems to have a time-dependent association with both prolonged and very prolonged extubation; traumatic ventilation, as defined in the methods, is associated with decreased odds of prolonged time to extubation when administered for up to 2 h, and increased odds of prolonged time to extubation when protracted for more than 2 h. We wonder if this is because of a ‘dose-dependent effect’, or if the potential injury caused by even shorter periods of nonprotective ventilation requires at least 2 h to contribute to pathophysiological changes that may compromise readiness for extubation.10

Finally, our methods do not allow for analysis of factors affecting judgement and decision-making of individual clinicians, nor their personal availability at the end of cases. For example, we could not assess the number of concurrently supervised rooms, the level of clinical experience and autonomy of other providers on the case (besides being residents or CRNAs),31 or the simultaneous occurrence of other acute events that could have distracted the responsible physician.

Conclusion

Notwithstanding these limitations, this study confirms that prolonged time to extubation is a frequent event that is associated with unplanned escalation of care, with a potential for higher costs of care and worse patient outcomes. These results suggest that anaesthesia clinicians should be especially vigilant when extubation conditions cannot be achieved within 15 min from the EOS. Future research is needed to determine whether preventing prolonged extubation, particularly by limiting the administration of opioids and better monitoring and reversing the effects of neuromuscular blocking agents, may result in lower rates of escalation of care before PACU discharge. Similarly, more research is needed to develop predictive models of prolonged time to extubation that may be of practical utility to clinicians to identify patients at risk and to adjust their anaesthetic management accordingly, and to operating room schedulers to anticipate case turnover time more accurately.

Acknowledgements relating to this article

Assistance with the study: the authors wish to recognise the outstanding support and assistance they have received from Anne DeSchryver and all colleagues of the Division of Clinical and Translation Research of the Department of Anesthesiology at Washington University in St Louis in securing and managing the funds to conduct the study, assisting with Institutional Review Board approval, and ensuring compliance with institutional research policies.

Financial support and sponsorship: this work was supported by a research grant provided by Washington University School of Medicine Faculty Practice Plan Patient Safety and Quality Improvement (Grant ID: 8014-88).

Conflicts of interest: none.

Presentation: preliminary results of the study were presented as a poster presentation at the Joint Meeting of the Association of University Anesthesiologists and International Anesthesia Research Society, 16 to 20 May 2019, Montreal, Canada.

References

1. Dexter F, Bayman EO, Epstein RH. Statistical modeling of average and variability of time to extubation for meta-analysis comparing desflurane to sevoflurane. Anesth Analg 2010; 110:570–580.
2. Dexter F, Epstein RH. Increased mean time from end of surgery to operating room exit in a historical cohort of cases with prolonged time to extubation. Anesth Analg 2013; 117:1453–1459.
3. Bayman EO, Dexter F, Todd MM. Prolonged operative time to extubation is not a useful metric for comparing the performance of individual anesthesia providers. Anesthesiology 2016; 124:322–338.
4. Vitez TS, Macario A. Setting performance standards for an anesthesia department. J Clin Anesth 1998; 10:166–175.
5. Dexter F, Epstein RH, Bayman EO, et al. Estimating surgical case durations and making comparisons among facilities: identifying facilities with lower anesthesia professional fees. Anesth Analg 2013; 116:1103–1115.
6. Wang J, Dexter F, Yang K. A behavioral study of daily mean turnover times and first case of the day start tardiness. Anesth Analg 2013; 116:1333–1341.
7. Agoliati A, Dexter F, Lok J, et al. Meta-analysis of average and variability of time to extubation comparing isoflurane with desflurane or isoflurane with sevoflurane. Anesth Analg 2010; 110:1433–1439.
8. Maheshwari K, Ahuja S, Mascha EJ, et al. Effect of sevoflurane versus isoflurane on emergence time and postanesthesia care unit length of stay: an alternating intervention trial. Anesth Analg 2020; 130:360–366.
9. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147:573–577.
10. Ball L, Costantino F, Orefice G, et al. Intraoperative mechanical ventilation: state of the art. Minerva Anestesiol 2017; 83:1075–1088.
11. Guldner A, Kiss T, Serpa Neto A, et al. Intraoperative protective mechanical ventilation for prevention of postoperative pulmonary complications: a comprehensive review of the role of tidal volume, positive end-expiratory pressure, and lung recruitment maneuvers. Anesthesiology 2015; 123:692–713.
12. Brower RG, Matthay MA, Morris A, et al. Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000; 342:1301–1308.
13. Lin A. Integrating PROC REG and PROC LOGISTIC for collinearity examination, sample scoring and model evaluation. SESUG 2015. Savannah, GA: The Proceedings of the SouthEast SAS Users Group; 2015.
14. Masursky D, Dexter F, Kwakye MO, et al. Measure to quantify the influence of time from end of surgery to tracheal extubation on operating room workflow. Anesth Analg 2012; 115:402–406.
15. Mathew JP, Rosenbaum SH, O’Connor T, et al. Emergency tracheal intubation in the postanesthesia care unit: physician error or patient disease? Anesth Analg 1990; 71:691–697.
16. Rujirojindakul P, Geater AF, McNeil EB, et al. Risk factors for reintubation in the postanaesthetic care unit: a case–control study. Br J Anaesth 2012; 109:636–642.
17. Tillquist MN, Gabriel RA, Dutton RP, et al. Incidence and risk factors for early postoperative reintubations. J Clin Anesth 2016; 31:80–89.
18. Sessler DI, Saugel B. Beyond ‘failure to rescue’: the time has come for continuous ward monitoring. Br J Anaesth 2019; 122:304–306.
19. Ramachandran SK, Nafiu OO, Ghaferi A, et al. Independent predictors and outcomes of unanticipated early postoperative tracheal intubation after nonemergent, noncardiac surgery. Anesthesiology 2011; 115:44–53.
20. Fernandez-Bustamante A, Frendl G, Sprung J, et al. Postoperative pulmonary complications, early mortality, and hospital stay following noncardiothoracic surgery: a multicenter study by the perioperative research network investigators. JAMA Surg 2017; 152:157–166.
21. Ellis SJ, Newland MC, Simonson JA, et al. Anesthesia-related cardiac arrest. Anesthesiology 2014; 120:829–838.
22. Brueckmann B, Villa-Uribe JL, Bateman BT, et al. Development and validation of a score for prediction of postoperative respiratory complications. Anesthesiology 2013; 118:1276–1285.
23. Sabate S, Mazo V, Canet J. Predicting postoperative pulmonary complications: implications for outcomes and costs. Curr Opin Anaesthesiol 2014; 27:201–209.
24. Neto AS, da Costa LGV, Hemmes SNT, et al. The LAS VEGAS risk score for prediction of postoperative pulmonary complications: an observational study. Eur J Anaesthesiol 2018; 35:691–701.
25. Rose DK, Cohen MM, Wigglesworth DF, et al. Critical respiratory events in the postanesthesia care unit. Patient, surgical, and anesthetic factors. Anesthesiology 1994; 81:410–418.
26. Ting P-C, Chou A-H, Yang M-W, et al. Postoperative reintubation after planned extubation: a review of 137,866 general anesthetics from 2005 to 2007 in a Medical Center of Taiwan. Acta Anaesthesiol Taiwan 2010; 48:167–171.
27. McLean DJ, Diaz-Gil D, Farhan HN, et al. Dose-dependent association between intermediate-acting neuromuscular-blocking agents and postoperative respiratory complications. Anesthesiology 2015; 122:1201–1213.
28. Olesnicky BL, Traill C, Marroquin-Harris FB. The effect of routine availability of sugammadex on postoperative respiratory complications: a historical cohort study. Minerva Anestesiol 2017; 83:248–254.
29. O’Reilly-Shah VN, Wolf FA, Jabaley CS, et al. Using a worldwide in-app survey to explore sugammadex usage patterns: a prospective observational study. Br J Anaesth 2017; 119:333–335.
30. O’Reilly-Shah VN, Lynde GC, Mitchell ML, et al. Initial experience with the unrestricted introduction of sugammadex at a large academic medical center: a retrospective observational study examining postoperative mechanical ventilation and efficiency outcomes. Korean J Anesthesiol 2018; 71:374–385.
31. House LM, Calloway NH, Sandberg WS, et al. Prolonged patient emergence time among clinical anesthesia resident trainees. J Anaesthesiol Clin Pharmacol 2016; 32:446–452.

Supplemental Digital Content

Copyright © 2020 European Society of Anaesthesiology and Intensive Care. Unauthorized reproduction of this article is prohibited.