Search results
The Ovid MEDLINE, Ovid Embase, EBSCOhost, Cochrane Central Register of Controlled Trials, and ISIWeb of Science were systematically searched until April 25, 2018. The search resulted in 3,834 articles. After initial evaluation, 781 studies were removed for being duplicates, 2,267 for being irrelevant (as determined by reading the title and abstracts). Ultimately, 14 studies for reasons determined by reading the full text (Figure 1).
Research characteristics and assessment of quality
Summarizes patient characteristics (i,e, author, population, age, gender, type of surgery, and outcomes). This analysis included 14 published double-blind, parallel RCTs, 1,554 patients need heart surgery and underwent extubation. All patients met risk criteria for mild or moderate cardiac surgery by the EuroSCORE [21] used as a basis for diagnosing. All characteristics of included study are presented in Table 1. All patients were adults, primarily with regard to CABG or valve or aortic valve replacement (AVR) surgery. In all of the included studies, 9 studies [7, 9, 10, 17-20, 26, 27] in the experimental group were ventilated within 4 hours, and 4 studies [5, 8, 28, 29] were performed within 8 hours. The overall risks of bias in the study are shown in Table 2.
Primary outcomes
ICU length of stay
A total of 8 studies [5, 7-9, 18, 27, 28, 30] , involving 912 patients with a summary of data reported the length of ICU hospitalization of early tracheal extubation and conventional extubation. Compared with the conventional extubation group, the early tracheal extubation was significantly shortened for the ICU stay time (MD=-9.90, 95%CI [-15.69, -4.51], I2=74%) in Figure 2.
In a subgroup analysis of extubation time (Table 3), the studies of average extubation time of less than four hours (MD=-17.70, 95%CI [-35.26, -0.15], I2=86%) compared with the traditional extubation, have remarkable difference. The rest of studies’ average extubation time of more than 4 hours (MD=-6.01, 95%CI [-10.6, -1.43], I2=55%) have obvious difference. In the different surgical types, the CABG group included 4 studies (MD=-5.53, 95%CI [-9.80, -1.26], I2=60%) have significant difference; the CABG/valve group included 4 studies (MD=-19.51, 95%CI [36.57, -2.45], I2=79%) have significant difference as well (Table 4). In the subgroup according to different extubation time criteria, the remarkable difference was observed in the within 8h group that included 3 studies (MD=-7.00, 95%CI [-7.41, -6.59], I2=0%); whereas the within 6h group included 3 studies (MD=-3.74, 95%CI [-7.49, 0.01], I2=0%) have no significant difference (Table 5).
Hospital length of stay
A total of 7 studies [5, 7, 8, 17, 18, 20, 29], aggregated data from 764 patients reported results on hospital stay (MD=-0.63, 95%CI [-1.22, -0.05], I2=65%) of early tracheal extubation and conventional extubation have statistical difference, compared with the conventional extubation group in Figure 3.
In a subgroup analysis of extubation time (Table 3), the studies of the average extubation time of less than four hours, (MD=-0.15, 95%CI [-0.58, 0.27], I2=0%) have no significant difference compared with the traditional extubation. Instead, the rest of studies that the average extubation time is more than 4 hours (MD=-1.1, 95%CI [-1.38, -0.83], I2=0%) have significant difference. In the different surgical types, the CABG group showed no obvious difference among these 6 studies (MD=-0.58, 95%CI [-1.22, 0.07], I2=71%); the CABG/AVR group included 1 study (MD=-1.1, 95%CI [-2.62, 0.42], I2=NA) and no significant difference was found (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 2 studies (MD=-0.57, 95%CI [−1.65, 0.51], I2=93%) and no significant difference had shown. On the contrary, the group within 6h involved 1 study (MD=-1.20, 95%CI [-2.38, -0.02], I2=NA) have significant difference; the group within 4h (MD=-0.87, 95%CI [-2.07, 0.32], I2=0%) have no remarkable difference in both 2 studies (Table 5).
Secondary outcomes
Mortality
There is no striking difference on the risk of death (RR=0.87, 95%CI [0.51, 1.49], I2=0%) reported by a total of 1262 patients in 8 studies [5, 7, 8, 18-20, 29, 30]. The risk of death of early was 3.3% in tracheal extubation group that is the same with the conventional extubation group. The researchers of this paper divided the deaths into in-hospital death (RR=0.20,95%CI[0.02,1.65],I2=0%。95%CI[0.1,4.46],I2=50%) and deaths in 30 days (RR = 0.68, 95%CI [0.1, 4.51], I2=50%) , and that both of the two outcomes have no significant difference in Figure 4.
In a subgroup analysis of extubation time (Table 3), no significant difference was illustrated by the studies that the average extubation time is less than four hours (RR=0.69, 95%CI [0.19, 2.53], I2=0%) compared with the traditional extubation. Likewise, the rest of studies’ average extubation time of more than 4 hours (RR=0.91, 95%CI [0.51, 1.65], I2=37%) have no obvious difference. In the different surgical types, the CABG group included 5 studies (RR=0.36, 95%CI [0.11, 1.15], I2=0%) and no significant difference was demonstrated; the valve group included one study, and it reported zero mortality; the CABG/AVR group included 1 study without death; the CABG/valve group only involved 1 study (RR=1, 95%CI [0.02, 49.06], I2=NA) and no significant difference was reported (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 2 studies (RR=0.32, 95%CI [0.03, 3.05], I2=0%) have no significant difference; the group within 6h included 2 studies (RR=0.48, 95%CI [0.04, 5.94], I2=69%) and no significant difference between studies; the group within 4h included 1 study, which reported zero mortality (Table 5).
Postoperative complications
Re-intubation
A total of 1,107 patients in 10 studies[5, 8, 9, 17-20, 28-30] recorded the number of patients requiring re-intubation. Compared with the conventional extubation grou, the risk of re-intubation in the early tracheal extubation group (RR=2.14, 95%CI [1.04, 4.38], I2=0%), have striking difference.
In a subgroup analysis of extubation time (Table 3), the average extubation time of 5 studies is less than 4 hours (RR=1.38, 95%CI [0.48, 4.00], I2=0%), and no statistical difference was illustrated. The rest studies that the average extubation time is more than 4 hours (RR=3.03, 95%CI [1.11, 8.29], I2=0%) have significant difference. In the different surgical types, the CABG group included 7 studies (RR=2.12, 95%CI [0.66, 6.83], I2=0%) and the CABG/valve group included 2 studies (RR=3.63, 95%CI [1.00, 13.14], I2=0%), have striking difference; the valve group included 1 study (RR=1, 95%CI [0.22, 4.63], I2=NA) and no significant difference was shown (Table 4). In the subgroup according to different extubation time criteria, the within 8h group included 2 studies (RR=2.13, 95%CI [0.20, 22.36], I2=0%) have no significant difference; the group within 6h included 3 studies (RR=3.20, 95%CI [0.96, 10.66], I2=0%) and no significant difference was represented; the group within 4h included 1 study (RR=2.84, 95%CI [0.12, 65.12], I2=NA) and no significant difference was shown (Table 5).
Myocardial infarction
A total of 1127 patients in 9 studies[5, 7-9, 19, 20, 27, 29, 30] recorded the number of patients with postoperative myocardial infarction, and 3 of them reported no patients with myocardial infarction. The risk of myocardial infarction was 1.6% in the early tracheal extubation group and 4.3% in the conventional extubation group. The risk of early extubation was lower than that of conventional extubation (RR = 0.43, 95% CI [0.23, 0.79], I2 = 0%), and the difference was statistically significant.
In a subgroup analysis of extubation time (Table 3), the average extubation time of the 6 studies was within 4 hours (RR = 0.34, 95% CI [0.15, 0.76], I2 = 0%) and there were significant differences. The average extubation time of the rest studies is more than 4 hours (RR=0.61, 95%CI [0.22, 1.65], I2=4%), which have significant difference. In the different surgical types, the CABG group included 6 studies (RR=0.43, 95%CI [0.16, 1.18], I2=0%) and no significant difference was shown; the CABG/valve group included 2 studies (RR=0.4, 95%CI [0.06, 2.53], I2=74%) and no significant difference was shown; the valve group included 1 study, the study reported zero incidence (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 3 studies (RR=0.55, 95%CI [0.15, 1.99], I2=0%) and no significant difference was stated; the group within 6h included 2 studies (RR=0.49, 95%CI [0.06, 4.10], I2=51%) and no significant difference was shown (Table 5).
Bleeding
A total of 684 patients in 8 studies [5, 7, 9, 17, 19, 20, 28] recorded major bleeding after surgery. The risk of major bleeding in the early tracheal extubation group was 11.2%, and the conventional care group is 9.8%. The risk of major bleeding in the early group was slightly higher than that in the conventional group. The difference between the two groups (RR=0.94, 95%CI [0.67, 1.32], I2 = 0%) have no obvious difference.
In a subgroup analysis of extubation time (Table 3), the average extubation time of the 5 studies is less than 4 hours (RR=1.06, 95%CI [0.75, 1.50], I2=0%) and no significant difference was shown. The average extubation time of the rest studies is more than 4 hours (RR=0.58, 95%CI [0.21, 1.62], I2=0%) and no significant difference was shown. In the different surgical types, the CABG group included 6 studies (RR=0.83, 95%CI [0.37, 1.90], I2=0%) and no significant difference was shown; the CABG/valve group included 2 studies (RR=2.74, 95%CI [0.05, 152.22], I2=73%) and no significant difference was shown; the valve group included 1 study (RR=1.03, 95%CI [0.74, 1.43], I2=NA) and no significant difference was shown (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 3 studies (RR=0.87, 95%CI [0.31, 2.44], I2=0%) and no significant difference was shown; the group within 6h included 1 study (RR=0.50, 95%CI [0.05, 5.37], I2=NA) and no significant difference was shown; the group within 4h included 1 study (RR=0.94, 95%CI [0.06, 13.93], I2=NA) and no significant difference was represented (Table 5).
Stroke
A total of 750 patients in 7 studies[5, 8, 9, 20, 21, 28, 30] recorded postoperative stroke, and there are no stroke patients in the three studies. The risk of stroke in the early tracheal extubation group was 0.02%, and that in the routine care group was 1.5%. After comparison between the two groups (RR = 0.52, 95% CI [0.17, 1.56], I2 = 0%), there was no significant difference.
In a subgroup analysis of extubation time (Table 3), average extubation time of less than 4 hours by 3 of the studies (RR=0.67, 95%CI [0.09, 4.90], I2=0%) have no significant difference. The average extubation time of the rest studies is more than 4 hours (RR=0.47, 95%CI [0.12, 1.75], I2=0%) and no significant difference was shown. In the different surgical types, the CABG group included 5 studies (RR=0.48, 95%CI [0.11, 2.10], I2=0%) and no significant difference was shown; the CABG/valve group included 2 studies (RR=0.57, 95%CI [0.11, 2.99], I2=0%) and no significant difference was shown (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 3 studies (RR=0.65, 95%CI [0.09, 4.81], I2=0%) and no significant difference was shown; the group within 6h included 2 studies (RR=0.56, 95%CI [0.11, 2.94], I2=0%) and no significant difference was presented (Table 5).
Acute renal failure
A total of 597 patients in 5 studies [8, 20, 21, 27, 28] recorded acute renal failure after surgery, and no patients with renal failure are shown in either study. The risk of severe acute renal failure in the early group was 1.1%, which was similar to the conventional care group (0.9%). Statistical differences between the two groups (RR=1.06, 95%CI [0.33, 3.41], I2=0%) have no striking difference.
In a subgroup analysis of extubation time (Table 3), average extubation time of less than 4 hours by 2 of the studies (RR=1.53, 95%CI [0.27, 8.72], I2=0%) have no significant difference. The average extubation time of rest studies is more than 4 hours (RR=0.78, 95%CI [0.16, 3.92], I2=0%) and no significant difference was shown. In the different surgical types, the CABG group included 2 studies, which reported that the incidence of acute renal failure is zero; the CABG/valve group included 3 studies (RR=1.06, 95%CI [0.29, 3.85], I2=0%) and have no striking difference (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 1 study and reported no Acute renal failure; the group within 6h included 1 study (RR=0.42, 95%CI [0.02, 10.11], I2=0%) and no significant difference was shown (Table 5).
Arrhythmia
A total of 6 studies[5, 8, 19, 20, 29, 30] , and a total of 750 patients got postoperative arrhythmias, and no arrhythmia patients was reported in one of the studies. The risk of arrhythmia was 16.6% in the early tracheal extubation group and 18.3% in the conventional extubation group. The statistical difference between the early tracheal extubation group and the conventional extubation group (RR=1.25, 95%CI [0.93, 1.68], I2=5%) have no conspicuous difference.
In a subgroup analysis of extubation time (Table 3), no significant difference was shown in 3 hours (RR=1.02, 95%CI [0.70, 1.51], I2=0%) of the study time with average extubation time less than 4 hours. The rest of studies’ average extubation time of more than 4 hours (RR=1.61, 95%CI [1.01, 2.57], I2=35%) have shown conspicuous difference. In the different surgical types, the CABG group included 4 studies (RR=0.97, 95%CI [0.65, 1.46], I2=0%) and have no significant difference; the CABG/valve group included 1 study (RR=2.3, 95%CI [1.23, 4.33], I2=NA) and significant difference was presented; the valve group included 1 study (RR=1.17, 95%CI [0.63, 2.16], I2=NA) and have no significant difference (Table 4). In the subgroup according to different extubation time criteria, the group within 8h included 1 study (RR=0.89, 95%CI [0.36, 2.17], I2=0%) and no significant difference was illustrated; the group within 6h included 2 studies (RR=2.03, 95%CI [1.16, 3.56], I2=0%) and have no significant difference (Table 5).
Meta-regression
The meta-regression results are presented below, the incidence of bleeding (β=-0.07, P=0.42, 95%CI[-0.25, 0.11]), acute renal failure (β=-0.14, P=0.45, 95%CI[-0.53, 0.24]), hospital stay (β=-0.16, P=0.003, 95%CI [-0.26, -0.05]), death (β=0.22, P=0.10, 95%CI[-0.04, 0.49]), ICU length of stay (β=1.11, P=0.39, 95%CI [-1.41, 3.64]), stroke (β=0.18, P=0.48, 95%CI [-0.31, 0.67]), re-intubation (β=0.12, P=0.25, 95%CI [-0.08, 0.32]), myocardial infarction (β=0.15, P=0.11, 95%CI [-0.03, 0.32]), arrhythmia (β = 0.1, P=0.19, 95%CI [-0.05, 0.26]).
Publication bias
The publication bias of five outcomes, including ICU length of stay [t=-0.86, p=0.43], hospital length of stay [t=0.89, p=0.41], re-intubation [t=0.68, p=0.53], bleeding [p=0.47, t=0.83], myocardial infarction [t=-0.44, p=0.68], and arrhythmia [t=0.83, p=0.47], weren’t detected.