In all, 281 studies were identified after a search for relevant articles; five articles [12-16] with data on 274 patients were eligible for this analysis. Of the 276 excluded studies, 41 were removed for duplications, 215 were removed after screening of their titles and abstracts according to the inclusion criteria, 1 was removed because the patient did not receive general anaesthesia, 5 were removed because they were not human studies, 12 were removed because they were not postoperative studies, 1 was removed because it did not focus on GI function, and 1 was removed because it was published repeatedly. Figure 1 shows the process of study selection.
Figure 1. PRISMA flowchart of the study selection process11. CNKI, China National Knowledge Infrastructure; VIP, Chinese VIP Information; GI, gastrointestinal.
- Summary characteristics of the included studies
The five studies included in this article were all published between 2013 and 2019 and met the inclusion criteria but not the exclusion criteria. The characteristics of the included studies are shown in table 1, which lists the characteristics of these studies in detail, including type of operation, time and side of SGB, type and volume of local anaesthetic, and other parameters.
- Quality of the included studies
The bias of the included studies was assessed according to the Rob-2.Figure 2 presents detailed information about this assessment. Two studies had low overall bias [12, 16], and three had some concerns regarding bias [13-15]. These concerns lay in the deviations from the intended interventions. All three of these studies performed SGB before induction and relied on the observation of Horner’s syndrome to ensure a successful block; thus, the care givers and those delivering the interventions might have been aware of the participants’ assigned interventions. In two of the included studies, no sham procedures were performed, so the allocation concealment and blinding of participants and personnel could not be assessed [14, 15]. Figure 3 shows a funnel plot of the included studies. Because we included only 5 studies, quantitative analyses such as the Begg and Egger tests were not performed. The overall quality of our meta-analysis is shown in table 2; the results indicate that we have a moderate overall certainty of evidence.
Table 1 Characteristics of the included studies.
Study ID
|
Kind of operation
|
Age year
(sd)
|
Sex
M/F
|
Time of SGB
|
Ultrasound guided
(yes or no)
|
Treatment of SGB group
(n; side of placement; medication)
|
Treatment of control group
(n; side of placement; medication)
|
Time of resuming peristaltic sound
hours(sd)
|
Flatus time
hours(sd)
|
Time of postoperative eating
hours(sd)
|
Incidence of abdominal bloating
|
Harms related to SGB
|
Chunying Z et al13
|
Laparoscopic gynecological surgery
|
43.57
(9.06)
|
0/41
|
After induction
|
yes
|
21; right; 1% Lidocaine, 2ml
|
20; right; 0.9% NS, 2ml
|
SGB group: 22 (20.298) Control group: 30.3 (20.145)
|
SGB group: 15.4 (4.4); Control group: 20.7 (5.1)
|
-
|
SGB group: 23.8%; Control group: 60%
|
No
|
Lihua C
et al14
|
GI surgery
|
57.69
(9.47)
|
41/14
|
Before induction
|
yes
|
18; left; 0.5% Ropivacaine, 7ml
|
37; left; 0.9% NS, 7ml
|
SGB group: 46 (31) Control group: 73 (36)
|
SGB group: 66 (34); Control group: 95 (45)
|
-
|
-
|
No
|
Renbo S
et al16
|
GI surgery
|
62.07 (15.86)
|
43/35
|
Before induction
|
no
|
39; -; 1% Lidocaine, 8-10ml
|
39; no treatment to SG
|
-
|
SGB group: 27.58 (6.72); Control group: 54.64 (8.91)
|
SGB group: 32.46 (6.29); Control group: 86.37 (9.54)
|
-
|
No
|
Peng K
et al17
|
Posterior spinal surgery
|
43.60 (9.05)
|
21/19
|
After induction
|
yes
|
20; right; 1% Lidocaine, 6ml
|
20; right; 0.9% NS,6ml
|
SGB group: 16.2 (24.901); Control group: 32.1 (24.787)
|
SGB group: 12 (4.4); Control group: 14.7 (4.6)
|
-
|
SGB group: 10%; Control group: 45%
|
No
|
Yuxin S
et al15
|
GI surgery
|
57.45 (8.12)
|
48/12
|
Before induction
|
no
|
30; right; 1% Lidocain, 8-10ml
|
30; no treatment to SG
|
-
|
SGB group: 72.15 (17.05); Control group: 85.8 (20.53)
|
SGB group: 179.8 (41.74); Control group: 229.35 (83.63)
|
-
|
N0
|
Data are presented as mean ± SD. Items that could not be extracted from the original articles are described as “-”. SGB, stellate ganglion block; GI, gastrointestinal; NS, normal saline; h, hours.
Table 2 Quality assessment according to the GRADE system.
Certainty assessment
|
Summary of findings
|
Participants (studies) Follow up
|
Risk of bias
|
Inconsistency
|
Indirectness
|
Imprecision
|
Publication bias
|
Overall certainty of evidence
|
Study event rates (%)
|
Relative effect (95% CI)
|
Anticipated absolute effects
|
With sham therapy or blank control
|
With SGB
|
Risk with sham therapy or blank control
|
Risk difference with SGB
|
Postoperative flatus time (Scale from: 0 to 100)
|
274 (5 RCTs)
|
not serious a
|
serious b
|
not serious
|
not serious c
|
none
|
⨁⨁⨁◯ MODERATE
|
146
|
128
|
-
|
The mean postoperative flatus time was 146 days
|
MD 15.07 days lower (27.58 lower to 2.56 lower)
|
CI: Confidence interval; MD: Mean difference
Explanations
- Four studies reported how the random sequence was generated, while one only stated that patients were randomly allocated into two groups but did not mention how randomization. Two included studies did not perform any sham procedures, so the allocation concealment and blinding of participants and personnel could not be assessed, leading to a high risk of selection and performance bias[15, 16]. Furthermore, no studies described blinding of outcome assessments.
- The number of included studies is small, and there are some differences among them.
- We calculate the optimal infprmation size, the largest one is 102, and the total sample of each group are both more than 102
Our primary outcome is comparison of the postoperative flatus time in the SGB and control groups; this comparison is shown in Figure 4. All included studies reported this outcome, but there was great heterogeneity (P<0.00001, I2=98%). Our analysis showed an overall effect size (mean difference, MD) of -15.07 h (95% CI: -27.58, 2.56) with a Z value of 2.36 (P=0.02<0.05). However, as shown in Figure 5, when we deleted one particular study [15], the heterogeneity was greatly reduced (P=0.02 < 0.05, I2=71%), and the overall effect size (MD) became -6.77 h (95% CI: -11.67, 1.88) with a Z value of 2.71 (P=0.007<0.05). We think that this phenomenon was caused by the low quality of that study, as shown in Figures 1 and 2; that study did not describe how the random sequence was generated, the SG did not receive any treatment in the control group, and the side on which SGB was performed in the SGB group was not reported. Furthermore, the population evaluated in this study was the oldest among the five studies, and the surgery was performed to treat GI tumours. However, despite the heterogeneity, inclusion of that study did not change the effect of SGB on the postoperative flatus time.
We also performed a subgroup analysis of 3 clinical trials in patients who underwent GI surgery surgery[13-15]. In these three studies, SGB was performed before induction of general anaesthesia. As shown in Figure 6, the overall effect size (MD) was -23.92 h (95% CI: -36.49, 11.35) with a Z value of 3.73 (P=0.0002<0.05), indicating that administration of SGB before anaesthesia significantly shortened the time to flatus after GI surgery.
Overall, our analysis suggests that SGB may shorten the postoperative flatus time in patients who undergo various types of surgery under general anaesthesia, with a mean reduction of more than 6 h. This reduction was most obvious in patients who underwent GI surgery, with a mean reduction of approximately an entire day.
3.4.1 Time to resumption of peristaltic sounds
Three of the included studies reported the time to resumption of peristaltic sounds after surgery; these studies involved patients who underwent GI, laparoscopic gynaecological and posterior spinal surgery, respectively [12, 13, 16]. However, the data were presented as the number of patients whose bowel sounds recovered within a given period of time after the operation, such as before 12 h, before 24 h, before 36 h, before 48 h, and before 72 h postoperatively. For convenience, we converted these data as follows: the average of the reported times was used as the time to resumption of peristaltic sounds in the patients; thus, this outcome was also analysed as continuous. Figure 7 shows a comparison of postoperative peristaltic sound resumption time in the SGB and control groups. The overall effect size (MD) was -14.67 h (95% CI: -23.21, -6.12) with a Z value of 3.36 (P=0.0008<0.05). Furthermore, the heterogeneity was low (P=0.25>0.05, I2=28%), indicating that the combined analysis of these three studies is reasonable. This outcome, to some degree, may show that SGB can promote GI movement after different types of surgery in patients who receive surgeries under general anaesthesia. However, additional studies are needed to confirm this.
3.4.2 Time to postoperative eating
Two of the included studies reported time to postoperative eating, and the parameters in these two studies were very similar [14, 15]. First, both studies involved patients who had undergone GI surgery. Second, SGB was performed before the induction of general anaesthesia in both studies. Third, the local anaesthetic used consisted of 8-10 ml of 1% lidocaine in both studies. Last, neither study used a sham group, and no treatment was applied in the control group. Figure 8 shows the data for this outcome. There was no heterogeneity between these 2 studies (P=0.80>0.05, I2=0%). The overall effect size (MD) was -53.86 h (95% CI: -57.43, -50.29) with a Z value of 29.60 (P<0.00001). This result suggests that in GI surgery under general anaesthesia, performing SGB before the induction of anaesthesia can significantly shorten the time to postoperative eating by more than 2 days.
3.4.3 Incidence of postoperative abdominal bloating
Two studies included data on the incidence of postoperative abdominal bloating [12, 16]. One study involved laparoscopic gynaecological surgery [12], and the other involved GI surgery [16].As shown in Figure 9, no heterogeneity was found (P=0.7070>0.05, I2=0%), and odds ratio, OR) was 0.18 (95% CI: 0.06, 0.51) with a Z value of 3.22 (P =0.001 < 0.05). The result indicates that SGB may reduce the incidence of postoperative abdominal bloating..