A total of 644 studies were identified after systematically searching the databases. After reviewing titles/abstracts and excluding duplicates, 249 studies were excluded, leaving 395 studies. After reviewing the abstracts, 283 studies were excluded. 112, which underwent full-text review and assessment for eligibility criteria, 104 studies were excluded as they did not fulfil the inclusion criteria, resulting in eight studies included in this meta-analysis. PRISMA flow chart is shown in Figure 1. These eight studies included 3,844 patients; 1,892 had had a robotic left colectomy for diverticular disease, and 1,952 had had a laparoscopic left colectomy for diverticular disease.
Original studies characteristics and quality:
A total of eight original studies [20-27] containing 1,892 patients who underwent robotic left-side colonic resections for diverticular disease and 1,952 patients who had laparoscopic left colonic resection for diverticular disease met the inclusion criteria and were included in this review. All studies were retrospective cohort studies, and no randomised clinical trials were identified. Five studies were from the USA [20, 24-27]. Three studies were from Germany (22), Turkey (23), and Italy (24).
The individual studies' characteristics and summary of results are shown in Tables 1 and 2.
Methodological appraisal of included studies:
The methodological appraisal of the included observational studies [20-27] is shown in (Table 3). The risk of bias was judged as low to moderate in two studies (19 and 26) and low in the rest of the observational studies.
Table 1. Characteristics of original studies included in the literature review.
Study Reference
|
Country
|
Publication year
|
Study design
|
Number of patients
|
|
Robotic
|
Laparoscopic
|
Total
|
Al-Temimi et al. [20]
|
USA
|
2019
|
Retrospective
|
439
|
439
|
878
|
Beltzer et al. [21]
|
Germany
|
2019
|
Retrospective
|
60
|
46
|
106
|
Bilgin et al. [22]
|
Turkey
|
2019
|
Retrospective
|
20
|
22
|
42
|
Cassini et al. [23]
|
Italy
|
2018
|
Retrospective
|
64
|
92
|
156
|
Elliot et al. [24]
|
USA
|
2015
|
Retrospective
|
11
|
20
|
31
|
Maciel et al. [25]
|
USA
|
2014
|
Retrospective
|
20
|
55
|
75
|
Ogilvie et al. [26]
|
USA
|
2013
|
Retrospective
|
69
|
69
|
138
|
Raskin et al. [27]
|
USA
|
2019
|
Retrospective
|
1209
|
1209
|
2418
|
Total
|
|
|
|
1892
|
1952
|
3844
|
Table 2. Summary of individual studies included in the analysis.
Study Reference
|
Summary of results (robotic versus laparoscopic technique)
|
Al-Temimi et al. [20]
|
- A shorter length of hospital stays in the robotic group (3.89 ± 2.18 days versus 4.75 ± 3.25 days, P < 0.001)
- Lower conversion rate in the robotic group (7.5% versus 14.3%, P = 0.001)
- Longer operative time in the robotic group (219.2 ± 95.6 min versus 188.8 ± 82.3 min, P < 0.001)
|
Beltzer et al. [21]
|
- Longer Operative time in the robotic group (130 versus 118 min; P = 0.23)
- Similar Anastomotic leakage (6.7% versus 6.5%; P = 1.0)
- More stoma rate in the robotic group (6.7% versus 4.3%; P = 1.0)
- Higher conversion to open in the robotic group (1.7% versus 0%; P = 0.36)
- Lower reoperation rate in the robotic group (8.3% versus 15.2%; P = 0.27)
- Higher mortality in the robotic group (1.7% versus 0%; P = 1.0)
- Shorter post-op ileus in the robotic group (2.2 versus 2.8 days; P = 0.01)
|
Bilgin et al. [22]
|
- The laparoscopic group had three conversions to open, none in the robotic group
- There is no difference in post-operative complications. One anastomotic leak in the laparoscopic group
|
Cassini et al. [23]
|
- There was no conversion in the robotic group versus 6.5% in the laparoscopic group
- No 30-day mortality in both groups
- Longer mean operative time in the robotic group 172.5 ± 55.64 min (110–325) versus 167.5 ± 54.4 min (80–420) (P = 0.079)
- Less mean intraoperative blood loss 138.4 ± 28.3 mL (20–185) in the robotic group versus 144.6 ± 40.6 mL (40–200) (P = 0.295)
- Similar mean hospital stay 5 ± 2.7 days (range 3–20) versus 5 ± 4.1 days (range 3–45) and (P = 0.974)
- Lower mortality rate in the robotic group 12.3% versus 21.6% (P = 0.067)
- Significant post-operative morbidity (Clavien–Dindo III and IV) 4.6% versus 13% (P = 0.091)
|
Elliott et al. [24]
|
- Similar operative times, complications, and readmission rates
- A higher conversion rate, diverting stoma, and more extended hospital stay were noticed in the robotic group, but this was attributed to higher rectal excision in the robotic group
|
Maciel et al. [25]
|
- Lower complication rate in the robotic group is 20.0% versus 29.09% (P = 0.69)
- Less estimated blood los for the robotic group was 101.25 mL (range, 40–250 mL, SD ± 55.76) vs 187.65 mL (range, 10–1000 mL, SD ± 182.26) for the laparoscopic group (P = 0.06)
- Eight patients in the laparoscopic group were converted to open (8 of 55, 14.55%), while No patients in the robotic group were converted to laparoscopic or open surgery (0 of 20, 0%) (P = 0.001)
- Length of stay was 3.5 days in the robotic group (range, 2–12 days, SD ± 2.78) versus 4.56 days (range, 1–22 days, SD ± 3.62) in the laparoscopic group (P = 0.08)
|
Ogilvie et al. [26]
|
- No difference in length of stay, pain control, and opioid use
- Time to bowel opening was shorter in the robotic group
|
Raskin et al. [27]
|
- The robotic group had a longer operative time than both open and laparoscopic groups (P = 0.0001)
- Significantly less conversion rate in the robotic group P = 0.0002
- Similar length of stay
|
Table 3. The methodological quality of the observational studies was assessed with the Newcastle-Ottawa scale.
Study
|
Is the case definition adequate?
|
Representativeness of the cases
|
Selection of Controls
|
Definition of Controls
|
Comparability of cases and controls on the basis of the design or analysis
|
Ascertainment of exposure
|
Same method of ascertainment for cases and controls
|
Non-Response rate
|
Bias
|
Al-Temimi et al. [20]
|
*
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
Low
|
Beltzer et al. [21]
|
*
|
*
|
|
*
|
*
|
*
|
*
|
*
|
Low - Moderate
|
Bilgin et al. [22]
|
*
|
*
|
|
*
|
**
|
*
|
*
|
*
|
Low
|
Cassini et al. [23]
|
*
|
*
|
|
*
|
**
|
*
|
*
|
*
|
Low
|
Elliot et al. [24]
|
*
|
*
|
|
*
|
*
|
*
|
*
|
*
|
Low - Moderate
|
Maciel et al. [25]
|
*
|
*
|
|
*
|
**
|
*
|
*
|
*
|
Low
|
Ogilvie et al. [26]
|
*
|
*
|
|
*
|
**
|
*
|
*
|
*
|
Low
|
Raskin et al. [27]
|
*
|
*
|
*
|
*
|
**
|
*
|
*
|
*
|
Low
|
‘*’ is granted if the item set and present within the study (low)
‘**’ is given if the item set and present within the study (Moderate)
Primary outcomes:
Conversion:
All eight studies have reported a conversion rate from robotic or laparoscopic to open as one of the outcome measures with a total number of 3,844 patients. 362 (9.4%) had been converted to open (Figure 2). The robotic group had a lower conversion to open rate when compared to the laparoscopic group, which was statistically significant [6.9% vs 11.9%, OR 0.56, 95% CI (0.44, 0.70), P < 0.00001]. There was no heterogeneity between the included studies (I2 = 0%, P = 0.46).
Stoma rate:
All eight studies have reported the formation of stoma rate in both groups. Of 3,844 patients, 181 (4.7%) required the formation of a stoma (Figure 3). The difference between the robotic and the laparoscopic groups was not statistically significant, with nearly similar numbers [4.7% for both groups OR 1.00, 95% CI (0.70, 1.42), P = 1.00]. There was no heterogeneity between the included studies (I2 = 5%, P = 0.39).
Secondary outcomes:
Anastomotic leakage:
Post-operative anastomotic leakage was reported in four studies [20, 21, 24, 26] with a total number of 1,153 patients, and 30 patients (2.6%) had an anastomotic leak reported (Figure 4). The difference between the robotic and the laparoscopic groups was not statistically significant, with a rate of 2.6% for both [OR 0.93, 95% CI (0.45, 1.95), P = 0.85]. There was no heterogeneity between the included studies (I2 = 0%, P= 0.78).
Operative time:
All eight studies have reported operative time as one of the outcome measures. However, seven studies were included in the meta-analysis, including 3,706 patients (Figure 5). Operative times were reported in mean/median minutes ± SD, except in the study by Ogilvie et al. (26), which reported operative time in hours and inter-quartile range. The laparoscopic approach had a longer operative time (MD 12.50, 95% CI (-21.46, 46.45)); however, this was not statistically significant (P = 0.47). A high level of heterogeneity was observed between included studies (I2 = 88%, P = 0.00001). Table 4 shows a breakdown of reported operative times.
Blood transfusion:
Five studies [20, 21, 23, 24, 27] reported results of post-operative transfusions (Table 5), with a total of 54 patients (2.3%) having had blood transfusions. The difference between the robotic and the laparoscopic groups was not statistically significant [1.5% vs 1.5%, OR 0.88, 95% CI (0.26, 3.01), P = 0.84]. There was low heterogeneity between the included studies (I 2= 35%, P = 0.20) (Figure 6).
Post-operative ileus:
Four studies reported rates of postoperative ileus [20, 22, 25, 27], including 3,413 patients (Table 6). The robotic approach had statistically significant lower rates of post-operative ileus when compared to laparoscopic [6.5% vs 9%, OR 0.70, 95% CI (0.54, 0.90), P = 0.005]. There was no heterogeneity between the two groups (I2 = 0%, P = 0.65) (Figure 7).
Hospital length of stay:
Seven studies included length of stay (LOS), including 3,738 patients (Figure 8). The robotic approach group had a trend to shorter length of stay (MD - 0.18 95% CI [-0.59, 0.23]) when compared to the laparoscopic group, and this was statistically significant P = 0.003). I2 = 76% P = 0.39. Table 7 shows a breakdown of the total length of hospital stay per included studies.
Morbidity:
534 (2.6%) patients with morbidity were reported in all studies (Figure 9). The difference between the robotic and the laparoscopic groups was statistically significant [12% versus 15.7%, OR 0.74, 95% CI (0.61, 0.89), P = 0.002]. There was no heterogeneity between the included studies (I2 = 0%, P = 0.90). As reported in five studies, sixty-five (1.7%) had Clavien-Dindo Grade III complications (Table 8). There was no significant difference [4.4% vs 6%, OR 0.70, 95% CI (0.42, 1.19), P = 0.19]. No heterogeneity was detected (I2 = 0%, P = 0.96) (Figure 10).
Mortality:
All studies except one [26] reported their 30-day mortality; out of 3,706, ten (0.3%) patients died within 30 days post-operatively (Figure 10). The difference between the robotic and the laparoscopic groups was not statistically significant [0.3% versus 0.2%, OR 1.40, 95% CI (0.41, 4.76), P = 0.59]. There was no heterogeneity between the included studies (I2 = 0%, P = 0.80). Table 9 shows post-operative morbidity and mortality in the included studies.
Sensitivity analysis:
There was no change in the direction of the pooled effect size when RD or RR was applied for dichotomous outcomes.
Table 4. Operative times (minutes) as reported in the included studies
Study Reference
|
Robotic
n = 1892
|
Operative time in minutes
mean ± SD
|
Laparoscopic
n = 1952
|
Operative time in minutes
mean ± SD
|
P-value
|
Al-Temimi et al. [20]
|
439
|
219.26 ± 95.66
|
439
|
188.87 ± 82.28
|
< 0.001
|
Beltzer et al. [21]
|
60
|
130.4 ± 42.4
|
46
|
118 ± 36.8
|
0.23
|
Bilgin et al. [22]
|
20
|
230 ± 88.3
|
22
|
243.6 ± 68.6
|
0.577
|
Cassini et al. [23]
|
64
|
172.5 ± 55.64
|
92
|
167.5 ± 54.4
|
0.797
|
Elliot et al. [24]
Median, (mean, range)
|
11
|
250 (272, 205–467)
|
20
|
247 (254, 168–360)
|
0.53
|
Maciel et al. [25]
|
20
|
207.67 ± 67.42
|
55
|
126.67 ± 49.63
|
0.0001
|
Ogilvie et al. [26]
Median hours [inter-quartile range]
|
69
|
(4.0 [1.0] )
|
69
|
2.1 [1.0]
|
0.0001
|
Raskin et al. [27]
|
1209
|
212.2 ± 75.0
|
1209
|
254.4 ± 74.8
|
0.0001
|
Table 5. Number of patients who received a post-operative blood transfusion as reported in included studies.
NR = Not Reported
Study Reference
|
Robotic
n = 1892
|
Number of patients
|
Laparoscopic
n = 1952
|
Number of patients
|
P-value
|
Al-Temimi et al. [20]
|
439
|
0
|
439
|
0
|
1
|
Beltzer et al. [21]
|
60
|
1 (1.7)
|
46
|
2 (4.3)
|
0.58
|
Bilgin et al. [22]
|
20
|
NR
|
22
|
NR
|
NR
|
Cassini et al. [23]
|
64
|
0
|
92
|
2 (2.1)
|
0.071
|
Elliot et al. [24]
|
11
|
2 (18%)
|
20
|
1 (5%)
|
0.23
|
Maciel et al. [25]
|
20
|
NR
|
55
|
NR
|
NR
|
Ogilvie et al. [26]
|
69
|
NR
|
69
|
NR
|
NR
|
Raskin et al. [27]
|
1209
|
24 (2.0)
|
1209
|
23 (1.9)
|
1.00
|
Table 6. Incidence of post-operative ileus (in days).
Study Reference
|
Robotic
n = 1892
|
Post-operative ileus in days
|
Laparoscopic
n = 1952
|
Post-operative ileus in days
|
P-value
|
Al-Temimi et al. [20]
|
439
|
26 (5.9%)
|
439
|
31 (7.1%)
|
0.488
|
Beltzer et al. [21]
|
60
|
2.2 ±6.3
|
46
|
2.8 ± 1.6
|
0.01
|
Bilgin et al. [22]
|
20
|
3(15)
|
22
|
2(9.1)
|
0.656
|
Cassini et al. [23]
|
64
|
NR
|
92
|
NR
|
NR
|
Elliot et al. [24]
|
11
|
NR
|
20
|
NR
|
NR
|
Maciel et al. [25]
|
20
|
1 (5%)
|
55
|
3 (5.45%)
|
NR
|
Ogilvie et al. [26]
|
69
|
1 [1]
|
69
|
2 (1.5)
|
0.09
|
Raskin et al. [27]
|
1209
|
80 (6.6)
|
1209
|
120 (9.9)
|
.004
|
NR = Not Reported
Table 7. Length of stay of patients in days as reported in the included studies.
Study Reference
|
Robotic
n = 1892
|
Length of stay
Mean ± SD, range
|
Laparoscopic
n = 1952
|
Length of stay
Mean ± SD, range
|
P-value
|
Al-Temimi et al. [20]
|
439
|
3.89 ± 2.18
|
439
|
4.75 ± 3.25
|
< 0.001
|
Beltzer et al. [21]
|
60
|
NR
|
46
|
NR
|
NR
|
Bilgin et al. [22]
(Median)
|
20
|
5 (4-6)
|
22
|
5 (4-6.3)
|
0.93
|
Cassini et al. [23]
|
64
|
5 ± 2.7 (3-20)
|
92
|
5 ± 4.1 (3–45)
|
0.190
|
Elliot et al. [24]
|
11
|
7 (6.5, 3–10)
|
20
|
4 (4.2, 3–10)
|
0.007
|
Maciel et al. [25]
|
20
|
3.50 ± 2.78 (2-12)
|
55
|
4.56 ± 3.62 (1-22)
|
0.08
|
Ogilvie et al. [26]
|
69
|
3.5 ± 1.6
|
69
|
3.6 [1.4]
|
0.64
|
Raskin et al. [27]
|
1209
|
5.4 ± 2.7
|
1209
|
5.7 ± 2.8
|
< 0.0001
|
NR = Not Reported
NA = Not applicable
Table 8. Grade > III Clavien-Dindo complications rate.
Study reference
|
Robotic
n = 1892
|
Number of patients with CD > III
n (%)
|
Laparoscopic
n = 1952
|
Number of patients with CD > III
n (%)
|
P-value
|
Al-Temimi et al. [20]
|
439
|
15 (3.4)
|
439
|
18 (4.1).
|
0.594
|
Beltzer et al. [21]
|
60
|
8 (13.3)
|
46
|
9 (21.7)
|
0.8
|
Bilgin et al. [22]
|
20
|
NR
|
22
|
NR
|
NR
|
Cassini et al. [23]
|
64
|
2 (3.1)
|
92
|
6 (6.5)
|
NR
|
Elliot et al. [24]
|
11
|
0
|
20
|
1 (5%)
|
0.34
|
Maciel et al. [25]
|
20
|
1 (5)
|
55
|
5 (9.09)
|
NR
|
Ogilvie et al [26]
|
69
|
NR
|
69
|
NR
|
NR
|
Raskin et al. [27]
|
1209
|
NR
|
1209
|
NR
|
NR
|
NR = Not Reported
Table 9. Post-operative morbidity and mortality rates in the reviewed articles.
Study Reference
|
Morbidity
|
Mortality
|
Robotic group
n (%)
|
Laparoscopic group
n (%)
|
Robotic group
n (%)
|
Laparoscopic group
n (%)
|
Al-Temimi et al. [20]
|
63 (14.35)
|
84 (19.13)
|
1 (0.23)
|
1 (0.23)
|
Beltzer et al. [21]
|
18 (30)
|
13 (28.26)
|
1 (1.7)
|
0
|
Bilgin et al. [22]
|
6 (30)
|
6 (27.3)
|
0
|
1 (4.5)
|
Cassini et al. [23]
|
8 (12.3)
|
20 (21.6)
|
0
|
0
|
Elliot et al. [24]
|
3 (27)
|
6 (30)
|
0
|
0
|
Maciel et al. [25]
|
4 (20)
|
16 (29.09)
|
0
|
0
|
Ogilvie et al [26]
|
10 (16.66)
|
9 (15)
|
NR
|
NR
|
Raskin et al. [27]
|
115 (9.5)
|
153 (12.7)
|
4 (0.3)
|
2 (0.2)
|
NR = Not Reported