Literature search results
A total of 1219 articles were obtained, and after eliminating duplicates by Endnote X9, and then eliminating irrelevant papers by reading the title and abstract, 15 articles were retained after reading the full text [8-22]. See Table 1 for basic information on the selected literature. A total of 45,702 patients, 11,618 patients in the RALM group and 34,084 patients in the LM group. The literature screening process and results are shown in Figure 1.
Table 1 Basic information about the included studies
First author/Year
|
Start time
|
Country
|
Group
|
Patients
|
Age(years)
|
BMI(kg/m2)
|
Bedient2009 [8]
|
2000-2008
|
United States
|
RALM
|
41
|
43.00±12.00
|
24.7±5.0
|
|
|
|
LM
|
40
|
40.90±6.60
|
25.3±5.4
|
Nezhat2009 [9]
|
2006-2007
|
United States
|
RALM
|
15
|
39.00±5.50
|
23(18-31)
|
|
|
|
LM
|
35
|
41.00±6.75
|
24(19-33)
|
Barakat2011 [10]
|
1995-2009
|
United States
|
RALM
|
89
|
37.00±1.75
|
25.15(22.14-29.44)
|
|
|
|
LM
|
93
|
38.00±2.25
|
24.10(22.00-28.01)
|
Gargiulo2012 [11]
|
2007-2009
|
United States
|
RALM
|
174
|
38.00±8.75
|
25.1(18.4-53)
|
|
|
|
LM
|
115
|
39.00±7.25
|
25.1(17.5-54.9)
|
Hsiao2013 [12]
|
2010-2011
|
Taiwan
|
RALM
|
20
|
47.00±1.25
|
24.2(22.9-25.6)
|
|
|
|
LM
|
22
|
48.00±1.00
|
23.1(21.9-25.1)
|
Ahmet2013 [13]
|
2008-2010
|
Turkey
|
RALM
|
15
|
34.20±5.65
|
25.64±3.29
|
|
|
|
LM
|
23
|
35.70±6.13
|
27.60±5.18
|
Gobern2013 [14]
|
2007-2009
|
United States
|
RALM
|
66
|
40.00±6.25
|
25(19-52)
|
|
|
|
LM
|
73
|
39.00±8.25
|
27(19-53)
|
Pluchino2014 [15]
|
1999-2001
|
Italy
|
RALM
|
70
|
34.72±5.95
|
22.86(17-35)
|
|
|
|
LM
|
69
|
36.40±7.14
|
23.84(20.5-28.5)
|
Ngan2017 [16]
|
2008-2012
|
United States
|
RALM
|
10677
|
NA
|
NA
|
|
|
|
LM
|
33088
|
NA
|
NA
|
MacKoul2018 [17]
|
2011-2013
|
United States
|
RALM
|
156
|
36.5±5.7
|
28.6±.7
|
|
|
|
LM
|
163
|
37.1±7.3
|
27.7±6.5
|
Takmaz2018 [18]
|
2016-2017
|
Turkey
|
RALM
|
31
|
38±5
|
23±4
|
|
|
|
LM
|
33
|
35±5
|
24±4
|
Chen2018 [19]
|
2012-2016
|
Taiwan
|
RALM
|
26
|
41(39-46)
|
23.7(20.7-26.5)
|
|
|
|
LM
|
52
|
47(44-49)
|
25.2(22.1-28.6)
|
Sheu2019 [20]
|
2014-2017
|
Taiwan
|
RALM
|
93
|
39±6.7
|
21.9±2.9
|
|
|
|
LM
|
110
|
39±6.1
|
22.4±3.5
|
Won2020 [21]
|
2017-2019
|
Korea
|
RALM
|
121
|
39.1±5.8
|
22.7±3.0
|
|
|
|
LM
|
144
|
39.3±5.6
|
22.9±4.1
|
Morales2022 [22]
|
2010-2018
|
Mexico
|
RALM
|
24
|
35.23±4.19
|
23.38±1.77
|
|
|
|
LM
|
24
|
37.24±5.65
|
24.62±3.28
|
Abbreviations: BMI body mass index, NA No relevant data available.
Quality evaluation of the included literature
The NOS scores for the included cohort studies are shown in Table 2 and all studies were of high quality with NOS scores between 7 and 9.
Table 2 NOS scores of included cohort studies (points)
Study
|
Study population
|
Comparability between groups
|
Outcome measurement
|
Scores
|
Exposure group representation
|
Non-exposure group selection method
|
Exposure factor determination methods
|
Outcomes
|
Evaluation of results
|
Is the follow-up long enough
|
Adequacy of follow-up visits
|
Bedient2009[8]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Nezhat2009[9]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Barakat2011[10]
|
1
|
1
|
1
|
0
|
1
|
1
|
1
|
1
|
7
|
Gargiulo2012[11]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Hsiao2013[12]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Ahmet2013[13]
|
1
|
1
|
1
|
1
|
1
|
1
|
0
|
1
|
7
|
Gobern2013[14]
|
1
|
1
|
1
|
0
|
1
|
1
|
1
|
1
|
7
|
Pluchino2014[15]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Ngan2017[16]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
MacKoul2018[17]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Takmaz2018[18]
|
1
|
1
|
1
|
0
|
1
|
1
|
1
|
1
|
7
|
Chen2018[19]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Sheu2019[20]
|
1
|
1
|
1
|
0
|
1
|
1
|
1
|
1
|
7
|
Won2020[21]
|
1
|
1
|
1
|
0
|
2
|
1
|
1
|
1
|
8
|
Morales2022[22]
|
1
|
1
|
1
|
0
|
1
|
1
|
1
|
1
|
7
|
Meta-analysis results
Comparison of operation times
A total of 14 studies were included [8-15,17-22] with a total of 1937 patients, 941 patients in the RALM group and 996 patients in the LM group. There was statistical heterogeneity between studies (P<0.00001, I2=94%), so a random effects model was used for data analysis. The results [MD=38.61, 95%CI (19.36, 57.86), P<0.0001] indicate a statistically significant difference between the two surgical approaches in terms of operative time, suggesting that the LM group had a shorter operative time than the RALM group, the results of the Meta-analysis are shown in Figure 2.
Comparison of intraoperative bleeding
A total of 11 studies [8-10,12-15,17-18,21-22] were included with a total of 1367 patients, 648 patients in the RALM group and 719 patients in the LM group. There was statistical heterogeneity between studies (P<0.1, I2=78%), so a random effects model was used for data analysis. The results [MD=-24.67, 95%CI (-41.91, -7.43), P=0.005] indicate a statistically significant difference between the two surgical approaches in terms of intraoperative bleeding, suggesting that the RALM group had less intraoperative bleeding than the LM group, the results of the Meta-analysis are shown in Figure 3.
Comparison of the incidence of blood transfusions
A total of 11 studies were included [8,10-12,14-17,19,21-22] with a total of 45,347 patients, 11,464 patients in the RALM group and 33,883 patients in the LM group. There was no statistical heterogeneity between studies (P=0.46>0.1, I2=0%), so a fixed effects model was used for data analysis. The results [OR=0.86, 95%CI (0.77, 0.97), P=0.01] indicate a statistically significant difference in the incidence of blood transfusions between the two procedures, suggesting that fewer people in the RALM group required blood transfusions than in the LM group, the results of the Meta-analysis are shown in Figure 4.
Comparison of length of stay in hospital
A total of 10 studies were included [9-10,12-15,17,20-22] with a total of 1425 patients, 669 patients in the RALM group and 756 patients in the LM group. There was statistical heterogeneity between studies (P=0.54, I2=0%), so a fixed effects model was used for data analysis. The results [OR=-0.11, 95%CI (-0.21, -0.01), P=0.03] illustrate a statistically significant difference in length of stay between the two surgical procedures, indicating that the RALM group had a lesser length of stay than the LM group, the results of the Meta-analysis are shown in Figure 5.
Comparison of transit open belly rates
A total of 11 studies were included [8-17,22] with a total of 45,092 patients, 11,347 patients in the RALM group and 33,745 patients in the LM group. Statistical heterogeneity between studies was not significant (P=0.32>0.1, I2=13%), so a fixed effects model was used for data analysis. The results [OR=0.82, 95%CI (0.73, 0.92), P=0.0006] indicate a statistically significant difference between the two surgical approaches in terms of intermediate openings, suggesting a lower rate of intermediate openings in the RALM group than in the LM group, the results of the Meta-analysis are shown in Figure 6.
Comparison of the incidence of post-operative complications
A total of 11 studies were included [8-15,17,19,21] with a total of 1622 patients, 793 patients in the RALM group and 829 patients in the LM group. Statistical heterogeneity between studies was not significant (P=0.20>0.1, I2=27%), so a fixed effects model was used for data analysis. The results [OR=0.58, 95%CI (0.40, 0.86), P=0.006] illustrate a statistically significant difference between the two surgical approaches in terms of postoperative complications, indicating that the RALM group had fewer postoperative complications than the LM group, the results of the Meta-analysis are shown in Figure 7.
Comparison of the largest myxoma diameters
A total of 12 studies were included [8-14,18-22] with a total of 1479 patients, 715 patients in the RALM group and 764 patients in the LM group. There was statistical heterogeneity between studies (P<0.00001, I2=87%), so a random effects model was used for data analysis. The results [MD=0.26, 95%CI (-0.17, 0.70), P=0.24] indicate that there was no significant difference in maximum myoma diameter between the RALM and LM groups, and the results of the Meta-analysis are shown in Figure 8.
Sensitivity analysis
We sequentially excluded individual studies before combining the analyses for each of the indicators measured, and the results did not change significantly, indicating that the findings of this study are reliable.
Publication bias
A funnel plot of each outcome indicator revealed that the studies were evenly distributed on both sides of the funnel plot, indicating a low publication bias for this study. The funnel plot plotted for the incidence of postoperative complications is shown in Figure 9.