Study quality assessment
The quality scores of the included observational studies were ranged from 5 to 8 with a mean score of 6.5, which suggested the relatively high quality in the meta analysis (Table 2, in online supplement 1). All studies did not mention the allocation concealment and most of studies did not Provide the information about the blinding method (in clinical surgery, family members are required to obtain informed consent and inform their plans). The lowest agreement was achieved in the incomplete outcome data, while the perfect agreement was achieved in the selective reporting. The risk of bias of each study was showed in Figures 2 A,B.
Primary outcomes
In terms of complications, the P values of Node-splitting analysis are all> 0.05, which means that the difference is not statistically significant. Therefore, the consistency test is used, and the probabilistic ranking results (Figure 4 A) are in order: 3D-MVS>M -MVS>T-MVS>R-MVS>C-MVS. Specific values: 3D-MVS(OR:0.65, 95%CI:0.13 to 3.00), M-MVS(OR=0.56, 95%CI:0.35 to 0.90) )], T-MVS (OR: 0.60, 95% CI:0.32 to 1.15), R-MVS (OR: 0.76, 95% CI:0.48 to 1.14)) are less than C-MVS.
In terms of blood transfusion rate, the ranking result is: T-MVS>3D-MVS>M-MVS>R-MVS>C-MVS. Specific value: T-MVS(OR:0.31, 95%CI:0.07 to 1.40), 3D -MVS(OR:0.35, 95%CI:0.05 to 2.62), M-MVS(OR:0.61, 95%CI:0.27 to 1.34), R-MVS(OR:0.77, 95%CI:0.38 to 1.55) Lower than C-MVS.
In terms of MR and paravalvular leakage, the ranking result is: 3D-MVS>M-MVS>T-MVS>R-MVS>C-MVS. The specific value is: 3D-MVS(OR:0.03, 95%CI:0.00 to 8315 ), M-MVS(OR: 0.46, 95%CI: 0.25 to 0.88), T-MVS(OR: 0.64, 95%CI: 0.25 to 1.47), R-MVS(OR: 0.69, 95%CI: 0.35 to 1.24) are lower than C-MVS.
Regarding the 30-day mortality rate, we use the node-split model (NM) for non-uniformity test. The last line indicates P<0.05, that is, the difference is statistically significant. According to the principles of statistical methods, we use non-uniformity test. Consistency test (Figure 5) and analysis of the source of inconsistency. The inconsistency suggests that there is a significant statistical difference between direct comparison and indirect comparison, so we should refer to the traditional direct comparison result to be more reliable.
Secondary outcomes
In terms of mechanical ventilation time, the robot takes the shortest time, and the ranking result is: R-MVS>3D-MVS>T-MVS>M-MVS>C-MVS.
In terms of recovery time, the robot group recovers the fastest, and the ranking results are: R-MVS>M-MVS>T-MVS>C-MVS.
In terms of postoperative thoracic drainage, the small incision drainage value is the smallest, and the ranking result (from less to more drainage) is: M-MVS>T-MVS>3D-MVS>R-MVS>C-MVS.
In terms of scoring, only 4 intervention measures were compared. In terms of physiological score, the ranking results of superiority and inferiority were as follows: R-MVS>M-MVS>T-MVS>C-MVS. In terms of psychological score, the order is: R-MVS>T-MVS>C-MVS>M-MVS.
Publication bias
A funnel plot of all included studies in this network meta-analysis was made for visual screening of any publication bias (figure 6). It revealed that all included studies were distributed around the vertical and oblique line within the 95% CIs, suggesting no obvious publication bias.
DISCUSSION
The above-mentioned surgical procedures have their own advantages and disadvantages, and even C-MVS is still used as a routine operation method in some institutions restricted by endoscopic conditions due to its remarkable curative effect.[27] The effect of different surgical methods in the treatment of mitral regurgitation is described, and the probabilistic results of their pros and cons are given.
It should be emphasized that in terms of total hospitalization costs, three of the studies [7,23,27] involved and performed statistical analysis, among which Zhu Yilin[27] (P=0.391, T-MVS vs M-MVS vs C- MVS), Coyan[7] (P=0.273, R-MVS vs C-MVS) directly compared the results with no statistical difference; while Liu’s study[23] suggested that the difference was statistically significant (P=0.002, T -MVS vs C-MVS). Perin[34] and others abroad reported that the total hospitalization cost of minimally invasive and C-MVS mitral valve surgery is not different: the high surgical cost of minimally invasive surgery is lowered by the lower postoperative cost. The cost is offset, and the length of hospital stay is shortened by about 2 days, which is consistent with the author’s research results to a certain extent. In addition, we can conclude that the minimally invasive surgery options can reduce trauma to a certain extent by comparing them with C-MVS. Saving blood products, reducing complications, and speeding up body recovery are beneficial to improve the quality of life of patients after surgery.
Although most of articles record the left ventricular ejection fraction (LVEF) values before, due to the large clinical heterogeneity between the studies, the postoperative EF value was not included in the outcome indicators for comparison. A Meta study in China[35] once reported that the Percutaneous Mitral Valve Clipping (Mitra-Clip treatment) can improve the cardiac function of patients with mitral regurgitation and heart failure to a certain extent, but further studies are still needed to prove its conclusions; recent studies by Khader [36] and others in the UK have also repeatedly emphasized the mid-term durability of MR valve surgical repair is better than Mitra-Clip.
Combined with this Network Meta, we believes that MR minimally invasive surgery will still be the standard treatment plan for mitral valve disease for a long period of time in the future. Which minimally invasive program needs to be determined by the institution based on the individual and the environment.
CONCLUSION
Based on the above network meta-analysis and ranking results, 3D-MVS operation method can be used as the best choice in the surgical scheme of mitral valve disease, followed by T-MVS and M-MVS. Of course, if there are resource conditions, the combination of two minimally invasive operations (such as small incision combined with thoracoscopy) can ensure safer operation and faster recovery of patients.
Figure 1 Study flow diagram
Figure 2 A-B. (A) risk of bias summary; (B) risk of bias graph.
Figure 3 a-d. Network diagram of the treatments. (a ): Complications; (b ): Blood transfusion rate; (c ) : Mitral regurgitation and perivalvular leakage during follow-up; (d ): Mortality within 30 days.
Figure 4 A-C. Rank probability of the best treatment. Rank 1 is worst, rank N is best. (A): Complications; (B): Blood transfusion rate; (C): Mitral regurgitation and perivalvular leakage during follow-up.
Figure 5 Node-splitting analysis of the 30-day mortality index. Data are presented as relative effect (95% credible interval).
Figure 6 A funnel plot of all included studies in this network meta-analysis was made for visual screening of any publication bias.