In the present study, the results of the observational studies were consistent with that of RCTs. Forest plots showed that laparoscope prolonged the time necessary to carry out catheter insertion in PD patients. The two groups did not significantly differ in hospital stays, early and late complications, including infection, dialysate leaks, catheter migration, pericannular bleeding, blockage, and hernia. The laparoscopic group had a lower incidence of complications than that of the open group, but the differences did not reach statistical significance.
A few articles about the two techniques had already been published. Recently, John H. Crabtree made a review about the use of the laparoscope for dialysis catheter implantation, and provided us some suggestions for catheter placement [23
]. Later, Jwo SC conducted a prospective randomized study for comparison of open surgery with laparoscopic-assisted placement of Tenckhoff peritoneal dialysis catheter and written a literature review, concluding that Laparoscopic assisted percutaneous puncture exhibited no superiority to open surgery [3
]. His point of view was consistent with ours. So far, no meta-analysis had been made to compare the two methods. Therefore we made a meta-analysis, in order to make clinicians convenient to select the appropriate surgical approach. In our study, we included all the studies mentioned in two the reviews above and also searched other database, and analyze both RCTs and observational studies, and we drew a relatively clear conclusion that open surgery had the shorter operative time but similar effect to laparoscope. Laparoscopy was seemed not to reduce the incidence of catheter-related complication rates. However, it could allow for the rescue of blocked catheters [8
]. It allowed immediate start dialysis without fluid leakage and permitted simultaneous performance of other laparoscopic procedures [4
]. It provided the patient reduced perioperative discomfort and earlier return to full mobility [14
]. It also allowed the diagnosis and treatment of the accompanying surgical pathologies during the same operation, such as intra-abdominal adhesions or preformed inguinal hernias [11
]. Compared to traditional peritoneal dialysis catheter placement, laparoscopic catheter placement has smaller scar, less pain, and quicker recovery. This approach is safe, feasible, and completely visible. Dialysis tube can be fixed under laparoscope, and the catheter position is more precise. Laparoscopic catheter placement is also suitable for patients with a history of abdominal surgery or with abdominal adhesions. Omentum can be fixed and trimmed, and postoperative complications may be reduced under laparoscopy. The above advantages induced the interest of clinicians on the laparoscopic approach. On the other hand, this approach has potential problems including advanced technique, high cost, and relatively high anesthesia risk. Weighing the pros and cons, which surgical approach to choose depends on the specific conditions and clinicians.
The heterogeneity of some variables in this study is worthy of comment. Explanations included the following. The heterogeneity of the time for operation depended on the skill and experience of surgeon and different operative methods. The heterogeneity of the length of stay at hospital might caused by different standards for discharge at different hospitals. The heterogeneity of the dialysate leaks, and blockage might due to the operations by different surgeons, the different study designs, catheter types, and operation techniques. What is more, patients in one of the observational study were children [12
], and it would affect the stability of results. In order to reduce the heterogeneity, we conducted a further research, and made subgroup analysis, but got the same conclusion. Moreover, Crabtree et al. divided laparoscopic catheterization into basic laparoscopy and advanced laparoscopy. The complications in Crabtree showed that mechanical flow obstruction was 1 in 200 implantation procedures in the advanced group, which was significantly less than that in the open dissection and basic laparoscopic groups [10
]. So we classified the complications in basic laparoscopy and advanced laparoscopy and made subgroup analysis. The random-effects statistical model revealed significant heterogeneity. The results also showed that laparoscopy did not reduce complications.
Patients in the two groups were given antibiotic prophylaxsis in most of the studies [3
]. Postoperative antibiotics were not prescribed only in one study [13
]. Use of prophylactic antibiotics before catheterization was found to be effective in reducing procedure-related peritonitis [24
]. Strippoli et al. stated that the use of perioperative intravenous antibiotics, compared to no treatment, significantly reduced the risk of early peritonitis [25
]. The routine use of vancomycin for prophylaxis before catheterization is not recommended because of the emergence of vancomycin-resistant enterococci [26
]. Other antibiotics, such as a cephalosporin, should be the first choice.
This study has several limitations. First, in observational studies, the proportion of patients who had previous abdominal surgery in laparoscopic group was larger than that of open group. More patients had chosen laparoscopy in observational studies [7
], because laparoscopy offered the advantage of entrance to abdominal cavity under direct visualization, and it was superior to open surgery for patients with a history of abdominal surgery. This would have impact on the outcomes. Moreover, as mentioned in the studies, the technique and condition of performance were of wide variability. The inevitable consequence of these practice traits was that there were almost as many laparoscopic techniques for placing catheters as there were surgeons performing them [3
]. In addition, the small number of participants, as well as the low quality of the studies, might not allow a reliable conclusion. All these factors can produce high selection bias, performance bias and measuring bias. Therefore, the studies in this review are limited, further trials of large-scale, high-quality RCTs are needed to find potential advantages or disadvantages.