This meta-analysis has been registered with the research registry reviewregistry1726. Electronic databases PubMed, EMBASE, MEDLINE and the Cochrane Library were reviewed and carefully searched. Relevant articles were identified with the use of MeSH terms (Robotic hernia repair, Laparoscopic hernia repair, Ventral hernia repair) and Boolean operators (AND, OR) and PICO approach was used to systematically refine and narrow down the search results. The references were further searched to identify the relevant articles for a detailed analysis.

The inclusion criteria for the systematic review was the randomised control trials (RCTs) comparing RVHR against LVHR, reporting perioperative outcomes and cost analysis. All trials regardless of their language of publication and number of recruited patients were deemed suitable for inclusion.

The published data was searched and collected by authors independently on a pre-planned standard data extraction sheet. The collected data was scrutinized involving all authors to detect any discrepancy and a mutual agreement was reached about accuracy. The main variables for data collection included were the list of published authors, the country where the RCT was conducted, year of publication, demographic details of the study population, hernia recurrence numbers, duration of operation, surgical site infections, wound seroma, wound break down, delayed wound healing, cost of the laparoscopic procedure and the cost of the robotic procedure.

RevMan version 5.4 (Review Manager 5.4, The Nordic Cochrane Centre, Copenhagen, Denmark) was used for the statistical analysis [10] of the data. In order to present the summated outcome of continuous variables such as cost of the procedure and duration of operation; the standardised mean difference (SMD) was used, and the risk ratio was used to present the summated outcomes of dichotomous data (Wound complication and hernia recurrence). The SMD and RR were calculated and presented with a 95% confidence interval (CI) under the random-effects model analysis [11, 12]. A forest plot was used for the graphical presentation of the results. The statistical heterogeneity was calculated by computing the chi² test, with significance set at P < 0.05 whereas the quantification of the heterogeneity was tested using the I² test with a maximum value of 30 per cent identifying low heterogeneity [13]. For the calculation of the SMD, the inverse-variance method was used and for the calculation of the risk ratio, the Mantel- Haenszel method was used under the random effect model analysis [14, 15]. If standard deviation was not reported in the published article on RCT, it was estimated either from the range or p-value or 0.5 was added in the cell frequency assuming the same variance in both the groups which might not be true in all the cases. The estimate of the difference between both techniques was pooled, depending upon the effect weights in results determined by each trial estimate variance.

The quality of the included RCTs was assessed by using various reported tools including the tool provided by the Cochrane Collaboration [16–18]. The quality of included studies is given in Table 1 and depicted in Figures 1, 2.

The recurrence of VH at the end of the follow-up period was considered the primary endpoint for this systematic review. The criteria to diagnose the VH recurrence included the symptomatic presentation of the patient with a recurrent lump at the site of previous surgery, clinical assessment by a senior surgeon and the reporting of the radiological diagnostic investigation to confirm it. The secondary endpoints were the surgical site complications, duration of operation, length of hospital stay and the cost of LVHR as well as the cost of RVHR.

The reported quality variables in the included RCTs used to assess their strength of published evidence is summarized in Table 1. The randomization technique used in the included RCTs was computer generated in all RCTs [19–22]; the concealment was done using sealed envelopes in three included RCTs [19–21]; single blinding was reported in three RCTs [19, 20, 22] and multi-blinding [21] was reported in one included RCT. All included studies reported ethical approval and were registered before the conduction of the trial. Power calculation was done and reported in three studies [20–22].

Four RCTs [19–22] on 337 patients were included to study for perioperative outcomes and cost comparison. A one year-follow up of Petro, et, al; was used to study the recurrence rate as well [23]. One included RCT was reported from Brazil [19] and three reported RCTs were from the USA [20–22]. The characteristics of the included RCTs are presented in Table 2 and the treatment protocols used in the included studies are presented in Table 3.

In the random effects model analysis, the incidence of VH recurrence was statistically similar between groups of laparoscopic versus robotic groups [RR 1.05, 95%, CI (0.22, 4.99), Z = 0.06, P = 0.95; Figure 4]. There was moderate heterogeneity among included RCTs (Tau² = 1.26; Chi² = 5.99, df = 2; (p = 0.05; I² = 67%).

In the random effects model analysis, the risk of surgical site complications (surgical site infection, seroma formation, wound break down, slow wound healing, failed wound healing) was statistically similar between LVHR group and RVHR group and there was no heterogeneity [RR, 0.85, 95%, CI (0.48, 1.50), Z = 0.55, p = 0.58; Figure 5], (Tau² = 0.00; Chi² = 1.06, df = 3; (p = 0.79; I² = 0%). The duration of operation was shorter in patient undergoing LVHR compared to RVHR indicating superiority of laparoscopic approach over robotic approach [SMD -48.07, 95%, CI (−78.06, −18.07), z = 3.14, p = 0.002; Figure 6]. However, there was statistically significant heterogeneity among included studies on the calculation of this variable (Tau² = 680.13; Chi² = 362.84, df = 2; (p = 0.00001; I² = 99%). In addition, LVHR was associated with reduced cost compared to RIVR in the random effects model analysis [SMD 0.82, 95%, CI (−1.48, −0.16), z = 2.45, p = 0.01; Figure 7]. However, there was statistically significant heterogeneity among included RCTs (Tau² = 0.15; Chi² = 2.87, df = 1; (p = 0.09; I² = 65%).

A previously published systematic review in 2023 [24] compared LVHR against RVHR and reported patient-related outcome measures. It was concluded that the available data on laparoscopic and robotic primary ventral hernia repair was scarce, and it was highly heterogeneous, thus making it difficult to assess the superiority of either approach in the management of VH. The current study provided evidence generated from the summated outcome of four RCTs and concludes that RVHR does not seem to have any proven clinical advantage over LVHR. Another published systematic review in 2020 [25] reported perioperative outcomes in a group of patients undergoing LVHR versus RVHR. The results of this review suggested that RVHR maintained some of the advantages of laparoscopic surgery and might provide an additional advantage of reduced hernia recurrence risk. This may well be explained by the ability to perform a more complex hernia repair with robotic assistance secondary to the ease of closure of the fascial defect. Whereas the current study analysed RCTs only and failed to demonstrate the previously reported advantage of the robotic approach. Several comparative trials have been reported with diverse outcomes [26, 27] and without any conclusive recommendations.

There was significant methodological and clinical diversity among included trials indicating heterogeneity. This study is based upon the findings of four RCTs on 337 patients and due to this reason, these findings cannot be generalised. The inclusion criteria were also different and patients with incisional hernia and primary ventral were jointly recruited in both limbs of trials which can potentially contribute to the biased outcome. The size of the hernia defect was reported in all studies, but it was of different size. The duration of follow-up in included RCTs varied from 1 year to 5 years which seems to be insufficient for the accurate estimation of the recurrence rate of hernia. Power calculations and intention to treat analysis were also not reported in the two included RCTs.

A major multicentre RCT is mandatory to validate the findings of the current study before drawing a stronger conclusion about the advantages of the robotic approach in the management of VH. Trials on primacy ventral hernia and incisional hernia should be conducted separately to assess which group of hernia can benefit more from either approach. The patient recruitment criteria should be strict in terms of hernia defect size, type of mesh used, and technique of mesh fixation used to reduce methodological diversity in the RCTs. Trials should be conducted using a gold standard radiological diagnostic tool to detect clinical and subclinical recurrence for accurate measurement of primary outcome.