Methods
Study selection
The Pubmed, EMBASE, Cochrane Library, Ovid, and Web of Science databases were searched systematically for all articles published in English before July 2012 that compared RT and ET. The terms used for the search were: ‘robotic’ and ‘thyroidectomy’.
Reference lists of all retrieved articles were also manually searched for additional studies. Two reviewers independently extracted the data from each study. All relevant text, tables, and figures were reviewed for data extraction. Discrepancies between the two reviewers were resolved by discussion and consensus.
Inclusion and exclusion criteria
Only studies in the English language were considered for inclusion. In addition, each study had to fulfill the following criteria: 1) it compared the outcomes of RT and ET, and 2) it reported on at least one of the outcome measures mentioned below. In cases where dual (or multiple) studies were reported by the same institution and/or authors, either the higher-quality or the most recent publication was included in the analysis.
Abstracts, letters, editorials and expert opinions, reviews without original data, case reports, and studies lacking control groups were excluded. The studies or data were also excluded when: 1) the outcomes and parameters of patients were not clearly reported (for example, with no clearly reported outcomes or standard deviations (SDs)); 2) it was impossible to extract the appropriate data from the published results; or 3) there was overlap between authors or centers in the published literature.
Outcomes of interest and data extraction
The following outcomes were used to compare the two operating techniques: 1) intra-operative data, which included operating time (min), and conversion; 2) post-operative data, which included complications, amount of drainage fluid (ml), and post-operative hospital stay (days); and 3) pathologic details, which included number of lymph nodes harvested.
Two reviewers independently extracted the following parameters from each study: 1) first author and year of publication; 2) study population characteristics; 3) number of subjects who underwent each technique; and lastly, 4) intra-operative data, post-operative data, and pathologic details.
Statistical analysis
The meta-analysis was performed using the Review Manager (RevMan) software, (version 4.2.2; Cochrane IMS;
http://ims.cochrane.org). We analyzed dichotomous variables using estimation of odds ratio (OR) with 95% confidence interval (CI) and continuous variables using weighted mean difference (WMD) with 95% CI. The pooled effect was calculated using either a fixed-effects or a random-effects model. Heterogeneity between studies was evaluated using the
χ2 and I
2 tests, and we considered heterogeneity to be present if the I
2 statistic was >50%.
P < 0.05 was considered significant.
Discussion
Meta-analysis can be used for both qualitative and quantitative evaluation of existing literature by comparing and integrating the results of different studies and taking into account variations in characteristics that could influence the overall estimate of the outcome of interest[
17]. Although meta-analysis has been traditionally applied and was mostly confined to randomized controlled trials (RCTs), meta-analytical techniques using nonrandomized controlled trials (NRCTs) might be a good method for use in some clinical settings in which either the number or the sample size of the RCTs is insufficient[
18,
19]. To our knowledge, this is the first comprehensive meta-analysis comparing RT versus ET.
RT is often perceived as being more time-consuming, because of the additional set-up time required[
20]. Operating times depend mainly on the experience and skill of the surgeon. In this meta-analysis, we found that there was no significant difference in operating time between RT and ET. This may be attributable to the shortened learning curve with RT, as it has been suggested that robotic systems make the technique easier to learn, even by relatively inexperienced endoscopic surgeons[
6]. With increasing experience, set-up time gradually decreased, and the actual time may be shorter in RT. There was no significant difference in conversion rates between RT and ET.
Although RT offers a number of advantages over ET, including improvements in manual dexterity, ergonomics, and visualization, the results of the present meta-analysis suggest that there is no additional clinical benefit for RT over ET. The disadvantages of RT are a higher rate of complications and a greater amount of drainage fluid. It has been suggested that the characteristics of RT might reduce complications because, using the Da Vinci Surgical System, robotic arms are used for retraction and dissection, and their use has been found to reduce unnecessary procedures and to minimize iatrogenic tissue injury during retraction. Consequently, our result is difficult to explain, and more studies are needed before such a conclusion can be drawn. There was no difference in post-operative hospital stay between the two groups, implying that the time required for patients to resume daily activities might be similar between RT and ET.
Oncologic outcomes after thyroid cancer surgery are affected by the extent of lymph-node dissection and the completeness of thyroidectomy[
21,
22]. Some studies have concluded that more lymph nodes are harvested via RT compared with ET, and that the robotic method may improve the long-term prognosis in patients who undergo surgery for thyroid cancer[
13,
14]. In this analysis, we found no significant differences between RT and ET in the number of lymph nodes harvested; however, long-term follow-up evaluation is necessary to evaluate the exact oncologic outcomes of RT for thyroid cancer.
In the studied articles we found significant heterogeneity in operating time and number of lymph nodes harvested, which may be explained by the differences in personnel skills, extension of lymph-node dissection, and period of the learning curve. Because of this heterogeneity, we used a random-effects model in this meta-analysis.
There are several limitations to this meta-analyis, and consequently, the results should be interpreted with caution. First, the data came from NRCTs, and the overall level of clinical evidence was low. It has been reported that NRCTs might either overestimate or underestimate the magnitude of the measured effect in an intervention study, regardless of quality scores[
23]. However, Abrahama
et al. found that meta-analyses carried out on well-designed NRCTs of surgical procedures were probably as accurate as those carried out on RCTs[
24], and all six studies included in this study were NRCTs. Second, there was heterogeneity between the two groups because it was impossible to match the patient characteristics across all of the studies. We applied a random-effects model to take variation between studies into consideration, and we believe that the heterogeneity would have had very limited influence. Finally, it is possible that investigative groups might be more likely to report positive results, and that studies with significant outcomes are more likely to be published, therefore, potential publication bias might be present in our analysis.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YJR designed the study; CZH and YJR performed the literature search and retrieved data; LS and JHG collected the data; and LS and CZH performed the research and wrote the paper. All authors read and approved the final manuscript.