Surg Endosc (2016) 30:4330–4352 and Other Interventional Techniques

DOI 10.1007/s00464-016-4752-x

Robotic surgery: disruptive innovation or unfulfilled promise?

A systematic review and meta-analysis of the first 30 years
Alan Tan1 • Hutan Ashrafian1 • Alasdair J. Scott1 • Sam E. Mason1 •

Leanne Harling1 • Thanos Athanasiou1 • Ara Darzi1,2

Received: 25 August 2015 / Accepted: 11 January 2016 / Published online: 19 February 2016
Ó The Author(s) 2016. This article is published with open access at Springerlink.com

Abstract (RoM 0.982, 0.936–1.027) and 30-day overall complication

Background Robotic surgery has been in existence for rate (RR 0.988, 0.822–1.188). In both comparisons, robotic
30 years. This study aimed to evaluate the overall periop- surgery prolonged operative time (OS: RoM 1.073,
erative outcomes of robotic surgery compared with open 1.022–1.124; MIS: RoM 1.135, 1.096–1.173). The benefits of
surgery (OS) and conventional minimally invasive surgery robotic surgery lacked robustness on RCT-sensitivity analy-
(MIS) across various surgical procedures. ses. However, many studies, including the relatively few
Methods MEDLINE, EMBASE, PsycINFO, and Clini- available RCTs, suffered from high risk of bias and inadequate
calTrials.gov were searched from 1990 up to October 2013 statistical power.
with no language restriction. Relevant review articles were Conclusions Our results showed that robotic surgery
hand-searched for remaining studies. Randomised con- contributed positively to some perioperative outcomes but
trolled trials (RCTs) and prospective comparative studies longer operative times remained a shortcoming. Better
(PROs) on perioperative outcomes, regardless of patient quality evidence is needed to guide surgical decision
age and sex, were included. Primary outcomes were blood making regarding the precise clinical targets of this inno-
loss, blood transfusion rate, operative time, length of hos- vation in the next generation of its use.
pital stay, and 30-day overall complication rate.
Results We identified 99 relevant articles (108 studies, Keywords Robotic surgery
Conventional surgery

14,448 patients). For robotic versus OS, 50 studies (11 RCTs, Perioperative outcomes
39 PROs) demonstrated reduction in blood loss [ratio of means
(RoM) 0.505, 95 % confidence interval (CI) 0.408–0.602],
transfusion rate [risk ratio (RR) 0.272, 95 % CI 0.165–0.449], Robotic surgery represents a fundamental innovation in
length of hospital stay (RoM 0.695, 0.615–0.774), and 30-day health care that is designed to enhance the quality of care
overall complication rate (RR 0.637, 0.483–0.838) in favour of for patients. Puma 560 was the first surgical robot applied
robotic surgery. For robotic versus MIS, 58 studies (21 RCTs, in a clinical setting to obtain neurosurgical biopsies in 1985
37 PROs) demonstrated reduced blood loss (RoM 0.853, [1]. The authors concluded that the robot contributed to
0.736–0.969) and transfusion rate (RR 0.621, 0.390–0.988) in improved accuracy. Since then, increasingly advanced
favour of robotic surgery but similar length of hospital stay surgical robots have been developed to assist in a rapidly
expanding range of operative procedures and anatomical
targets (Fig. 1). The drivers for continuous innovation stem
& Hutan Ashrafian from the potential to offer greater operative precision that
[email protected] may translate into enhanced clinical outcomes and the
1 accompanying background of corporate revenues within
Department of Surgery and Cancer, Imperial College
London, 10th Floor QEQM Building, St. Mary’s Hospital, the healthcare technology sector.
London W2 1NY, UK To achieve these goals, current robotic platforms are
2
Institute of Global Health Innovation, Imperial College designed to incorporate advanced features, such as, (i) dex-
London, London SW7 2NA, UK terous capability with accompanying instrumentation, (ii)

123
Surg Endosc (2016) 30:4330–4352 4331

Fig. 1 Timeline demonstrating

selected events in the history
and development of surgical
robots

augmented visualisation, (iii) improved stability, (iv) natural of robotic surgery collectively as a single entity has not been
coordination, (v) accurate cutting capacity, (vi) reliable performed. As we approach the end of the third decade
execution, and (vii) enhanced surgeon ergonomics. These following the pioneering use of the first surgical robot, an
features can theoretically increase surgical precision by overview of this innovation may be useful for understanding
rendering difficult operative tasks easier to perform safely. the adoption of innovations in health care.
Moreover, surgical robots have retained the capacity to The aim of this comprehensive systematic review and
enable surgery through smaller incisions. Collectively, these meta-analysis was to draw evidence from comparative
characteristics aim to enhance outcomes beyond that studies in robotic surgery, regardless of specialty and pro-
achievable through conventional operative methods. cedure type, and irrespective of patient age and sex. We
The adoption and diffusion of robotic surgery demonstrate avoided the biases of retrospective studies that dominate the
a positive trend in some geographical areas, particularly for literature by focussing only on randomised controlled trials
advanced economies. This can be illustrated by the prominent (RCTs) and non-randomised prospective studies. In com-
application of the da VinciÒ Surgical System (dVSS; Intuitive paring potentially very heterogeneous studies, we empha-
Surgical Inc., Mountain View, Sunnyvale, California, USA), sised a methodology that identified the proportional benefit
which has US Food and Drug Administration (FDA) clear- of robotic surgical outcomes compared with controls in each
ance across a multitude of specialties [2], demonstrating its study. This offered internal consistency from each study. We
greatest exposure for urological and gynaecological proce- were then able to calculate a pooled proportional benefit for
dures [3]. For example, more than half of radical prostatec- specific robotic surgical outcomes for all studies.
tomies and about a third of benign hysterectomies are already In this review, we evaluated core perioperative variables
performed robotically in the USA [3, 4]. as our primary outcomes. These were (i) blood loss, (ii)
Despite offering some elements of innovative technology, blood transfusion rate, (iii) operative time, (iv) length of
the necessary evidence to justify the expanding investment hospital stay, and (v) 30-day overall complication rate. In
in robotic surgery remains ambiguous. Whilst the concept of robotic surgical studies, these perioperative variables were
robotic surgery is almost universally favoured, its wide- most commonly addressed. Analyses were performed
spread promotion across all healthcare sectors requires separately for robotic versus open surgery (OS) and robotic
robust justification, not least because it can be very costly versus minimally invasive surgery (MIS). As a secondary
[5]. Studies comparing outcomes of robotic surgery with outcome, we calculated the proportion of studies that
conventional approaches for specific robots and procedures demonstrated adequate statistical power for the evaluation
are certainly not scarce. However, the systematic assessment of these clinical outcomes.

123
4332 Surg Endosc (2016) 30:4330–4352

Materials and methods (MESH, exp) OR computer-assisted surg* (tw) OR com-

puter-aided surg* (tw)] AND [intervention* (tw) OR surg*
This review was performed according to the Preferred (tw)]. Studies from 1990 to the search dates were included.
Reporting Items for Systematic Reviews and Meta-analy- There was no language restriction. Relevant review articles,
ses (PRISMA) statement [6]. including health technology assessments, found through our
search strategy were also hand-searched to identify any
remaining studies.
Inclusion and exclusion criteria
Data collection and analysis
We defined surgery as any interventional procedure
involving alteration in anatomy and that either requires a
Study selection
skin (or mucosal) incision or puncture. Patients requiring
surgery for which a robotic approach was a feasible alter-
Articles were screened from titles and abstracts by three
native approach to OS or MIS were included. There was no
authors independently (AT, SM, and AS). Potentially rel-
age or sex restriction. Controls were eligible only if
evant articles that appear to fit the inclusion and exclusion
patients underwent surgery and no robotic assistance was
criteria were obtained in full text. These were indepen-
provided. RCTs and prospective studies that addressed one
dently assessed for eligibility by the same authors. Articles
or more core perioperative surgical outcomes (blood loss,
were excluded if they had duplicate or incomplete data, or
blood transfusion rate, operative time, length of hospital
if they were only available in abstract form. Any dis-
stay, and 30-day overall complication rate) were included.
agreement was resolved through discussion with a senior
For operative time, we included studies that explicitly
author (HA).
defined it as starting from skin incision to skin closure (for
intravascular procedures, we used procedure time, which
Dealing with duplicate publications
was generally defined as time from first venous puncture to
sheath withdrawal at the end of the procedure). Whilst this
If several articles reported outcomes from a single study,
measure does not represent the total theatre occupation
the article with the most comprehensive results (most
time, it was selected to improve comparability because
number of patients and/or most recent publication) was
operative time was variedly defined in the literature.
included. If this article failed to report outcomes that were
We excluded studies where surgical robots were used for
otherwise available in the duplicate article, then the addi-
stereotactic, endoscopic, or single-incision laparoscopic
tional data from the duplicate article were included.
surgery. Robotic instrument positioners without concurrent
use of other robotic instrumentation tools were also exclu-
Data extraction
ded, as were innovations that are generally not considered
robotic technology, such as remote magnetic catheter navi-
One author (AT) extracted data into an Excel 2011 data-
gation and pure computer navigation systems. We also
base (Microsoft Corp., Redmond, Washington, USA),
discounted studies with historical controls that preceded the
which were then reviewed independently by three authors
robotic arm considerably (that is, greater than a year) as well
(SM, AS, and HA). For each article, the year of publica-
as those that retrospectively reviewed and analysed
tion, study design, total number of patients, number of
prospective databases. Laboratory studies involving syn-
patients in each arm, robot and control type, baseline
thetic models, animals, or cadavers were not considered.
characteristics, and results of outcome measures of interest
were extracted. For continuous outcomes, we extracted the
Search methodology mean and standard deviation (or if unavailable, the median
and standard error, range, or interquartile range). For cat-
Using the OvidSP search engine, the MEDLINE, EMBASE, egorical outcomes, we recorded the number of events.
and PsycINFO databases were searched on 2 September
2013 with the terms: robot* (tw) AND [intervention* (tw) Risk of bias assessment
OR surg* (tw)]. The same search terms were used to search
the ClinicalTrials.gov registry to identify potentially relevant Three authors (AT, SM, and AS) independently assessed the
trials. On 26 May 2014, these trials were reviewed to risk of bias of eligible articles. Quality of articles with more
identify any relevant published data. To avoid losing gen- than one study was assessed on their overall methodology.
erally older papers which had used the term computer-as- The Cochrane risk of bias tool [7] was applied to RCTs.
sisted instead of robot, we also performed a search on 7 Seven key domains were assessed: method of random
October 2013 with the terms: [surgery, computer-assisted sequence generation, allocation concealment, blinding of

123
Surg Endosc (2016) 30:4330–4352 4333

participants and personnel, blinding of outcome assessors, for both continuous and categorical outcomes. This was
completeness of outcome data, selective reporting, and other accomplished using Stata 13 (StataCorp., College Station,
potential sources of bias. Based on a set of listed criteria, Texas, USA). Sensitivity analysis on RCTs was also per-
each domain was judged to have either a low, high, or formed. The I2 statistic was used to estimate the degree of
unclear risk of bias. If a study had unclear or high risk of heterogeneity between studies, where larger values indicate
bias for one or more key domains, then it was classified as increasing heterogeneity [12].
having, respectively, an unclear or high risk of bias overall. Post hoc power analysis (significant at the 5 % level,
If instead all the key domains had low bias risk, then the two-tailed t test) was conducted for all eligible studies
study was judged to have a low risk of bias overall [7]. using the G*Power 3.1 programme [13]. Power was cal-
For prospective studies, the Newcastle–Ottawa scale culated for large (d = 0.8), medium (d = 0.5), and small
(NOS) [8] was used for quality scoring. The NOS judges (d = 0.2) effect sizes. We defined adequate statistical
studies on three categories: the selection of the study power as [80 %. We also identified studies with clearly
groups (comprising four numbered items: representative- specified primary outcomes and where power analysis was
ness of exposed cohort, selection of non-exposed cohort, performed to determine the required sample size for ade-
ascertainment of exposure, demonstration that outcomes quate assessment of these outcomes.
were not present at start of study), the comparability of the
groups (comprising one numbered item: comparability of
cohorts on basis of study design or analysis), and outcomes Results
(comprising three numbered items: assessment of outcome,
appropriateness of length of follow-up, adequacy of fol- Search results
low-up of cohorts). From a set of listed criteria, a maxi-
mum of one star can be awarded for each numbered item, A total of 43,132 articles were identified from the databases.
except for comparability where a maximum of two stars This included 104 trials from the ClinicalTrials.gov registry,
can be awarded. The possible NOS score ranges from 0 to 9 of which one [14] was subsequently found to contain relevant
stars. We classified studies with C7 stars as ‘‘higher’’ published data. After removing duplicates, 28,574 articles
quality and \7 stars as ‘‘lower’’ quality. were screened based on their titles and abstracts. Of these,
Risk of bias assessment was made at the level of out- 1702 potentially relevant full-text articles were retrieved for
comes. We assessed perioperative outcomes together as a further evaluation. We found 97 articles that met the inclu-
class [7, 9]. If a study addressed several perioperative sion criteria. Two additional articles were identified through
outcomes, the risk of bias for a particular domain was hand-searching. In total, 99 articles, involving 14,448
judged based on the outcome that was most affected by the patients overall, were included in this review (Fig. 2).
study methodology. Any disagreement with risk of bias
assessment was resolved through discussion with a senior Description of included studies
author (HA).
Of the included articles, all but one [15] investigated out-
Data synthesis and statistical methods comes in adult patients. Overall, there were 31 and 68
articles, respectively, that were based on RCT and non-
Meta-analysis was based on control type, that is, either randomised prospective comparative designs. They
robotic versus OS or robotic versus MIS. Wherever possible, encompassed a wide range of specialties and procedures
we used results from intention-to-treat analyses. Continuous (Tables 1, 2). Some articles comprised more than one
outcomes were analysed by calculating the ratio of means comparison or study [16–23].
(RoM) for each study, with expression of uncertainty of each
result represented by the 95 % confidence intervals (CI) Robotic versus OS
[10]. We substituted median for mean in studies where only
the median was reported. When the calculated RoM was 1, For robotic versus OS, there were 50 studies (11 RCTs and
computation was not possible. Consequently, these results 39 prospective studies) (Table 1). The year of publication
were excluded. Categorical outcomes were analysed using ranged from 1998 to 2013. In total, there were 5910 and 4237
risk ratio (RR) with 95 % CI [7]. Studies reporting cate- patients in the robotic and OS groups, respectively. The
gorical outcomes with no events in both the robotic and smallest and largest sample sizes were 14 and 1738,
control groups were excluded, as their effect sizes were not respectively. The surgical robots used in these studies were
computable. We performed meta-analysis if two or more the dVSS, ZeusÒ Robotic Surgical System (ZRSS; Computer
separate studies were available. The inverse-variance, ran- Motion Inc., Santa Barbara, California, USA), ROBODOCÒ
dom-effects model of DerSimonian and Laird [11] was used Surgical System (Curexo Technology Corp., Fremont,

123
4334 Surg Endosc (2016) 30:4330–4352

Fig. 2 Flow chart of included

studies. *Some articles
contained more than one
comparison or study (see text).
OS open surgery, MIS
minimally invasive surgery,
RCT randomised controlled trial

California, USA), AcrobotÒ Surgical System (The Acrobot patients, respectively. Sample sizes ranged from 12 to 390.
Co. Ltd., London, UK), CASPAR system (OrtoMaquet, The surgical robots used were the dVSS, ZRSS, Mona
Rastatt, Germany), and SpineAssistÒ (Mazor Robotics Ltd., (Intuitive Surgical), and SenseiÒ Robotic Catheter System
Caesarea, Israel). (Hansen Medical Inc., Mountain View, California, USA).

Robotic versus MIS Risk of bias assessment

For robotic versus MIS, there were 58 studies (21 RCTs All included articles were assessed for the quality of their
and 37 prospective studies), which were published between methodology. Of note, all 31 RCT articles suffered from a
2001 and 2014 (Table 2). Taking into account all studies, high risk of bias because they all showed a high risk of bias
the robotic and MIS groups consisted of 1991 and 2310 in the performance bias domain (Fig. 3). This was primarily

123
Surg Endosc (2016) 30:4330–4352 4335

Table 1 Studies comparing robotic versus open surgery

References Procedure Design No. of patients, n Robot Perioperative Power
outcomes addresseda
Total R C Effect size
Large Medium Small

Bertani et al. Rectal resection PRO 86 52 34 dVSS BL, LOS, C 0.948 0.611 0.146
[16]b
Kim et al. [68] Rectal resection PRO 200 100 100 dVSS BT, LOS, C 1.000 0.940 0.291
Bertani et al. Colectomy PRO 79 34 45 dVSS BL, LOS, C 0.935 0.584 0.140
[16]b
Lee et al. [47] Thyroidectomy PRO 84 41 43 dVSS BL, OT, LOS, C 0.952 0.619 0.148
Kim et al. [48] Thyroidectomy PRO 37 19 18 dVSS BL 0.657 0.315 0.091
Ryu et al. [96] Thyroidectomy PRO 90 45 45 dVSS LOS, C 0.964 0.738 0.180
Menon et al. [34] Prostatectomy PRO 60 30 30 dVSS BL, BT, OT, LOS, C 0.861 0.478 0.119
Tewari et al. [36] Prostatectomy PRO 300 200 100 dVSS BL, BT, OT, LOS, C 1.000 0.983 0.370
Farnham et al. Prostatectomy PRO 279 176 103 dVSS BL, BT 1.000 0.980 0.362
[31]
Wood et al. [35] Prostatectomy PRO 206 117 89 dVSS BL, BT, LOS, C 1.000 0.943 0.293
Nelson et al. [95] Prostatectomy PRO 1003 629 374 dVSS LOS, C 1.000 1.000 0.864
Ham et al. [37] Prostatectomy PRO 298 188 110 dVSS BL, LOS, C 1.000 0.986 0.383
Ficarra et al. [32] Prostatectomy PRO 208 103 105 dVSS BL, BT, LOS, C 1.000 0.948 0.300
Carlsson et al. Prostatectomy PRO 1738 1253 485 dVSS C 1.000 1.000 0.962
[111]
Hong et al. [38] Prostatectomy PRO 51 26 25 dVSS BL, BT 0.799 0.417 0.108
Doumerc et al. Prostatectomy PRO 714 212 502 dVSS BT, LOS, C 1.000 1.000 0.684
[69]
Kordan et al. [33] Prostatectomy PRO 1244 830 414 dVSS BL, BT 1.000 1.000 0.913
Di Pierro et al. Prostatectomy PRO 150 75 75 dVSS BT, C 0.998 0.860 0.229
[70]
Kim et al. [112] Prostatectomy PRO 763 528 235 dVSS C 1.000 1.000 0.721
Ludovico et al. Prostatectomy PRO 130 82 48 dVSS BL, LOS, C 0.992 0.780 0.194
[39]
Rhee et al. [42] Cystectomy PRO 30 7 23 dVSS BL, BT, OT, LOS 0.432 0.201 0.073
Nix et al. [25] Cystectomy RCT 41 21 20 dVSS BL, LOS, C 0.704 0.345 0.096
Ng et al. [41] Cystectomy PRO 187 83 104 dVSS BL, OT, LOS, Cc 1.000 0.922 0.272
Martin et al. [40] Cystectomy PRO 33 19 14 dVSS BL, LOS 0.595 0.280 0.085
Khan et al. [17]b Cystectomy PRO 100 48 52 dVSS BL, BT, OT, LOS, C 0.977 0.696 0.167
Parekh et al. [26] Cystectomy RCT 40 20 20 dVSS BL, BT, OT, LOS, C 0.693 0.338 0.095
Masson-Lecomte Nephrectomy PRO 100 42 58 dVSS BL, BT, OT, LOS, C 0.974 0.686 0.165
et al. [49]
Parekattil et al. Vasovasostomy PRO 94 66 28 dVSS OT 0.939 0.592 0.142
[23]b
Parekattil et al. Vasoepididymostomy PRO 61 44 17 dVSS OT 0.787 0.406 0.106
[23]b
Bucerius et al. CABG PRO 117 24 93 dVSS LOS 0.934 0.581 0.139
[18]b
Kiaii et al. [94] CABG PRO 100 50 50 ZRSS LOS, C 0.977 0.697 0.168
Poston et al. [43] CABG PRO 200 100 100 dVSS BL, LOS, C 1.000 0.940 0.291
Bachinsky et al. CABG PRO 52 25 27 dVSS BT, LOS 0.807 0.424 0.109
[71]
Balduyck et al. Anterior mediastinal PRO 36 14 22 dVSS C 0.623 0.295 0.088
[110] mass resection
Hoekstra et al. Endometrial cancer PRO 58 32 26 dVSS BL, OT, LOS, C 0.846 0.461 0.116
[20]b staging

123
4336 Surg Endosc (2016) 30:4330–4352

Table 1 continued
References Procedure Design No. of patients, n Robot Perioperative Power
outcomes addresseda
Total R C Effect size
Large Medium Small

Göçmen et al. Endometrial cancer PRO 22 10 12 dVSS BL, BT, OT, LOS, C 0.428 0.199 0.073
[45] staging
Jung et al. [19]b Endometrial cancer PRO 84 28 56 dVSS BT, OT, LOS, C 0.927 0.569 0.137
staging
Lowe et al. [46] Hysterectomy— PRO 14 7 7 dVSS BL, OT, LOS, C 0.281 0.139 0.064
cervical cancer
Collins et al. [44] Sacrocolpopexy PRO 48 30 18 dVSS BL 0.748 0.375 0.101
Bargar et al. [28] Total hip arthroplasty RCT 136 70 66 ROBODOC BL, LOS, C 0.996 0.825 0.212
Bach et al. [109] Total hip arthroplasty PRO 50 25 25 ROBODOC C 0.791 0.410 0.107
Honl et al. [76] Total hip arthroplasty RCT 141 61 80 ROBODOC OT, C 0.997 0.832 0.215
Siebel et al. [113] Total hip arthroplasty PRO 71 36 35 CASPAR C 0.914 0.547 0.132
Nishihara et al. Total hip arthroplasty RCT 156 78 78 ROBODOC BL, BT 0.999 0.873 0.237
[27]
Nakamura et al. Total hip arthroplasty RCT 146 75 71 ROBODOC C 0.998 0.851 0.224
[107]
Cobb et al. [106] Unicompartmental RCT 28 13 15 Acrobot C 0.529 0.246 0.080
knee arthroplasty
Park et al. [108] Total knee arthroplasty RCT 62 32 30 ROBODOC C 0.872 0.490 0.121
Song et al. [29] Total knee arthroplasty RCT 60 30 30 ROBODOC BL, C 0.861 0.478 0.119
Song et al. [30] Total knee arthroplasty RCT 100 50 50 ROBODOC BL, C 0.977 0.697 0.168
Ringel et al. [77] Spinal pedicle screw RCT 60 30 30 SpineAssist OT, LOS 0.861 0.478 0.119
insertion
Total 10,147 5910 4237
BL blood loss, BT blood transfusion rate, OT skin-to-skin operative (or procedure) time, LOS length of hospital stay, C 30-day overall
complication rate, CABG coronary artery bypass grafting, RCT randomised controlled trial, PRO non-randomised prospective comparative
studies
a
Relevant to this review
b
More than one comparison or study in an article
c
Not computable, as there were more complications than the number of patients in the open group—complication data were excluded from
meta-analysis as a result; for robotic studies on hips and knees, n = number of limbs

due to the lack of surgeon blinding, which is unlikely to be studies also selected their control cohort from the same
possible in clinical trials of robotic surgery. As periopera- community as the robotic cohort and showed adequate fol-
tive outcomes are especially vulnerable to performance low-up. Many of them suffered from poor comparability, as
bias, this risk was judged to be high. The subject of patient expected from the lack of randomisation where selection
blinding, which is difficult in surgical trials but potentially bias is a caveat. In some cases, the representativeness of the
feasible [24], was frequently unaddressed or unreported by robotic cohort in the community was felt not be adequate.
authors. Most RCTs showed low risk of attrition bias, with
complete perioperative outcome data. In many trials, how-
Meta-analyses of perioperative surgical outcomes
ever, the risk of bias related to sequence generation, allo-
cation concealment, blinding of outcome assessor, and
(i) Blood loss
selective reporting was unclear, as sufficient information
was not available due to poor reporting.
Of 68 articles of non-randomised prospective design, 55 Robotic versus OS
(80.9 %) were of ‘‘higher’’ quality (Tables 3, 4). All
prospective studies met the criteria for ascertainment of There were six RCT [25–30] and 23 prospective [16, 17,
exposure, absence of outcome at the start of study, outcome 20, 31–49] studies reporting on blood loss, giving a total of
assessment, and duration of follow-up. Most prospective 29 studies overall. Meta-analysis demonstrated blood loss

123
Surg Endosc (2016) 30:4330–4352 4337

Table 2 Studies comparing robotic versus minimally invasive surgery

References Procedure Design No. of patients, n Robot Perioperative Power
outcomes addresseda
Total R C Effect size
Large Medium Small

Pigazzi et al. [56] Rectal resection PRO 12 6 6 dVSS BL, LOS, C 0.241 0.123 0.061
Patriti et al. [55] Rectal resection PRO 66 29 37 dVSS BL, BT, LOS, C 0.888 0.510 0.125
Baik et al. [89] Rectal resection PRO 113 56 57 dVSS OT, LOS, C 0.984 0.726 0.176
Kim et al. [75] Rectal resection PRO 209 62 147 dVSS BT, OT, LOS, C 1.000 0.908 0.260
Bertani et al. [16]b Colectomy PRO 64 34 30 dVSS BL, LOS, C 0.882 0.502 0.123
Park et al. [51] Colectomy RCT 70 35 35 dVSS BL, BT, LOS, C 0.910 0.541 0.131
Jiménez Rodrı́guez et al. [74] Colorectal RCT 56 28 28 dVSS BT, LOS, C 0.836 0.451 0.114
resection
Heemskerk et al. [102] Rectopexy PRO 33 14 19 dVSS LOS 0.595 0.280 0.085
Wong et al. [57] Rectopexy PRO 63 23 40 dVSS BL, LOS, C 0.853 0.468 0.117
Cadière et al. [97] Fundoplication RCT 21 10 11 Mona LOS, C 0.412 0.193 0.072
Melvin et al. [84] Fundoplication PRO 40 20 20 dVSS OT, C 0.693 0.338 0.095
Draaisma et al. [50] Fundoplication RCT 50 25 25 dVSS BL, OT, LOS, C 0.791 0.410 0.107
Morino et al. [78] Fundoplication RCT 50 25 25 dVSS OT, LOS, C 0.791 0.410 0.107
Nakadi et al. [99] Fundoplication RCT 20 9 11 dVSS LOS, C 0.392 0.184 0.071
Lehnert et al. [15] Fundoplication PRO 20 10 10 dVSS OT, C 0.395 0.185 0.071
Müller-Stitch et al. [98] Fundoplication RCT 40 20 20 dVSS LOS, C 0.693 0.338 0.095
Hartmann et al. [101] Fundoplication PRO 80 18 62 dVSS LOS, C 0.839 0.454 0.114
Sanchez et al. [100] RYGB RCT 50 25 25 dVSS LOS, C 0.791 0.410 0.107
Benizri et al. [85] RYGB PRO 200 100 100 dVSS OT, LOS, C 1.000 0.940 0.291
Mühlmann et al. [104] Various bariatricc PRO 20 10 10 dVSS LOS, C 0.395 0.185 0.071
Park et al. [91] Gastrectomy PRO 150 30 120 dVSS OT, LOS, C 0.973 0.682 0.164
Ruurda et al. [79] Cholecystectomy RCT 20 10 10 dVSS OT 0.395 0.185 0.071
Nio et al. [105] Cholecystectomy PRO 20 10 10 ZRSS LOS, C 0.395 0.185 0.071
Zhou et al. [52] Cholecystectomy RCT 40 20 20 ZRSS BL, LOS, C 0.693 0.338 0.095
Kornprat et al. [90] Cholecystectomy PRO 46 20 26 ZRSS OT 0.749 0.376 0.101
Berber et al. [58] Liver resection PRO 32 9 23 dVSS BL, OT, C 0.504 0.234 0.078
Brunaud et al. [93] Adrenalectomy PRO 28 14 14 dVSS OT, LOS, C 0.531 0.247 0.080
Morino et al. [81] Adrenalectomy RCT 20 10 10 dVSS OT, LOS, C 0.395 0.185 0.071
Wu et al. [66] Adrenalectomy PRO 12 5 7 ZRSS BL, LOS, C 0.236 0.121 0.061
Ploussard et al. [59] Prostatectomy PRO 288 83 205 dVSS BL, BT, LOS, C 1.000 0.969 0.335
Gosseine et al. [67] Prostatectomy PRO 247 122 125 dVSS BL, BT, LOS, C 1.000 0.975 0.347
Asimakopoulos et al. [72] Prostatectomy RCT 112 52 60 dVSS BT, C 0.987 0.744 0.182
Porpiglia et al. [53] Prostatectomy RCT 120 60 60 dVSS BL, OT, LOS, C 0.991 0.775 0.192
Khan et al. [17]b Cystectomy PRO 106 48 58 dVSS BL, BT, OT, LOS, C 0.982 0.719 0.174
Caruso et al. [63] Nephrectomy PRO 20 10 10 dVSS BL, BT, LOS, C 0.395 0.185 0.071
Hemal et al. [60] Nephrectomy PRO 30 15 15 dVSS BL, BT, LOS, C 0.562 0.262 0.083
Kural et al. [62] Nephrectomy PRO 31 11 20 dVSS BL, BT, LOS, C 0.540 0.251 0.081
Masson-Lecomte et al. [61] Nephrectomy PRO 265 220 45 dVSS BL, BT, LOS, C 0.998 0.861 0.230
Bucerius et al. [18]b CABG PRO 97 24 73 dVSS LOS 0.920 0.557 0.134
Mierdl et al. [92] CABG PRO 46 30 16 dVSS OT, C 0.715 0.352 0.097
Sarlos et al. [54] Hysterectomy— RCT 95 47 48 dVSS BL, OT, LOS, C 0.971 0.674 0.162
benign disease
Paraiso et al. [73] Hysterectomy— RCT 52 26 26 dVSS BT, OT 0.807 0.424 0.109
benign disease
Hoekstra et al. [20]b Endometrial cancer PRO 39 32 7 dVSS BL, OT, LOS, C 0.463 0.215 0.075
staging

123
4338 Surg Endosc (2016) 30:4330–4352

Table 2 continued
References Procedure Design No. of patients, n Robot Perioperative Power
outcomes addresseda
Total R C Effect size
Large Medium Small
b
Jung et al. [19] Endometrial cancer PRO 53 28 25 dVSS BT, OT, LOS, C 0.814 0.430 0.110
staging
Paraiso et al. [80] Sacrocolpopexy RCT 68 35 33 dVSS OT, LOS, C 0.901 0.528 0.128
Seror et al. [64] Sacrocolpopexy PRO 67 20 47 dVSS BL, LOS, C 0.839 0.454 0.114
Anger et al. [14] Sacrocolpopexy RCT 78 40 38 dVSS BL, C 0.937 0.587 0.141
El Hachem et al. [65] Various PRO 91 39 52 dVSS BL, LOS, C 0.962 0.646 0.154
gynaecological—
unspecified
Kolvenbach et al. [103] AAA repair PRO 39 8 31 ZRSS LOS 0.502 0.233 0.078
Malcolme-Lawes et al. [21]b AF ablation—robot RCT 20 10 10 Sensei BT, OT, C 0.395 0.185 0.071
30 s
Malcolme-Lawes et al. [21]b AF ablation—robot RCTd 20 10 10 Sensei BT, OT, C 0.395 0.185 0.071
60 s
Steven et al. [83] AF ablation RCT 50 25 25 Sensei OT, C 0.791 0.410 0.107
Kautzner et al. [87] AF ablation PRO 38 22 16 Sensei OT, C 0.659 0.316 0.091
Di Biase et al. [86] AF ablation PRO 390 193 197 Sensei OT, C 1.000 0.998 0.504
Steven et al. [82] AF ablation RCT 60 30 30 Sensei OT 0.861 0.478 0.119
Tilz et al. [22]b AF ablation—robot PRO 29 4 25 Sensei OT 0.299 0.146 0.065
30W
Tilz et al. [22]b AF ablation—robot PRO 35 10 25 Sensei OT 0.546 0.254 0.081
20W
Rillig et al. [88] AF ablation PRO 70 50 20 Sensei OT 0.846 0.461 0.116
Total 4301 1991 2310
BL blood loss, BT blood transfusion rate, OT skin-to-skin operative (or procedure) time, LOS length of hospital stay, C 30-d overall complication
rate, RYGB Roux-en-Y gastric bypass, CABG coronary artery bypass grafting, AAA abdominal aortic aneurysm, AF atrial fibrillation/flutter, RCT
randomised controlled trial, PRO non-randomised prospective comparative study
a
Relevant to this review
b
More than one comparison or study in an article
c
Gastric banding, implantable gastric stimulator, band revision
d
Quasi-RCT (10 patients who underwent robotic AF ablation of 60-s duration were not randomized compared with 10 control patients that were
randomised); for Baik 2009, n = 57 (control) for C and n = 51 (control) for OT and LOS, as 6 converted cases were excluded from analysis by
authors; for Sarlos 2012, n = 47 (robotic) and n = 48 (control) for analysis of C, as no operations were performed in 5 patients, and for BL, OT,
and LOS, n = 50 in each arm, as missing values were replaced with median of available measurements in respective study arm; for Mierdl 2005,
n = 30 (robotic) for analysis of C but n = 24 for OT, as data not shown for 6 patients

in the robotic arm to be 50.5 % of that in the OS arm 85.3 % of that experienced by patients in the MIS arm
(Fig. 4). This reduction was significant (95 % CI (95 % CI 0.736–0.969) (Fig. 4). The heterogeneity was
0.408–0.602). There was high heterogeneity in the results high (I2 = 98.2 %). Sensitivity analysis performed on
(I2 = 98.0 %). Sensitivity analysis on RCTs showed RCTs and, however, revealed a non-robust result (pooled
reduction in blood loss, but this was no longer significant RoM: 0.830, 95 % CI 0.653–1.008, I2 = 95.9 %).
(pooled RoM: 0.807, 95 % CI 0.563–1.051, I2 = 96.3 %).
(ii) Blood transfusion rate
Robotic versus MIS
Robotic versus OS
Twenty-two studies reported blood loss as an outcome
measure. Of these, six were RCT studies [14, 50–54] and Blood transfusion rate was investigated in two RCT [26,
16 were prospective studies [16, 17, 20, 55–67]. Meta- 27] and 16 prospective [17, 19, 31–36, 38, 42, 45, 49, 68–
analysis of these studies confirmed a significant reduction 71] studies. Forty-two of 2127 patients (2.0 %) in the
in blood loss in favour of robotic surgery, which was robotic group needed blood transfusion compared with 249

123
Surg Endosc (2016) 30:4330–4352 4339

Fig. 3 Risk of bias graphs of

randomised controlled trials
comparing robotic versus open
surgery (above) and robotic
versus minimally invasive
surgery (below)

of 1869 patients (13.3 %) in the open group. One study Robotic versus OS
[27] was excluded from quantitative synthesis, as its effect
size was not computable. Meta-analysis of the remaining Sixteen studies assessed operative time. These comprised
17 studies demonstrated the risk of blood transfusion with three RCT [26, 76, 77] and 13 prospective [17, 19, 20, 23,
robotic surgery to be 27.2 % of that of OS. This reduction 34, 36, 41, 42, 45–47, 49] studies. Meta-analysis showed
in favour of robotic surgery was significant (95 % CI robotic surgery to increase operative time by 7.3 %, which
0.165–0.449). The results showed moderate heterogeneity was significant (95 % CI 1.022–1.124). High heterogeneity
(I2 = 55.2 %). Sensitivity analysis on RCTs was not done, was found (I2 = 91.8 %). Sensitivity analysis on RCTs
as only one study was available. In this RCT, no significant showed a consistent result (pooled RoM: 1.162, 95 % CI
difference in blood transfusion requirement was demon- 1.016–1.308, I2 = 86.8 %).
strated (RR 0.800, 95 % CI 0.400–1.600) [26].
Robotic versus MIS
Robotic versus MIS
Operative time was investigated by 12 RCT [21, 50, 53, 54,
Six RCT [21, 51, 72–74] and ten prospective [17, 55, 59–
73, 78–83] and 18 prospective [15, 17, 19, 20, 22, 58, 75,
63, 67, 75] studies reported blood transfusion requirement.
84–93] studies. There was a significant prolongation of
Taking all these studies together, 4.2 % (33/789) of
operative time by 13.5 % over MIS when surgical robots
patients who underwent robotic intervention compared
were utilised (95 % CI 1.096–1.173). Heterogeneity was
with 6.5 % (56/856) of MIS patients received blood
high (I2 = 92.3 %). When only RCTs were considered in a
transfusion. Computation of valid RR was not possible in
sensitivity analysis, the result remained robust (pooled
three studies [21, 51, 63], hence their exclusion from meta-
RoM: 1.202, 95 % CI 1.119–1.286, I2 = 87.1 %).
analysis. From the remaining 13 studies, we demonstrated
a significant reduction in the requirement for blood trans- (iv) Length of hospital stay
fusion in patients who underwent robotic surgery compared
with MIS (pooled RR 0.621, 95 % CI 0.390–0.988). The
Robotic versus OS
heterogeneity was low (I2 = 0.0 %). Nevertheless, the
result of sensitivity analysis on RCTs was inconsistent
Thirty studies compared length of hospital stay between
(pooled RR 1.329, 95 % CI 0.325–5.438, I2 = 0.0 %).
robotic and open interventions. There were 4 RCT [25, 26,
(iii) Operative time (skin-to-skin) 28, 77] and 26 prospective [16–20, 32, 34–37, 39–43, 45–

123
 Table 3 Risk of bias of non-randomised prospective comparative cohort studies comparing robotic versus open surgery based on the Newcastle–Ottawa scale
4340

References Selection Comparability Outcome Total score

123
Representativeness Selection of Ascertainment Absence of outcome Assessment of outcome Duration of Adequacy
of robotic cohort control cohort of exposure at start of study follow-up of follow-up

Menon et al. [34] 0 1 1 1 1 1 1 1 7 Higher

Bucerius et al. [18]a 0 1 1 1 1 1 1 1 7 Higher
Bach et al. [109] 0 1 1 1 1 1 1 1 7 Higher
Tewari et al. [36] 1 1 1 1 1 1 1 1 8 Higher
Siebel et al. [113] 0 1 1 1 1 1 1 1 7 Higher
Farnham et al. [31] 1 1 1 1 1 1 1 1 8 Higher
Rhee et al. [42] 0 1 1 1 0 1 1 1 6 Lower
Kiaii et al. [94] 1 1 1 1 1 1 1 1 8 Higher
Wood et al. [35] 1 1 1 1 1 1 1 1 8 Higher
Nelson et al. [95] 1 1 1 1 0 1 1 1 7 Higher
Ham et al. [37] 1 1 1 1 1 1 1 1 8 Higher
Poston et al. [43] 1 0 1 1 2 1 1 1 8 Higher
Ficarra et al. [32] 1 1 1 1 1 1 1 1 8 Higher
Hoekstra et al. [20]a 1 1 1 1 0 1 1 1 7 Higher
Lowe et al. [46] 0 1 1 1 1 1 1 1 7 Higher
Göçmen et al. [45] 0 1 1 1 1 1 1 1 7 Higher
a
Jung et al. [19] 0 1 1 1 2 1 1 1 8 Higher
Carlsson et al. [111] 1 1 1 1 0 1 1 1 7 Higher
Hong et al. [38] 0 1 1 1 0 1 1 1 6 Lower
Doumerc et al. [69] 0 1 1 1 0 1 1 1 6 Lower
Kordan et al. [33] 1 1 1 1 0 1 1 1 7 Higher
Lee et al. [47] 1 1 1 1 2 1 1 1 9 Higher
Ng et al. [41] 1 1 1 1 1 1 1 1 8 Higher
Bertani et al. [16]a 1 1 1 1 2 1 1 1 9 Higher
Di Pierro et al. [70] 1 1 1 1 1 1 1 1 8 Higher
Kim et al. [112] 1 1 1 1 0 1 1 1 7 Higher
Martin et al. [40] 0 0 1 1 0 1 1 1 5 Lower
Balduyck et al. [110] 0 1 1 1 1 1 1 1 7 Higher
Kim et al. [68] 1 1 1 1 1 1 1 1 8 Higher
Khan et al. [17]a 0 1 1 1 0 1 1 1 6 Lower
Parekattil et al. [23]a 0 1 1 1 1 1 1 1 7 Higher
Bachinsky et al. [71] 0 1 1 1 1 1 1 1 7 Higher
Collins et al. [44] 0 1 1 1 2 1 1 1 8 Higher
Surg Endosc (2016) 30:4330–4352
Surg Endosc (2016) 30:4330–4352 4341

Higher
Higher
Higher
Higher
47, 49, 68, 69, 71, 94–96] studies. The result for one study
Total [26] was not computable. Meta-analysis of the remaining
score
29 studies revealed length of stay for patients who under-

9
7
7
8
of follow-up went robotic surgery to be 69.5 % of those who underwent
OS. This decrease was significant (95 % CI 0.615–0.774).
Adequacy

Heterogeneity was high (I2 = 98.5 %). In contrast, when

only RCTs were considered, the improvement in length of
1
1
1
1
stay was lost (pooled RoM: 1.038, 95 % CI 0.878–1.197,
Duration of

I2 = 89.4 %).
follow-up

Robotic versus MIS

1
1
1
1
Assessment of outcome

Length of hospital stay was addressed by 40 studies, of

which 13 were RCT [50–54, 74, 78, 80, 81, 97–100] and
were prospective [16–20, 55–57, 59–66, 75, 85, 89, 91, 101–
105] studies. Ten studies [16, 20, 50, 52, 57, 91, 97, 100,
Outcome

104, 105] were excluded from meta-analysis, as their effect

sizes were not computable. Meta-analysis of the remaining
1
1
1
1

30 studies showed no significant difference in duration of

Comparability

stay (pooled RoM: 0.982, 95 % CI 0.936–1.027). High

heterogeneity was noted (I2 = 93.4 %). Sensitivity analysis
on RCTs remained robust (pooled RoM: 1.001, 95 % CI
0.955–1.047, I2 = 80.2 %).
2
0
1
1
Absence of outcome

(v) Overall complication rate (30 day)

at start of study

Robotic versus OS
Quality of articles with more than one study was assessed on their overall methodology

Overall complications were compared in nine RCT [25, 26,

1
1
1
1

28–30, 76, 106–108] and 28 prospective [16, 17, 19, 20, 32,
Ascertainment

34–37, 39, 41, 43, 45–47, 49, 68–70, 94–96, 109–113]

of exposure

studies. From these studies, the overall complication rate

was 11.6 % (515/4453) in the robotic arm compared with
21.4 % (693/3245) in the open arm. Results from three
1
1
1
1

studies [29, 96, 109] did not allow for computable RRs.
control cohort
Selection of

From the remaining 34 studies, meta-analysis demonstrated

a significant decrease in overall complication rate in favour
of robotic surgery, which was 63.7 % of that with OS
1
1
1
1

(95 % CI 0.483–0.838). High heterogeneity was present

(I2 = 81.9 %). Sensitivity analysis on RCTs was, however,
Representativeness
of robotic cohort

inconsistent with the primary analysis (pooled RR 1.090,

95 % CI 0.631–1.881, I2 = 59.9 %).
Selection

Robotic versus MIS

1
1
0
1
Masson-Lecomte et al. [49]

Forty-eight studies investigated complications. There were

18 RCT [14, 21, 50–54, 72, 74, 78, 80, 81, 83, 97–100] and
Ludovico et al. [39]

30 prospective [15–17, 19, 20, 55–66, 75, 84–87, 89, 91,

Table 3 continued

92, 101, 104, 105] studies. Taking all these studies into
Kim et al. [48]
Ryu et al. [96]

consideration, the overall complication rate in the robotic

References

arm was 16.1 % (288/1789) compared with 15.7 % (317/

2025) in the MIS arm. Valid effect sizes in the form of RR
were not producible from results of nine studies [15, 52, 66,
a

123
 Table 4 Risk of bias of non-randomised prospective comparative cohort studies comparing robotic versus minimally invasive surgery based on the Newcastle–Ottawa scale
4342

References Selection Comparability Outcome Total score

123
Representativeness Selection of Ascertainment Absence of outcome Assessment Duration of Adequacy
of robotic cohort control cohort of exposure at start of study of outcome follow-up of follow-up

Melvin et al. [84] 1 0 1 1 0 1 1 1 6 Lower

Bucerius et al. [18]a 0 1 1 1 1 1 1 1 7 Higher
Mühlmann et al. [104] 0 1 1 1 0 1 1 1 6 Lower
Brunaud et al. [93] 1 1 1 1 1 1 1 1 8 Higher
Nio et al. [105] 1 1 1 1 2 1 1 1 9 Higher
Kolvenbach et al. [103] 0 1 1 1 0 1 1 0 5 Lower
Mierdl et al. [92] 0 1 1 1 0 1 1 0 5 Lower
Pigazzi et al. [56] 0 1 1 1 0 1 1 1 6 Lower
Lehnert et al. [15] 1 1 1 1 0 1 1 1 7 Higher
Kornprat et al. [90] 0 1 1 1 2 1 1 1 8 Higher
Caruso et al. [63] 1 0 1 1 2 1 1 1 8 Higher
Heemskerk et al. [104] 0 1 1 1 1 1 1 1 7 Higher
Wu et al. [68] 1 1 1 1 1 1 1 1 8 Higher
Patriti et al. [57] 0 1 1 1 0 1 1 0 5 Lower
Baik et al. [91] 1 1 1 1 2 1 1 0 8 Higher
Hartmann et al. [102] 1 1 1 1 2 1 1 1 9 Higher
Ploussard et al. [59] 1 1 1 1 1 1 1 1 8 Higher
Gosseine et al. [67] 1 1 1 1 1 1 1 1 9 Higher
Hemal et al. [60] 1 1 1 1 2 1 1 1 9 Higher
Kural et al. [62] 0 1 1 1 2 1 1 1 8 Higher
Hoekstra et al. [20]a 1 1 1 1 0 1 1 1 7 Higher
Kautzner et al. [87] 0 0 1 1 0 1 1 1 5 Lower
Di Biase et al. [86] 1 1 1 1 2 1 1 1 9 Higher
Berber et al. [58] 1 0 1 1 2 1 1 1 8 Higher
Jung et al. [19]a 0 1 1 1 2 1 1 1 8 Higher
Tilz et al. [22]a 0 1 1 1 0 1 1 1 6 Lower
Bertani et al. [16]a 1 1 1 1 2 1 1 1 9 Higher
Wong et al. [57] 1 1 1 1 1 1 1 1 8 Higher
Kim et al. [75] 0 1 1 1 1 1 1 1 7 Higher
Park et al. [91] 0 1 1 1 2 1 1 1 8 Higher
Khan et al. [17]a 0 1 1 1 0 1 1 1 6 Lower
Seror et al. [64] 1 1 1 1 1 1 1 1 8 Higher
Rillig et al. [88] 0 1 1 1 2 1 1 1 8 Higher
Surg Endosc (2016) 30:4330–4352
Surg Endosc (2016) 30:4330–4352 4343

Fig. 4 Forest plots of blood loss; robotic versus open surgery c

Higher
Higher
Higher
Total score (above), robotic versus minimally invasive surgery (below)

7
7
8
78, 83, 84, 87, 104, 105]. Meta-analysis involving the
of follow-up

remaining 39 studies demonstrated no significant differ-

Adequacy

ence in overall complication rate between robotic and MIS

(pooled RR 0.988, 95 % CI 0.822–1.188). Heterogeneity
1
1
1

was low (I2 = 23.0 %). When sensitivity analysis was

Duration of

performed on RCTs, the result remained robust (pooled

follow-up

RR 1.187, 95 % CI 0.851–1.654, I2 = 15.4 %).

Results of our meta-analyses are summarised in Fig. 5.
1
1
1

Post hoc power analyses

Assessment
of outcome
Outcome

With respect to RCT studies, for large effect sizes, just 17

1
1
1

[14, 27–30, 51, 53, 54, 72–74, 76, 77, 80, 82, 107, 108] of
32 studies (53.1 %) had adequate statistical power (that is,
Comparability

power [80 %). This fell to four studies [27, 28, 76, 107]
(12.5 %) for medium effect sizes. For small effect sizes,
no RCT study had adequate power.
1
0
1

Analysis of the 76 prospective studies revealed that just

47 [16–20, 23, 31–37, 39, 41, 43, 47, 49, 55, 57, 59, 61, 64,
Absence of outcome

65, 67–71, 75, 85, 86, 88, 89, 91, 94–96, 101, 111–113] of
at start of study

them (61.8 %) had adequate power for outcome evalua-

tion, assuming large effect sizes. For medium effect sizes,
20 studies [31–33, 35–37, 41, 43, 59, 61, 67–70, 75, 85,
86, 95, 111, 112] (26
3 %) were sufficiently powered.
Quality of articles with more than one study was assessed on their overall methodology
1
1
1

Only three studies [33, 95, 111] (4.2 %) had adequate

Ascertainment

power for small effect sizes.

of exposure

The lack of statistical power in many studies is not sur-

prising given that in only 16 RCT (50 %) and six prospective
(7.9 %) studies were primary outcomes clearly defined and a
1
1
1

priori power analysis performed (Table 5). Furthermore,

control cohort

only a handful of these studies [51, 54, 73, 80, 82, 85] were
Selection of

powered to the outcomes investigated in this review.

Results of post hoc power analyses for individual
studies are presented in Tables 1 and 2.
1
1
1
Representativeness
of robotic cohort

Discussion
Selection

The term ‘‘disruptive innovation’’ represents a process

where a product establishes itself at the bottom of a market
0
1
1

and climbs through this sector to displace competitors

Masson-Lecomte et al. [61]

[114]. Initial characteristics of a disruptive innovation

model include: (i) simpler products and services, (ii)
El Hachem et al. [65]

smaller target markets, and (iii) lower gross margins. As a

Table 4 continued

Benizri et al. [85]

result, these innovations can ‘‘create space’’ at the bottom

of the market to allow new disruptive competitors to
References

emerge. Currently in the field of robotic surgery, the

promise of simplicity has yet to be translated into daily
practice. Furthermore, the evidence regarding cost efficacy
a

123
4344 Surg Endosc (2016) 30:4330–4352

123
Surg Endosc (2016) 30:4330–4352 4345

Fig. 5 Pooled proportional

change in perioperative
outcomes for robotic versus
open surgery and robotic versus
minimally invasive surgery,
with 95 % confidence interval.
RoM ratio of means, RR risk
ratio, OS open surgery, MIS
minimally invasive surgery.
*Significant effect

and gross margins has been poorly documented so that When RCTs were analysed separately, the proportional
decisions regarding the adoption of robotic surgery remain benefits of robotic surgery were lost. Given their higher level
controversial. of evidence, these RCTs may be considered as more rep-
However, to disregard robotic surgery completely as an resentative of the true population effect, although they are
unfulfilled promised in its 30 years of existence may be limited by a profound lack of numbers. We identified only
imbalanced. Our meta-analyses of all RCTs and prospective 31 clinical RCTs on robotic surgery, which is a fraction
studies to date, regardless of specialty and procedure type, (0.1 %) of the 28,574 potentially relevant articles. Many
revealed a decrease in blood loss and blood transfusion rate RCTs failed to clearly define primary outcomes and perform
with robotic surgery when compared with both OS and MIS. a priori power analysis, which led to inadequate sample sizes
Additionally, comparison against OS demonstrated a reduc- and hence, statistical power necessary for outcome evalua-
tion in length of hospital stay and overall complication rate in tion. Through post hoc analyses, we showed that just over
favour of robotic surgery. half of all RCTs were adequately powered to detect a true
The ability of robotic surgery to reduce blood loss and difference in outcomes for large effect sizes. For smaller
need for blood transfusion may be attributed to its effect sizes, this deficiency, inevitably, was further ampli-
advanced features, which could improve surgical precision. fied. These findings are probably related to common barriers
This would be important in avoiding injury to vessels and in undertaking successful surgical RCTs, including ethical
other structures that can cause unintended bleeding. The issues, challenging patient recruitment and randomisation
additional benefits of robotic surgery over OS, in the form due partly to lack of equipoise, learning curve, inexperience
of shorter length of hospital stay and fewer complications, in designing trials, inadequate medical statistical knowledge,
may partly be due to its capacity for minimal access. These problematic long-term follow-up, and insufficient funding
benefits have been demonstrated in conventional minimally and resources [24, 119]. Furthermore, difficulty in blinding
invasive surgical procedures [115–118], where the positive is a major methodological barrier [120, 121]. Consequently,
effect of reduced tissue trauma has been implicated [118]. all included RCTs were considered to suffer from a high risk
Given its added features, the inability of robotic surgery to of performance bias, and accordingly, a high risk of bias
achieve improved length of stay and complication rate over overall [7]. Together, these factors could explain the non-
MIS can be considered surprising. This may be reflective of robust results.
the status that surgical robots have not yet exceeded their The demonstration of longer operative time with robotic
effects beyond those of conventional minimally invasive surgery contradicts its proposed aims of facilitating oper-
platforms for these outcomes. Alternatively, these out- ative tasks that would otherwise be difficult to perform
comes may be inadequate markers for accurately capturing efficiently with conventional tools. One possible explana-
the increased precision of robotic surgery. More sensitive tion is the requirement for additional steps in their
assessment tools of precision are advocated in future trials, deployment. For example, docking is needed for surgical
which might include video appraisal of intraoperative tis- robots such as the dVSS [73, 80]. Hardware issues could
sue handling, errors, and efficiency [52, 105]. also explain the longer operative time, as surgical robotic

123
4346 Surg Endosc (2016) 30:4330–4352

Table 5 Studies with clearly defined primary outcomes and where power analysis was undertaken a priori
Study Procedure Design Primary outcome
OT LOS C Onc Func Cost Other

Robot versus MIS

Draaisma et al. [50] Fundoplication RCT 4 Barium swallow,
manometry, ph study
Morino et al. [78] Fundoplication RCT 4
Steven et al. [83] AF ablation RCT 4 Radiofrequency duration
Steven et al. [82] AF ablation RCT 4
Asimakopoulos et al. Prostatectomy RCT 4
[72]
Paraiso et al. [80] Sacrocolpopexy RCT 4
Park et al. [51] Colectomy RCT 4
Sarlos et al. [54] Hysterectomy—benign disease RCT 4
Porpiglia et al. [53] Prostatectomy RCT 4
Paraiso et al. [73] Hysterectomy—benign disease RCT 4
Benizri et al. [85] Roux-en-Y gastric bypass PRO 4
El Hachem et al. [65] Various gynae—unspecified PRO 4 Pain
Anger et al. [14] Sacrocolpopexy RCT 4
Robot versus OS
Cobb et al. [106] Unicompartmental knee RCT 4 Leg alignment
arthroplasty
Wood et al. [35] Prostatectomy PRO 4 Quality of life
Nix et al. [25] Cystectomy RCT 4
Hong et al. [38] Prostatectomy PRO 4 Venous gas embolism
Song et al. [29] Total knee arthroplasty RCT 4 Leg alignment
Ringel et al. [77] Spinal pedicle screw insertion RCT 4 Implant position
Collins et al. [44] Sacrocolpopexy PRO 4 Return to baseline activity
(accelerometer)
Song et al. [30] Total knee arthroplasty RCT 4 Leg alignment
Kim et al. [48] Thyroidectomy PRO 4 Intraocular pressure
Note absence of outcome for blood loss and blood transfusion
OT operative time (includes fluoroscopy time), LOS length of stay, C complication, Onc oncological (includes lymph node yield), Func
functional (includes erectile function, continence), RYGB Roux-en-Y gastric bypass, AF atrial flutter/fibrillation, RCT randomised controlled
trial, PRO non-randomised prospective comparative study

instruments may be cumbersome to place or switch effi- competences. Furthermore, many surgeons may view sur-
ciently, or may be insufficiently adapted for the specific gical robots as an ‘‘enabling technology’’, without which it
purpose [78, 80, 81, 97]. would not be possible for them to perform certain complex
The surgical learning curve has implications on our minimally invasive procedures [122]. Pure laparoscopic
findings. Before study commencement, individual surgeons radical prostatectomy, which demonstrates significant tech-
have typically performed far fewer robotic cases than con- nical challenges, is an example of a procedure where robotic
ventional ones [51, 53, 54, 72, 73, 107]. This disparity could assistance in suturing and other laparoscopic tasks is
disadvantage robotic surgery due to relatively less famil- important [123]. Although robotic surgery needs to
iarity. This could further explain the prolonged operative demonstrate more than just equivalent patient outcomes to
time of robotic surgery. Nevertheless, our demonstration of be cost-effective due to its substantial costs, its potential
at least equivalent outcomes for other perioperative variables positive effects on surgeon ability must also be considered.
may be regarded as a favourable effect of robotic surgery. This systematic review has some limitations. Our focus
By allowing achievement of similar or better outcomes on blood loss, blood transfusion rate, operative time, length
despite the relative lack of user experience, surgical robots of hospital stay, and complications was based primarily on
may be important in facilitating training and attainment of the fact that these were the most commonly reported

123
Surg Endosc (2016) 30:4330–4352 4347

outcomes in the robotic surgery literature. However, these reporting. Many authors failed to comply with the quality
standard parameters may not fully demonstrate the true criteria [125] for complication reporting. There was also a
value of robotic surgery, especially when the overall ben- lack of agreement in terms of what constitutes complica-
efits are not always clearly perceptible in the short term. tions, such as with regard to blood transfusion and con-
Utilisation of dedicated research parameters should be version. Nevertheless, this issue is not unique to our
encouraged [124]. Already, there is an increasing inclina- included studies [126, 127]. Additionally, studies on
tion towards such parameters that are probably more rele- robotic surgery continue to suffer from several method-
vant, including functional, oncological, and quality of life ological flaws, including a lack of studies that offer mul-
outcomes, specific anatomical–pathological endpoints tiple endpoint analysis [128] in such a complex field.
(such as nerve damage control), and ergonomics. With The Society of American Gastrointestinal and Endo-
continuing improvement in outcome parameter selection scopic Surgeons [122] and European Association of Endo-
by clinical research teams, future evidence synthesis cen- scopic Surgeons [124] consensus statements on robotic
tred on these parameters may better reflect the added value surgery have also highlighted the lack of high-quality data in
of robotic surgery. evaluating the health outcomes of this technology. Upcom-
Our appraisal of robotic surgery through an exclusively ing research efforts should improve on current method-
clinical viewpoint has also meant that other elements of ological deficiencies. The implementation of outcome
innovation evaluation could not be incorporated into our registries for robotic surgery is important to document and
conclusions. These include the impact of surgical robotics compare benefits and harms and in identifying the direction
on intellectual property and patent generation, resource for future development [122]. More robust controlled trials
management, healthcare leadership, mentorship, training, should be undertaken, particularly in areas where robotic
cost efficacy, marketing strategy, business strategy, and surgery has shown some potential, such as complex hepa-
stakeholder value generation. tobiliary surgery, bariatric and upper gastrointestinal revi-
When meta-analyses were possible, the heterogeneity sional surgery, gastric and oesophageal cancer surgery,
was frequently high. However, this is not unexpected given rectal surgery, and surgery for large adrenal masses [124].
the wide variability in patient cohorts and interventions.
There was additional variability within specific procedures.
For instance, Nissen [50, 78, 84, 97–99], Toupet [84], Dor Conclusions
[101], and Thal [15] fundoplications were variant tech-
niques performed in different studies. Furthermore, the After the promising pioneering clinical application of PUMA
extent of robotic assistance varied from its utilisation in 560 in 1985, the stage was set for robotic surgery to assume
anastomotic suturing only [103] to totally robotic proce- the role of a significant disruptive innovation in health care.
dures [21, 22, 82, 83, 85–88, 92, 100]. Methodological Three decades on, our analysis across a wide range of sur-
diversity in the form of different study designs and risks of gical robots identified their overall positive contribution in
bias also contributed to the heterogeneity. reducing blood loss and blood transfusion rate over OS and
We incorporated different surgical robots in our review, MIS. Additionally, against OS, they showed overall pro-
including those that are no longer in use, such as the ZRSS. portional improvement in length of hospital stay and overall
However, our intention was not to compare outcomes of complication rate. These beneficial effects were lost when
specific procedures obtainable through currently available only RCTs were appraised, although these RCTs were
robots but to evaluate, via an overview of commonly themselves limited. Longer operative time was a common
addressed perioperative outcomes, whether the goals of caveat. Further well-conducted surgical trials are needed to
robotic surgery in general have been achieved. Hence, we confirm these findings. Whilst the barriers for these trials
offered a unique perspective on robotic surgery by cover- may seem insurmountable, solutions to overcoming them are
ing the 30 years of its existence. Accordingly, we also now increasingly recognised. These may involve ensuring
elected not to stratify our analysis based on robot or pro- protocol transparency, improving trial dissemination, creat-
cedure type. Consequently, this restricts the applicability of ing specialised trial units, establishing dedicated outcome
this review, so that the individual stakeholder interested in monitoring groups, implementing appropriate minimum
outcomes for a specific intervention may not be able to surgeon experience to reduce the impact of learning curves,
draw sufficiently relevant evidence from our results. and incorporating research training in the surgical curriculum
Prospective studies were included to address the paucity [119]. To ensure better outcomes for future robotic surgery, a
of RCTs. Although practical, their inclusion inevitably multidisciplinary approach during product development
introduces other biases associated with this study design. involving close collaboration between surgeons and engi-
Moreover, caution is advised in the interpretation of neers, in addition to inclusive patient engagement, is
complication data, as there were inconsistencies in their mandatory. With the advent of more affordable, enriching

123
4348 Surg Endosc (2016) 30:4330–4352

technologies can be modularly incorporated into conven- 13. Faul F, Erdfelder E, Lang AG, Buchner A (2007) G*Power 3: a
tional surgical approaches such as intraoperative fluores- flexible statistical power analysis program for the social, behav-
ioral, and biomedical sciences. Behav Res Methods 39:175–191
cence imaging, high-definition 3-D visualisation, wristed 14. Anger JT, Mueller ER, Tarnay C, Smith B, Stroupe K, Rosen-
endoscopic hand tools, and navigation systems, robotic sur- man A, Brubaker L, Bresee C, Kenton K (2014) Robotic com-
gery risks degenerating into an unfulfilled promise if it fails pared with laparoscopic sacrocolpopexy: a randomized
to innovate in line with stakeholders’ needs. controlled trial. Obstet Gynecol 123:5–12
15. Lehnert M, Richter B, Beyer PA, Heller K (2006) A prospective
study comparing operative time in conventional laparoscopic
Compliance with ethical standards
and robotically assisted Thal semifundoplication in children.
J Pediatr Surg 41:1392–1396
Disclosures Alan Tan, Hutan Ashrafian, Alasdair J. Scott, Sam E. 16. Bertani E, Chiappa A, Biffi R, Bianchi PP, Radice D, Branchi V,
Mason, Leanne Harling, Thanos Athanasiou, and Ara Darzi have no Cenderelli E, Vetrano I, Cenciarelli S, Andreoni B (2011)
conflicts of interest or financial ties to disclose. Assessing appropriateness for elective colorectal cancer surgery:
clinical, oncological, and quality-of-life short-term outcomes
Open Access This article is distributed under the terms of the employing different treatment approaches. Int J Colorectal Dis
Creative Commons Attribution 4.0 International License (http://crea 26:1317–1327
tivecommons.org/licenses/by/4.0/), which permits unrestricted use, 17. Khan MS, Challacombe B, Elhage O, Rimington P, Coker B,
distribution, and reproduction in any medium, provided you give Murphy D, Grieve A, Dasgupta P (2012) A dual-centre, cohort
appropriate credit to the original author(s) and the source, provide a comparison of open, laparoscopic and robotic-assisted radical
link to the Creative Commons license, and indicate if changes were cystectomy. Int J Clin Pract 66:656–662
made. 18. Bucerius J, Metz S, Walther T, Falk V, Doll N, Noack F,
Holzhey D, Diegeler A, Mohr FW (2002) Endoscopic internal
thoracic artery dissection leads to significant reduction of pain
References after minimally invasive direct coronary artery bypass graft
surgery. Ann Thorac Surg 73:1180–1184
19. Jung YW, Lee DW, Kim SW, Nam EJ, Kim JH, Kim JW, Kim
1. Kwoh YS, Hou J, Jonckheere EA, Hayati S (1988) A robot with
YT (2010) Robot-assisted staging using three robotic arms for
improved absolute positioning accuracy for CT guided stereo-
endometrial cancer: comparison to laparoscopy and laparotomy
tactic brain surgery. IEEE Trans Biomed Eng 35:153–161
at a single institution. J Surg Oncol 101:116–121
2. Cooper MA, Ibrahim A, Lyu H, Makary MA (2013) Underre-
20. Hoekstra AV, Jairam-Thodla A, Rademaker A, Singh DK, Buttin
porting of robotic surgery complications. J Healthc Qual. doi:10.
BM, Lurain JR, Schink JC, Lowe MP (2009) The impact of
1111/jhq.12036
robotics on practice management of endometrial cancer: transi-
3. Investor Presentation Q2 (2015) http://investor.intuitivesurgical.
tioning from traditional surgery. Int J Med Robot 5:392–397
com/phoenix.zhtml?c=122359&p=irol-irhome. Accessed 29
21. Malcolme-Lawes LC, Lim PB, Koa-Wing M, Whinnett ZI, Jamil-
May 2015
Copley S, Hayat S, Francis DP, Kojodjojo P, Davies DW, Peters
4. Yu HY, Hevelone ND, Lipsitz SR, Kowalczyk KJ, Hu JC (2012)
NS, Kanagaratnam P (2013) Robotic assistance and general
Use, costs and comparative effectiveness of robotic assisted,
anaesthesia improve catheter stability and increase signal atten-
laparoscopic and open urological surgery. J Urol 187:1392–1398
uation during atrial fibrillation ablation. Europace 15:41–47
5. Barbash GI, Glied SA (2010) New technology and health care
22. Tilz RR, Chun KR, Metzner A, Burchard A, Wissner E,
costs—the case of robot-assisted surgery. N Engl J Med
Koektuerk B, Konstantinidou M, Nuyens D, De Potter T, Neven
363:701–704
K, Furnkranz A, Ouyang F, Schmidt B (2010) Unexpected high
6. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred
incidence of esophageal injury following pulmonary vein iso-
reporting items for systematic reviews and meta-analyses: the
lation using robotic navigation. J Cardiovasc Electrophysiol
PRISMA statement. BMJ 339:b2535
21:853–858
7. Higgins J, Green S (2011) Cochrane handbook for systematic
23. Parekattil SJ, Gudeloglu A, Brahmbhatt J, Wharton J, Priola KB
reviews of interventions Version 5.1.0 [updated March 2011].
(2012) Robotic assisted versus pure microsurgical vasectomy
The Cochrane Collaboration. http://www.cochrane-handbook.org
reversal: technique and prospective database control trial. J Re-
8. Stang A (2010) Critical evaluation of the Newcastle–Ottawa
constr Microsurg 28:435–444
scale for the assessment of the quality of nonrandomized studies
24. McCulloch P, Taylor I, Sasako M, Lovett B, Griffin D (2002)
in meta-analyses. Eur J Epidemiol 25:603–605
Randomised trials in surgery: problems and possible solutions.
9. Guyatt GH, Oxman AD, Vist G, Kunz R, Brozek J, Alonso-
BMJ 324:1448–1451
Coello Montori V, Akl EA, Djulbegovic B, Falck-Ytter Y,
25. Nix J, Smith A, Kurpad R, Nielsen ME, Wallen EM, Pruthi RS
Norris SL, Williams JA Jr, Atkins D, Meerpohl J, Schunemann
(2010) Prospective randomized controlled trial of robotic versus
HJ (2011) GRADE guidelines: 4. Rating the quality of evi-
open radical cystectomy for bladder cancer: perioperative and
dence—study limitations (risk of bias). J Clin Epidemiol
pathologic results. Eur Urol 57:196–201
64:407–415
26. Parekh DJ, Messer J, Fitzgerald J, Ercole B, Svatek R (2013)
10. Friedrich JO, Adhikari NK, Beyene J (2008) The ratio of means
Perioperative outcomes and oncologic efficacy from a pilot
method as an alternative to mean differences for analyzing
prospective randomized clinical trial of open versus robotic
continuous outcome variables in meta-analysis: a simulation
assisted radical cystectomy. J Urol 189:474–479
study. BMC Med Res Methodol 8:32
27. Nishihara S, Sugano N, Nishii T, Miki H, Nakamura N, Yosh-
11. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials.
ikawa H (2006) Comparison between hand rasping and robotic
Control Clin Trials 7:177–188
milling for stem implantation in cementless total hip arthro-
12. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Mea-
plasty. J Arthroplasty 21:957–966
suring inconsistency in meta-analyses. BMJ 327:557–560

123
Surg Endosc (2016) 30:4330–4352 4349

28. Bargar WL, Bauer A, Borner M (1998) Primary and revision versus traditional off-pump coronary artery bypass grafting
total hip replacement using the Robodoc system. Clin Orthop techniques. Ann Surg 248:638–646
Relat Res 354:82–91 44. Collins SA, Tulikangas PK, O’Sullivan DM (2012) Effect of
29. Song EK, Seon JK, Park SJ, Jung WB, Park HW, Lee GW surgical approach on physical activity and pain control after
(2011) Simultaneous bilateral total knee arthroplasty with sacral colpopexy. Am J Obstet Gynecol 206:438.e1-6
robotic and conventional techniques: a prospective, randomized 45. Gocmen A, Sanlikan F, Ucar MG (2010) Comparison of robotic-
study. Knee Surg Sports Traumatol Arthrosc 19:1069–1076 assisted surgery outcomes with laparotomy for endometrial
30. Song EK, Seon JK, Yim JH, Netravali NA, Bargar WL (2013) cancer staging in Turkey. Arch Gynecol Obstet 282:539–545
Robotic-assisted TKA reduces postoperative alignment outliers 46. Lowe MP, Hoekstra AV, Jairam-Thodla A, Singh DK, Buttin
and improves gap balance compared to conventional TKA. Clin BM, Lurain JR, Schink JC (2009) A comparison of robot-as-
Orthop Relat Res 471:118–126 sisted and traditional radical hysterectomy for early-stage cer-
31. Farnham SB, Webster TM, Herrell SD, Smith JA Jr (2006) vical cancer. J Robot Surg 3:19–23
Intraoperative blood loss and transfusion requirements for 47. Lee J, Nah KY, Kim RM, Ahn YH, Soh EY, Chung WY (2010)
robotic-assisted radical prostatectomy versus radical retropubic Differences in postoperative outcomes, function, and cosmesis:
prostatectomy. Urology 67:360–363 open versus robotic thyroidectomy. Surg Endosc 24:3186–3194
32. Ficarra V, Novara G, Fracalanza S, D’Elia C, Secco S, Iafrate 48. Kim JA, Kim JS, Chang MS, Yoo YK, Dim DK (2013) Influ-
M, Cavalleri S, Artibani W (2009) A prospective, non-ran- ence of carbon dioxide insufflation of the neck on intraocular
domized trial comparing robot-assisted laparoscopic and pressure during robot-assisted endoscopic thyroidectomy: a
retropubic radical prostatectomy in one European institution. comparison with open thyroidectomy. Surg Endosc 27:1587–
BJU Int 104:534–539 1593
33. Kordan Y, Barocas DA, Altamar HO, Clark PE, Chang SS, 49. Masson-Lecomte A, Yates DR, Hupertan V, Haertig A, Char-
Davis R, Herrell SD, Baumgartner R, Mishra V, Chan RC, tier-Kastler E, Bitker MO, Vaessen C, Roupret M (2013) A
Smith JA Jr, Cookson MS (2010) Comparison of transfusion prospective comparison of the pathologic and surgical outcomes
requirements between open and robotic-assisted laparoscopic obtained after elective treatment of renal cell carcinoma by open
radical prostatectomy. BJU Int 106:1036–1040 or robot-assisted partial nephrectomy. Urol Oncol 31:924–929
34. Menon M, Tewari A, Baize B, Guillonneau B, Vallancien G 50. Draaisma WA, Ruurda JP, Scheffer RC, Simmermacher RK,
(2002) Prospective comparison of radical retropubic prostatec- Gooszen HG, Rijnhart-de Jong HG, Busken E, Broeders IA
tomy and robot-assisted anatomic prostatectomy: the Vattikuti (2006) Randomized clinical trial of standard laparoscopic versus
Urology Institute experience. Urology 60:864–868 robot-assisted laparoscopic Nissen fundoplication for gastro-
35. Wood DP, Schulte R, Dunn RL, Hollenbeck BK, Saur R, Wolf oesophageal reflux disease. Br J Surg 93:1351–1359
JS Jr, Montie JE (2007) Short-term health outcome differences 51. Park JS, Choi GS, Park SY, Kim JH, Ryuk JP (2012) Ran-
between robotic and conventional radical prostatectomy. Urol- domized clinical trial of robot-assisted versus standard laparo-
ogy 70:945–949 scopic right colectomy. Br J Surg 99:1219–1226
36. Tewari A, Srivasatava A, Menon M (2003) A prospective 52. Zhou HX, Guo YH, Yu XF, Bao SY, Liu JL, Zhang Y, Ren YG
comparison of radical retropubic and robot-assisted prostatec- (2006) Zeus robot-assisted laparoscopic cholecystectomy in
tomy: experience in one institution. BJU Int 92:205–210 comparison with conventional laparoscopic cholecystectomy.
37. Ham WS, Park SY, Kim WT, Koo KC, Lee YS, Choi YD (2008) Hepatobiliary Pancreat Dis Int 5:115–118
Open versus robotic radical prostatectomy: a prospective anal- 53. Porpiglia F, Morra I, Lucci Chiarissi M, Manfredi M, Mele F,
ysis based on a single surgeon’s experience. J Robot Surg Grande S, Ragni F, Poggio M, Fiori C (2013) Randomised
2:235–241 controlled trial comparing laparoscopic and robot-assisted rad-
38. Hong JY, Kim JY, Choi YD, Rha KH, Yoon SJ, Kil HK (2010) ical prostatectomy. Eur Urol 63:606–614
Incidence of venous gas embolism during robotic-assisted 54. Sarlos D, Kots L, Stevanovic N, von Felten S, Schar G (2012)
laparoscopic radical prostatectomy is lower than that during Robotic compared with conventional laparoscopic hysterec-
radical retropubic prostatectomy. Br J Anaesth 105:777–781 tomy: a randomized controlled trial. Obstet Gynecol 120:604–
39. Ludovico GM, Dachille G, Pagliarulo G, D’Elia C, Mondaini N, 611
Gacci M, Detti B, Malossini G, Bartoletti R, Cai T (2013) 55. Patriti A, Ceccarelli G, Bartoli A, Spaziani A, Biancafarina A,
Bilateral nerve sparing robotic-assisted radical prostatectomy is Casciola L (2009) Short- and medium-term outcome of robot-
associated with faster continence recovery but not with erectile assisted and traditional laparoscopic rectal resection. JSLS
function recovery compared with retropubic open prostatec- 13:176–183
tomy: the need for accurate selection of patients. Oncol Rep 56. Pigazzi A, Ellenhorn JD, Ballantyne GH, Paz IB (2006)
29:2445–2450 Robotic-assisted laparoscopic low anterior resection with total
40. Martin AD, Nunez RN, Castle EP (2011) Robot-assisted radical mesorectal excision for rectal cancer. Surg Endosc 20:1521–
cystectomy versus open radical cystectomy: a complete cost 1525
analysis. Urology 77:621–625 57. Wong MT, Meurette G, Rigaud J, Regenet N, Lehur PA (2011)
41. Ng CK, Kauffman EC, Lee MM, Otto BJ, Portnoff A, Ehrlich Robotic versus laparoscopic rectopexy for complex rectocele: a
JR, Schwartz MJ, Wang GJ, Scherr DS (2010) A comparison of prospective comparison of short-term outcomes. Dis Colon
postoperative complications in open versus robotic cystectomy. Rectum 54:342–346
Eur Urol 57:274–281 58. Berber E, Akyildiz HY, Aucejo F, Gunasekaran G, Chalikonda
42. Rhee JJ, Lebeau S, Smolkin M, Theodorescu D (2006) Radical S, Fung J (2010) Robotic versus laparoscopic resection of liver
cystectomy with ileal conduit diversion: early prospective tumours. HPB (Oxford) 12:583–586
evaluation of the impact of robotic assistance. BJU Int 59. Ploussard G, Xylinas E, Paul A, Gillion N, Salomon L, Allory
98:1059–1063 Y, Vordos D, Hoznek A, Yiou R, Abbou CC, de la Taille A
43. Poston RS, Tran R, Collins M, Reynolds M, Connerney I, (2009) Is robot assistance affecting operating room time com-
Reicher B, Zimrin D, Griffith BP, Bartlett ST (2008) Compar- pared with pure retroperitoneal laparoscopic radical prostatec-
ison of economic and patient outcomes with minimally invasive tomy? J Endourol 23:939–943

123
4350 Surg Endosc (2016) 30:4330–4352

60. Hemal AK, Kumar A (2009) A prospective comparison of 76. Honl M, Dierk O, Gauck C, Carrero V, Lampe F, Dries S,
laparoscopic and robotic radical nephrectomy for T1-2N0M0 Quante M, Schwieger K, Hille E, Morlock MM (2003) Com-
renal cell carcinoma. World J Urol 27:89–94 parison of robotic-assisted and manual implantation of a primary
61. Masson-Lecomte A, Bensalah K, Seringe E, Vaessen C, de la total hip replacement. A prospective study. J Bone Joint Surg
Taille A, Doumerc N, Rischmann P, Bruyere F, Soustelle L, Am 85-A:1470–1478
Droupy S, Roupret M (2013) A prospective comparison of 77. Ringel F, Stuer C, Reinke A, Preuss A, Behr M, Auer F, Stoffel
surgical and pathological outcomes obtained after robot-assisted M, Meyer B (2012) Accuracy of robot-assisted placement of
or pure laparoscopic partial nephrectomy in moderate to com- lumbar and sacral pedicle screws: a prospective randomized
plex renal tumours: results from a French multicentre collabo- comparison to conventional freehand screw implantation. Spine
rative study. BJU Int 111:256–263 (Phila Pa 1976) 37:E496–E501
62. Kural AR, Atug F, Tufek I, Akpinar H (2009) Robot-assisted 78. Morino M, Pellegrino L, Giaccone C, Garrone C, Rebecchi F
partial nephrectomy versus laparoscopic partial nephrectomy: (2006) Randomized clinical trial of robot-assisted versus
comparison of outcomes. J Endourol 23:1491–1497 laparoscopic Nissen fundoplication. Br J Surg 93:553–558
63. Caruso RP, Phillips CK, Kau E, Taneja SS, Stifelman MD 79. Ruurda JP, Visser PL, Broeders IA (2003) Analysis of procedure
(2006) Robot assisted laparoscopic partial nephrectomy: initial time in robot-assisted surgery: comparative study in laparo-
experience. J Urol 176:36–39 scopic cholecystectomy. Comput Aided Surg 8:24–29
64. Seror J, Yates DR, Seringe E, Vaessen C, Bitker MO, Chartier- 80. Paraiso MF, Jelovsek JE, Frick A, Chen CC, Barber MD (2011)
Kastler E, Roupret M (2012) Prospective comparison of short-term Laparoscopic compared with robotic sacrocolpopexy for vaginal
functional outcomes obtained after pure laparoscopic and robot- prolapse: a randomized controlled trial. Obstet Gynecol
assisted laparoscopic sacrocolpopexy. World J Urol 30:393–398 118:1005–1013
65. El Hachem L, Acholonu UC Jr, Nezhat FR (2013) Postoperative 81. Morino M, Beninca G, Giraudo G, Del Genio GM, Rebecchi F,
pain and recovery after conventional laparoscopy compared with Garrone C (2004) Robot-assisted vs laparoscopic adrenalec-
robotically assisted laparoscopy. Obstet Gynecol 121:547–553 tomy: a prospective randomized controlled trial. Surg Endosc
66. Wu JC, Wu HS, Lin MS, Chou DA, Huang MH (2008) Com- 18:1742–1746
parison of robot-assisted laparoscopic adrenalectomy with tra- 82. Steven D, Servatius H, Rostock T, Hoffmann B, Drewitz I,
ditional laparoscopic adrenalectomy—1 year follow-up. Surg Mullerleile K, Sultan A, Aydin MA, Meinertz T, Willems S (2010)
Endosc 22:463–466 Reduced fluoroscopy during atrial fibrillation ablation: benefits of
67. Gosseine PN, Mangin P, Leclers F, Cormier L (2009) Pure robotic guided navigation. J Cardiovasc Electrophysiol 21:6–12
laparoscopic versus robotic-assisted laparoscopic radical 83. Steven D, Rostock T, Servatius H, Hoffmann B, Drewitz
prostatectomy: comparative study to assess functional urinary Mullerleile K, Meinertz T, Willems S (2008) Robotic versus
outcomes. Prog Urol 19:611–617 conventional ablation for common-type atrial flutter: a
68. Kim JC, Yang SS, Jang TY, Kwak JY, Yun MJ, Lim SB (2012) prospective randomized trial to evaluate the effectiveness of
Open versus robot-assisted sphincter-saving operations in rectal remote catheter navigation. Heart Rhythm 5:1556–1560
cancer patients: techniques and comparison of outcomes 84. Melvin WS, Needleman BJ, Krause KR, Scheneider C, Ellison
between groups of 100 matched patients. Int J Med Robot EC (2002) Computer-enhanced vs. standard laparoscopic antire-
8:468–475 flux surgery. J Gastrointest Surg 6:11–15 (discussion 15–16)
69. Doumerc N, Yuen C, Savdie R, Rahman MB, Rasiah KK, Pe 85. Benizri EI, Renaud M, Reibel N, Germain A, Ziegler O,
Benito R, Delprado W, Matthews J, Haynes AM, Stricker PD Zarnegar R, Ayav A, Bresler L, Brunaud L (2013) Perioperative
(2010) Should experienced open prostatic surgeons convert to outcomes after totally robotic gastric bypass: a prospective
robotic surgery? The real learning curve for one surgeon over 3 nonrandomized controlled study. Am J Surg 206:145–151
years. BJU Int 106:378–384 86. Di Biase L, Wang Y, Horton R, Gallinghouse GJ, Mohanty P,
70. Di Pierro GB, Baumeister P, Stucki P, Beatrice J, Danuser H, Sanchez J, Patel D, Dare M, Canby R, Price LD, Zagrodzky JD,
Mattei A (2011) A prospective trial comparing consecutive series Bailey S, Burkhardt JD, Natale A (2009) Ablation of atrial
of open retropubic and robot-assisted laparoscopic radical fibrillation utilizing robotic catheter navigation in comparison to
prostatectomy in a centre with a limited caseload. Eur Urol 59:1–6 manual navigation and ablation: single-center experience.
71. Bachinsky WB, Abdelsalam M, Boga G, Kiljanek L, Mumtaz J Cardiovasc Electrophysiol 20:1328–1835
M, McCarty C (2012) Comparative study of same sitting hybrid 87. Kautzner J, Peichl P, Cihak R, Wichterle D, Mlcochova H
coronary artery revascularization versus off-pump coronary (2009) Early experience with robotic navigation for catheter
artery bypass in multivessel coronary artery disease. J Interv ablation of paroxysmal atrial fibrillation. Pacing Clin Electro-
Cardiol 25:460–468 physiol 32(Suppl 1):S163–S166
72. Asimakopoulos AD, Pereira Fraga CT, Annino F, Pasqualetti P, 88. Rillig A, Meyerfeldt U, Tilz RR, Talazko J, Arya A, Zvereva V,
Calado AA, Mugnier C (2011) Randomized comparison Birkemeyer R, Miljak T, Hajredini B, Wohlmuth P, Fink U,
between laparoscopic and robot-assisted nerve-sparing radical Jung W (2012) Incidence and long-term follow-up of silent
prostatectomy. J Sex Med 8:1503–1512 cerebral lesions after pulmonary vein isolation using a remote
73. Paraiso MF, Ridgeway B, Park AJ, Jelovsek JE, Barber MD, robotic navigation system as compared with manual ablation.
Falcone T, Einarsson JI (2013) A randomized trial comparing Circ Arrhythm Electrophysiol 5:15–21
conventional and robotically assisted total laparoscopic hys- 89. Baik SH, Kwon HY, Kim JS, Hur H, Sohn SK, Cho CH, Kim H
terectomy. Am J Obstet Gynecol 208(368):e1–e7 (2009) Robotic versus laparoscopic low anterior resection of
74. Jimenez Rodriguez RM, Diaz Pavon JM, de La Portilla de Juan rectal cancer: short-term outcome of a prospective comparative
F, Prendes Sillero E, Hisnard Cadet Dussort JM, Padillo J (2011) study. Ann Surg Oncol 16:1480–1487
Prospective randomised study: robotic-assisted versus conven- 90. Kornprat P, Werkgartner G, Cerwenka H, Bacher H, El-Shab-
tional laparoscopic surgery in colorectal cancer resection. Cir rawi A, Rehak P, Mischinger HJ (2006) Prospective study
Esp 89:432–438 comparing standard and robotically assisted laparoscopic
75. Kim YW, Lee HM, Kim NK, Min BS, Lee KY (2012) The cholecystectomy. Langenbecks Arch Surg 391:216–221
learning curve for robot-assisted total mesorectal excision for 91. Park JY, Jo MJ, Nam BH, Kim Y, Eom BW, Yoon HM, Ryu
rectal cancer. Surg Laparosc Endosc Percutan Tech 22:400–405 KW, Kim YW, Lee JH (2012) Surgical stress after robot-

123
Surg Endosc (2016) 30:4330–4352 4351

assisted distal gastrectomy and its economic implications. Br J of cementless total hip arthroplasty. Clin Orthop Relat Res
Surg 99:1554–1561 468:1072–1081
92. Mierdl S, Byhahn C, Lischke V, Aybek T, Wimmer-Greinecker 108. Park SE, Lee CT (2007) Comparison of robotic-assisted and
G, Dogan S, Viehmeyer S, Kessler P, Westphal K (2005) Seg- conventional manual implantation of a primary total knee
mental myocardial wall motion during minimally invasive arthroplasty. J Arthroplasty 22:1054–1059
coronary artery bypass grafting using open and endoscopic 109. Bach CM, Winter P, Nogler M, Gobel G, Wimmer C, Ogon M
surgical techniques. Anesth Analg 100:306–314 (2002) No functional impairment after Robodoc total hip arthro-
93. Brunaud L, Bresler L, Ayav A, Tretou S, Cormier L, Klein M, plasty: gait analysis in 25 patients. Acta Orthop Scand 73:386–391
Boissel P (2003) Advantages of using robotic Da Vinci system for 110. Balduyck B, Hendriks JM, Lauwers P, Mercelis R, Ten Broecke
unilateral adrenalectomy: early results. Ann Chir 128:530–535 P, Van Schil P (2011) Quality of life after anterior mediastinal
94. Kiaii B, McClure RS, Stitt L, Rayman R, Dobkowski WB, mass resection: a prospective study comparing open with
Jablonsky G, Novick RJ, Boyd WD (2006) Prospective angio- robotic-assisted thoracoscopic resection. Eur J Cardiothorac
graphic comparison of direct, endoscopic, and telesurgical Surg 39:543–548
approaches to harvesting the internal thoracic artery. Ann Tho- 111. Carlsson S, Nilsson AE, Schumacher MC, Jonsson MN, Volz
rac Surg 82:624–628 DS, Steineck G, Wiklund PN (2010) Surgery-related compli-
95. Nelson B, Kaufman M, Broughton G, Cookson MS, Chang SS, cations in 1253 robot-assisted and 485 open retropubic radical
Herrell SD, Baumgartner RG, Smith JA Jr (2007) Comparison prostatectomies at the Karolinska University Hospital, Sweden.
of length of hospital stay between radical retropubic prostatec- Urology 75:1092–1097
tomy and robotic assisted laparoscopic prostatectomy. J Urol 112. Kim SC, Song C, Kim W, Kang T, Park J, Jeong IG, Lee S, Cho
177:929–931 YM, Ahn H (2011) Factors determining functional outcomes
96. Ryu HR, Lee J, Park JH, Kang SW, Jeong JJ, Hong JY, Chung after radical prostatectomy: robot-assisted versus retropubic. Eur
WY (2013) A comparison of postoperative pain after conventional Urol 60:413–419
open thyroidectomy and transaxillary single-incision robotic thy- 113. Siebel T, Kafer W (2005) Clinical outcome following robotic
roidectomy: a prospective study. Ann Surg Oncol 20:2279–2284 assisted versus conventional total hip arthroplasty: a controlled
97. Cadiere GB, Himpens J, Vertruyen M, Bruyns J, Germay O, and prospective study of seventy-one patients. Z Orthop Ihre
Leman G, Izizaw R (2001) Evaluation of telesurgical (robotic) Grenzgeb 143:391–398
NISSEN fundoplication. Surg Endosc 15:918–923 114. Christensen CM (1997) The innovator’s dilemma: when new
98. Muller-Stich BP, Reiter MA, Wente MN, Bintintan VV, Konin- technologies cause great firms to fail. Harvard Business School
ger J, Buchler MW, Gutt CN (2007) Robot-assisted versus con- Press, Boston
ventional laparoscopic fundoplication: short-term outcome of a 115. Cheng Q, Pang TC, Hollands MJ, Richardson AJ, Pleass H,
pilot randomized controlled trial. Surg Endosc 21:1800–1805 Johnston ES, Lam VW (2014) Systematic review and meta-
99. Nakadi IE, Melot C, Closset J, DeMoor V, Betroune K, Feron P, analysis of laparoscopic versus open distal gastrectomy. J Gas-
Lingier P, Gelin M (2006) Evaluation of da Vinci Nissen fun- trointest Surg 18:1087–1099
doplication clinical results and cost minimization. World J Surg 116. Rondelli F, Trastulli S, Avenia N, Schillaci G, Cirocchi R, Gulla N,
30:1050–1054 Mariani E, Bistoni G, Noya G (2012) Is laparoscopic right colec-
100. Sanchez BR, Mohr CJ, Morton JM, Safadi BY, Alami RS, Curet tomy more effective than open resection? A meta-analysis of ran-
MJ (2005) Comparison of totally robotic laparoscopic Roux-en- domized and nonrandomized studies. Colorectal Dis 14:e447–e469
Y gastric bypass and traditional laparoscopic Roux-en-Y gastric 117. Tang K, Li H, Xia D, Hu Z, Zhuang Q, Liu J, Xu H, Ye Z (2014)
bypass. Surg Obes Relat Dis 1:549–554 Laparoscopic versus open radical cystectomy in bladder cancer:
101. Hartmann J, Menenakos C, Ordemann J, Nocon M, Raue W, a systematic review and meta-analysis of comparative studies.
Braumann C (2009) Long-term results of quality of life after PLoS ONE 9:e95667
standard laparoscopic vs. robot-assisted laparoscopic fundopli- 118. Vennix S, Pelzers L, Bouvy N, Beets GL, Pierie JP, Wiggers T,
cations for gastro-oesophageal reflux disease. A comparative Breukink S (2014) Laparoscopic versus open total mesorectal
clinical trial. Int J Med Robot 5:32–37 excision for rectal cancer. Cochrane Database Syst Rev 4:Cd005200
102. Heemskerk J, de Hoog DE, van Gemert WG, Baeten CG, Greve 119. Garas G, Ibrahim A, Ashrafian H, Ahmed K, Patel V, Oka-
JW, Bouvy ND (2007) Robot-assisted vs. conventional laparo- bayashi K, Skapinakis P, Darzi A, Athanasiou T (2012) Evi-
scopic rectopexy for rectal prolapse: a comparative study on dence-based surgery: barriers, solutions, and the role of evidence
costs and time. Dis Colon Rectum 50:1825–1830 synthesis. World J Surg 36:1723–1731
103. Kolvenbach R, Schwierz E, Wasilljew S, Miloud A, Puerschel 120. Lilford R, Braunholtz D, Harris J, Gill T (2004) Trials in sur-
A, Pinter L (2004) Total laparoscopically and robotically gery. Br J Surg 91:6–16
assisted aortic aneurysm surgery: a critical evaluation. J Vasc 121. Stirrat GM, Farrow SC, Farndon J, Dwyer N (1992) The chal-
Surg 39:771–776 lenge of evaluating surgical procedures. Ann R Coll Surg Engl
104. Muhlmann G, Klaus A, Kirchmayr W, Wykypiel H, Unger A, 74:80–84
Holler E, Nehoda H, Aigner F, Weiss HG (2003) DaVinci 122. Herron DM, Marohn M (2008) A consensus document on
robotic-assisted laparoscopic bariatric surgery: is it justified in a robotic surgery. Surg Endosc 22:313–325 (discussion 311–312)
routine setting? Obes Surg 13:848–854 123. Harrell AG, Heniford BT (2005) Minimally invasive abdominal
105. Nio D, Bemelman WA, Busch OR, Vrouenraets BC, Gouma DJ surgery: lux et veritas past, present, and future. Am J Surg
(2004) Robot-assisted laparoscopic cholecystectomy versus 190:239–243
conventional laparoscopic cholecystectomy: a comparative 124. Szold A, Bergamaschi R, Broeders I, Dankelman J, Forgione A,
study. Surg Endosc 18:379–382 Lango T, Melzer A, Mintz Y, Morales-Conde S, Rhodes M,
106. Cobb J, Henckel J, Gomes P, Harris S, Jakopec M, Rodriguez F, Satava R, Tang CN, Vilallonga R (2015) European Association
Barrett A, Davies B (2006) Hands-on robotic unicompartmental of Endoscopic Surgeons (EAES) consensus statement on the use
knee replacement: a prospective, randomised controlled study of of robotics in general surgery. Surg Endosc 29:253–288
the Acrobot system. J Bone Joint Surg Br 88:188–197 125. Martin RC 2nd, Brennan MF, Jaques DP (2002) Quality of
107. Nakamura N, Sugano N, Nishii T, Kakimoto A, Miki H (2010) complication reporting in the surgical literature. Ann Surg
A comparison between robotic-assisted and manual implantation 235:803–813

123
4352 Surg Endosc (2016) 30:4330–4352

126. Haidich AB, Birtsou C, Dardavessis T, Tirodimos I, Arvanitidou 128. Almeida GM, Duarte TL, Farmer PB, Steward WP, Jones GD
M (2011) The quality of safety reporting in trials is still sub- (2008) Multiple end-point analysis reveals cisplatin damage
optimal: survey of major general medical journals. J Clin Epi- tolerance to be a chemoresistance mechanism in a NSCLC
demiol 64:124–135 model: implications for predictive testing. Int J Cancer
127. Ioannidis JP, Lau J (2001) Completeness of safety reporting in 122:1810–1819
randomized trials: an evaluation of 7 medical areas. JAMA
285:437–443

123