Excitement about robotic surgery continues to grow with the obvious benefits in surgeon ergonomics, high-definition 3D vision, dexterity, truly objective metrics for assessment, the application of artificial intelligence and augmented reality. Minimally invasive surgery (MIS) has evidence of non-inferiority in mortality outcomes compared to open surgery, but with superior outcomes in terms of patient morbidity and length of stay [1]. For robotic surgery, evidence is mounting as institutional learning curves are realised and long-term data becomes available, with improved patient outcomes when compared to laparoscopy in terms of morbidity [2,3,4,5,6] including lower blood loss, conversion-to-open, pain and shorter hospital stay. However, the evidence is mixed, for example in a meta-analysis comparing different approaches of total mesorectal excision in rectal cancer [7], and often equivalent outcomes in “smaller” operations [8, 9].

Roughly 82% of robotic surgery performed is within urology, general surgery, and gynaecology [10], but it is still only used in a minority of operations worldwide, due to availability and cost. To address this, many robotic platforms are in development or have recently reached the market, providing competition but also different approaches to broaden the capacity and capabilities of surgeons. As such there has been a rapid increase in robotic operations, one study of 73 hospitals stated an increase of 1.8% to 15.1% of all general surgery procedures were performed robotically between 2012 to 2018 [11] and the robotic surgery market globally was valued at $5.32 billion in 2019, estimated to grow to $19 billion in 2027 [12].

Along with this expansion there have been calls for reporting on safe implementation of novel platforms and standardisation of training and accreditation within robotics, due to concerns over errors and patient safety [13, 14]. Now the surgical community is faced with the additional challenge of evaluating multiple robotic platforms.

To our knowledge, there is no comprehensive, up-to-date review of existing platforms which can help guide the end-user, the surgeon, to decide which robot would be ideal for their purpose and the evidence to support it.

This scoping review aims to provide an update of current and emerging robotic platforms within minimally invasive surgical specialties, including a stage of evaluation using The Idea, Development, Explore, Assessment and Long-term study (IDEAL) Framework [15].

Methods

A scoping review was performed, screening articles from PubMed, Google Scholar, journal reports, company websites and review articles. The search focused on minimally invasive robotic surgical platforms used within general surgery, gynaecology, urology, head and neck, cardiothoracics, given the application is predominantly in these specialties, and robots are broadly comparable in terms of function. Systems operating purely outside the abdomen or thorax and endoluminal or natural orifice platforms were excluded as these are potentially not comparable.

Information from the public domain was also collected and each company approached via email for virtual interview to discover more about the systems and quality check data collection. This was a structured hour-long interview with a template of questions used (Fig. 1), and the company was given an opportunity for a short presentation. Clinical data to aid the IDEAL stage of evaluation was identified through PubMed, Google and company websites. Data collection included: company, founding year, development and testing including pre-clinical/clinical trials, price, system and device descriptors, training and support available, and additional information distinguishing robots from their competitors. The IDEAL Framework (Fig. 2) was applied to assess the stage of evaluation for each system in the clinical setting. All companies who responded reviewed their respective data in this review as part of the quality assurance process. Results are accurate to the authors’ knowledge at the time of publication, however, may have changed or inaccuracies present, particularly in companies who have not responded.

Fig. 1
figure 1

Virtual Interview/Data recording template

Full size image
Fig. 2
figure 2

IDEAL Framework stages modified from the website

Full size image

Results

A total of 36 robotic system platforms were identified for potential data extraction. Of these, 15 were excluded, 10 of which were endoluminal or natural orifice robotic systems and the other 5 were robotic devices rather than surgical systems. A total of 21 robotic platforms were scrutinised for data extraction and analysis, presented in full in Table 1, with accompanying images in Figs. 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22.

Table 1 Robotic systems description and evaluation
Full size table
Fig. 3
figure 3

da Vinci 4th generation robots [101]

Full size image
Fig. 4
figure 4

KANDUO Robot® Surgical System provided by and permission from Suzhuo Kangduo Robot Co., Ltd

Full size image
Fig. 5
figure 5

The Senhance® Surgical System, Asensus Surgical. Provided by and with permission from the company

Full size image
Fig. 6
figure 6

Hinotori Surgical Robot, Medicaroid Corporation [102]

Full size image
Fig. 7
figure 7

Hugo™ RAS system, Medtonic. Photos from media kit and with permission

Full size image
Fig. 8
figure 8

WEGO MicroHand S Surgical Robot System [103]

Full size image
Fig. 9
figure 9

Revo-I system, Meerecompany Inc. Downloaded from the website media kit

Full size image
Fig. 10
figure 10

Toumai Laparoscopic Surgical Robot, MEDBOT [104]

Full size image
Fig. 11
figure 11

Versius Surgical Robotic System, provided by and permission from CMT (left picture- surgeon console and modular bed side units, right picture- hand controller) [105]

Full size image
Fig. 12
figure 12

MP1000 and SP1000 robots, Shenzen Edge Medical Robotics Co [76]

Full size image
Fig. 13
figure 13

Dexter, Distalmotion. Provided by and permission from the company

Full size image
Fig. 14
figure 14

Avatera® System, avateramedical GmbH, provided by and permission from company

Full size image
Fig. 15
figure 15

Mantra Surgical Robotic System, M/S. Sudhir, Srivastava Innovations (SSI) PVT. Ltd. Photos from product brochure with permission

Full size image
Fig. 16
figure 16

The MIRA Surgical System, Virtual Incision Corp. Images provided from company with permission and modified for the publication to fit size

Full size image
Fig. 17
figure 17

Shurui Robot, Beijing Shurui Technology Co., Ltd [106]

Full size image
Fig. 18
figure 18

Bitrack System, RobSurgical [93]

Full size image
Fig. 19
figure 19

Carina, Ronovo Surgical [94]

Full size image
Fig. 20
figure 20

Enos™, Titan Medical Inc. Photo from website and permission from the company

Full size image
Fig. 21
figure 21

MiroSurge, DLR/Alexandra Beier (CC BY-NC-ND 3.0). Photo and permission provided by company

Full size image
Fig. 22
figure 22

Vicarious surgical system available online via media kit

Full size image

Twenty companies were approached via email for virtual interview. Twelve replied, seven met virtually and five via email. Eight companies did not reply and one there was no available email, therefore, the corresponding author of a review paper was approached, again, with no reply. Instead, for these companies, information was collected solely from the public domain.

Of these companies, China and USA have six different platforms each, Germany with two and the UK, Canada, Spain, Republic of Korea, Japan, Switzerland and India all with one platform. There was a total of 23 surgical robots for analysis. Fifteen of 21 companies represent multiport robots, four single ports, and two with both. Of the multiport systems, eight were modular and nine had a single-unit patient console.

Thirteen robotic companies have national or international regulatory approvals, with eight having none, although one robot, DLR MiroSurge from Germany, will never be used in the clinical setting.

All supporting evidence on robotic systems have been reported only in urology, general surgery and gynaecology.

Of the 17 multiport systems, one is fully evaluated at stage 4, two are stage 3, six stage 2b, two stage 2a, two stage 1, and four at the pre-IDEAL stage 0. Of the six single-port systems none have been fully evaluated with one at stage 3, three at stage 1 and two at stage 0. Pooling the 23 systems together; six have been evaluated at stage 0, and 13 at stage 1 to 2b, with four at stage 3 to 4.

Long-term data was reported for three companies. Intuitive Surgical Inc, and Suzhou Kangduo Robot Co., Ltd robotic systems both had randomised control trial data supporting evidence, whilst Asensus Surgical Ltd., have formed a multispecialty registry.

Comparison data is reported for robotic devices and platforms (Table 1). Two studies, a systematic review [21] and meta-analysis [22], found that the da Vinci single port compared to multiport, had reduced time with the catheter post prostatectomy [22], reduced length of hospital stay and opioid/analgesia administration, with equivalent oncological and continence outcomes [21, 22]. The KANGDUO Robot® Surgical System is compared to the da Vinci Si system in two studies, one on robotic assisted radical prostatectomy (RARP) demonstrating comparable short-term functional and oncological outcomes, but longer operating times in 16 patients [41]. The second, a two-centre blinded randomised control trial, showed non-inferiority in robotic assisted partial nephrectomy (RAPN) for T1a renal tumours, but with longer docking and suturing times [42]. The Revo-i system was compared to da Vinci Si in RARP, producing similar short-term functional and oncological outcomes [69]. MicroHand Surgical Robot was compared to da Vinci, reporting shorter length of stay and reduced hospital costs in 45 patients undergoing sigmoidectomy, although did not specify the da Vinci generation [63]. It also demonstrated no difference in faecal continence following total mesorectal excision when compared to da Vinci Si [64] (Table 2).

Table 2 Key messages
Full size table

Comparing total costs, da Vinci X and Xi is reported to be at $1.2 and $2 million respectively and an average cost per operation of $2500 [17], although clearly this will have a significant range. Other comparable systems are touted to be cheaper with KANGDUO at $1 to 1.4million, with no comparable clinical data to X and Xi systems, only Si. The Senhance® Surgical System is stated to cost between $1 to 1.2 million [17], with per procedure comparisons with da Vinci stated to be cheaper in one study, $559 versus $1393, and comparable operative times [44]. Hinotori™ Surgical Robot System and Hugo™ RAS surgical system both state that their system is cheaper than the Xi, and the avatera® system has been quoted at $1.1 million [84].

Discussion

Our review has highlighted that full evaluation for robotic platforms has not been reported even on established robots, with the majority currently validated at stages 0 to 2b. Understandably, Intuitive is the only platform which has been fully evaluated, as it has had over twenty years to achieve long-term outcome, including randomised control trial, data. Publication of full evaluations for other systems is eagerly awaited.

The lack of evaluation reports represents a challenge for the surgical community given the rapid adoption of new systems. This situation, however, is likely to improve with time due to emerging platforms becoming commercially available.

We have provided an initial, comprehensive analysis of the platforms in this review, using The IDEAL Framework. Its intended use is for the evaluation of new, complex treatments within surgery through a logical, methodical pathway. Professor McCulloch (Chair of IDEAL) and the IDEAL team offer an explanation that competition, in this case between robotic companies, can often drive rapid adoption without full evaluation as defined by the framework. Although, safe evaluation exists with regulatory approvals before implementation into the clinical setting, it is possible that devices are introduced too quickly and not fully evaluated for certain procedures, given the increasingly competitive industry. On the other hand, it is worth noting that it would not be possible to reach stage 4 of evaluation without a platform being used in the clinical setting. Other explanations for rapid adoption pertain to the feasibility of performing multiple evaluations, across many different types of operations, within and between specialties. This would require considerable time and is unlikely to be cost-effective for robotic companies to wait for full evaluation [15]. Ultimately this would lead to the failure of bringing many platforms to market and the undesirable outcome of hindering technological progress within surgical specialties. It is also recognised that attitudes and process in healthcare differ worldwide including the adoption of new technology, therefore, evaluation of these will as well. However, broadly speaking clinicians should evaluate outcomes in the same way as IDEAL suggests i.e. through case report and series, prospective observational studies, randomised control trials against the gold standard, and long-term follow-up. Therefore, with rigorous regulatory approval and sound methodology from stage 0 to 2a, implementation of new and emerging robotic platforms is likely to be safe. Regarding long-term outcomes, multicentre, international registries could be an alternative solution to provide large data on evaluation across platforms and specialties with the European Association of Endoscopic Surgery (EAES) well positioned to provide this function for its members and beyond.

Another consideration when discussing further research within this area is whether it could distract or deviate finite resources from other fields in need. However, given that the IDEAL Framework evaluation relies on studies investigating clinical outcomes, the research required is likely to be transferable.

Several comparative studies were highlighted in Table 1, looking at clinical outcomes, however, there is greater research needed in this field. Studies are limited in scope, are often not independent from funding or involvement from the robotic company and none compare their systems to the fourth generation of Intuitive robots which are predominantly in use. Equally, these studies should not be ignored as often they demonstrate non-inferiority to the da Vinci Si i.e. safety of their use and clinical efficacy.

Considering the costs of each system it appears that the da Vinci Xi is the most expensive, but it is perhaps difficult to compare, with Intuitive Surgical Inc. producing its fourth generation. In fact, some of the systems highlighted have been created to have different capabilities and accessibility, therefore will be cheaper, but not comparable. For example, the Revo-I robot has been developed to do just this, improve accessibility, and is currently being used in Uzbekistan. It is also important to note that some of the costings quoted in the table were released by other companies or in news articles, so the reliability of this should be questioned. Lastly, details for the cost of many systems were not publicly available.

The environmental impact of robotic surgery is another important consideration. A systematic review [107] reported that robotics compared to laparoscopy had 43.5% greater greenhouse gas emissions and 24% higher waste production. Many, but not all, of the robotic systems highlighted have reusable instruments (Table 1) which will undoubtedly help to offset this. Current and emerging robotic companies should take the environmental impact of their product into account, especially for future generations of robot. This should extend beyond the procedure itself, into a more holistic approach of the perioperative pathway.

This study has a number of limitations. Firstly, although we have carried out a comprehensive search through various channels, there is a chance we may have missed emerging platforms. Evaluation is also a dynamic process and requires regular updates to provide a true account of platforms’ status.

Challenges were observed in performing a comprehensive search strategy to identify new systems, including a lack of visibility for some. Reports were occasionally not found despite being mentioned on a company’s website, making it difficult to ascertain the stage of evaluation. We had to utilise multiple resources including a literature and Google search, screening old reviews, technology articles and reaching out multiple times to company emails. These efforts are not feasible outside the research setting and it is unrealistic to expect a practicing surgeon to investigate new devices or platforms to this level of evaluation, in order to provide the user guidance.

The IDEAL Framework stage of evaluation has been awarded based on studies investigating only a limited number of operation types. There is an argument to evaluate and assign an IDEAL Framework stage for each operation type, as it would certainly differ. This topic deserves discussion and expert consensus on how to evaluate new surgical technologies and/or how the IDEAL Framework should be implemented.

It has also been argued that the framework is not optimally suited for the evaluation of future robotic systems [16]. Despite this, it is likely the best framework available which can be adapted to evaluate new technology, providing a standardised and quality-assured pathway. Importantly, the framework has been globally accepted to ensure the safe implementation of novel interventions.

Conclusion

The majority of existing robotic platforms are currently at the preclinical to developmental and exploratory stage of evaluation. Using the IDEAL framework will ensure that emerging robotic platforms are fully evaluated with long-term data, to inform the surgical workforce and ensure patient safety.