Robot-assisted surgery for soft tissue procedures: early value assessment

Access to minimally invasive surgery can depend on the physical characteristics of the person, the type of condition and the type of procedure. Experts said that robot assistance can make minimally invasive surgery an option for some procedures and for people who did not have this option before. Expert opinions and evidence in the external assessment report indicated that improved ergonomics with robot assistance makes it easier for surgeons to do technically challenging surgery. So, some procedures that could only be done through open surgery can now be done using minimally invasive techniques. Also, more surgeons may be able to do more challenging surgery. Experts said that the technologies have made it possible for different specialties to collaborate on complex multidisciplinary procedures. Experts said that the groups of people that minimally invasive surgery could become available to will vary between specialties and procedures. But, experts gave examples such as people with multimorbidity, people with high body mass index, and people needing complex surgery or neoadjuvant treatment. Experts said that hospitals with robotic systems and training programmes may attract candidates for surgical training.

The committee noted that a wider NHS England robot-assisted surgery steering group is coordinating national strategies for training, procurement and implementation of robot-assisted surgery services, and guidance on surveillance of robot-assisted surgery programmes. The committee highlighted that national strategy from the steering group on the procurement and implementation of new robotic systems may reduce inequalities to access (see section 3.10).

Many centres already have a robotic system and most of the technologies included in this assessment are already used in the NHS. The committee noted that some centres will be newly introducing a robotic system. Many centres will aim to maximise benefits from existing robotic systems, and some may aim to expand their robot-assisted surgery service. The NHS England robot-assisted surgery steering group has noted that there is rapid adoption in some specialties such as colorectal and gynaecological surgery. But, adoption and expansion are unequally distributed across the UK and across specialties. A coordinated approach to provide for future demand is part of the remit of the NHS England robot-assisted surgery steering group.

In NHS practice, minimally invasive surgery is done less often in areas of the country that are more deprived than in those that are less deprived. Also, there has been lower uptake of robot-assisted surgery in some areas of the UK than others. The highest volume centres are mostly in or around London. Experts said that the geographic placement of additional robotic systems, and the availability of training, resources and staff to implement robot-assisted surgery services for soft tissue procedures, could worsen these disparities. These concerns were reiterated by patient organisation and patient expert feedback. The NHS England robot-assisted surgery steering group may be influential in moderating this with future national strategy. It is actively analysing and mapping current robot-assisted surgery provision in England. A key priority will be equitable provision of robot-assisted surgery based on need rather than current configuration.

Responses from 5 patient organisations said that shorter length of stay, faster recovery, and faster return to work and usual activities were among the key perceived benefits of robot-assisted surgery among patients. Patients believe that robot-assisted surgery could widen access to minimally invasive surgery. Also, patients reported experiencing fewer side effects, and less pain and scarring. Patient experts reiterated these points in the committee meeting.

Patient organisations and experts said that the main concerns were around access to robot-assisted surgery, including the need to travel and wait times if there are not enough robots available. They reflected the potential of these factors to exacerbate health inequalities. Also, patient organisation submissions and patient experts said that clear and transparent information about robot-assisted surgery and reassurance about quality assurance was needed because it is an unfamiliar method of surgery. The patient organisations and the patient experts were aware of the high cost of the technologies and saw this as a concern.

The EAG did literature searches across 8 databases and 2 clinical trial registries for studies that named eligible technologies. A total of 492 full texts were retrieved and examined. In addition, 62 relevant studies identified from company submissions were considered. The EAG prioritised studies most relevant to the UK on a per-technology basis as outlined in the protocol for this assessment. They prioritised comparative studies, done in the most UK-relevant contexts, that explicitly assessed the technologies in scope. Twenty comparative studies were prioritised for assessment:

• 1 UK-based randomised controlled trial (RCT)

• 5 non-randomised prospective studies

• 4 historically controlled cohort studies

• 10 retrospective cohort studies, including 2 done in the UK.
  
  Evidence for all 5 technologies was included in the EAG report, but the amount of evidence per technology differed. Most of the evidence (13 out of 20 studies) was for Da Vinci X and Xi. The committee noted that this technology has been in use the longest and is currently the most used in the UK. The Senhance Surgical System was used in 2 studies, whereas the Hugo robotic-assisted surgery system, Versius Surgical System and Da Vinci SP were used in 1 study each. The EAG noted that, in the prioritised evidence, there was only 1 RCT, few studies were done in the UK and the studies were small. All the prioritised evidence compared the technology against laparoscopic surgery or open surgery. Colorectal, hernia repair, gastrointestinal, gynaecological, pancreatic, testicular and urological surgical procedures were represented in the prioritised evidence. The EAG noted that none of the studies explored the generalisability of clinical findings across different specialties. The committee heard from experts that robot-assisted surgery practices differ in the UK compared with Europe and the US, so evidence from those regions may not be generalisable to the NHS.

Experts said that the effect of learning curve on outcomes was the most important factor that limited interpretation of the evidence and generalisability of findings. Experts said evidence from outside the UK may be useful to understand clinical outcomes from surgeons and centres that are further along the learning curve. But, they said evidence from outside the UK on the learning curve itself may be less relevant. This is because the time taken to move along the learning curve is affected by opportunity and volume of surgery, which differs between the UK and other countries. Experts said there may be limited generalisability for outcomes between procedures in different specialties and studies. For example, length of hospital stay may depend on factors unrelated to the surgical technique used.

The committee commented that there was little evidence in paediatric groups. It noted that only 1 technology is indicated for use in children (Da Vinci X and Xi). Experts said that fewer UK centres currently use robot-assisted surgery in children compared with adults.

Experts noted that there is additional evidence that did not meet the EAG's search and inclusion criteria. In response to a call to experts and companies for additional evidence that could fill evidence gaps in the report, the EAG included 10 additional studies in an addendum to its report:

• 5 RCTs

• 1 RCT with economic analysis

• 1 large real-world database study

• 1 case-control survey of surgeon ergonomics

• 1 matched cohort study

• 1 retrospective comparative study.
  
  The studies either did not specify which robot was used (but the model could be inferred given the date and location of the study), or used older models of the robotic systems included in the scope of this assessment. The EAG also reviewed and summarised 17 recent systematic reviews. The committee considered the evidence in the addendum and concluded that the additional evidence generally supported the findings of the 20 prioritised studies. This was that robot-assisted surgery is generally comparable with standard minimally invasive surgery across a range of outcomes.

The EAG developed a cost-comparison model to investigate the potential benefit of robot-assisted surgery over a 1-year time horizon. It explained that modelling a longer time horizon would be associated with substantial uncertainty because of the differences in procedures, populations and surgical settings. But, the EAG did a scenario analysis to investigate the long-term benefit needed for robot-assisted surgery to be cost effective. The committee agreed that the EAG's approach was suitable for modelling multiple procedures and populations in light of the uncertainty for most outcomes. Over a 1-year time horizon, the cost-comparison base case and scenario analysis found that robot-assisted surgery for soft tissue procedures was likely to be cost-incurring. But, the scenario analysis indicated that it could plausibly become cost effective if robot-assisted surgery led to long-term quality-adjusted life year (QALY) gains. The EAG and experts emphasised that the range of scenario analyses should be considered carefully alongside the base case because the scope included a wide variety of potential procedures, patients, robot utilisation and costing structures.

In the base case, the technologies were estimated to increase healthcare costs. These were driven by the upfront cost of the robot and additional consumable equipment needed to do the procedures. Short-term costs that could be reduced by robot-assisted surgery (for example, from reduced complications, readmissions and surgery conversions) were not considered likely to outweigh the cost of using the technologies. All short-term scenarios led to cost-incurring results, but results were more favourable when robot-assisted surgery replaced open surgery instead of standard minimally invasive surgery. See section 8.3.1 of the report for more details on the scenario analysis. One-way sensitivity analysis showed that the key drivers of the model were likely:

• proportion of surgeries that were open surgery, standard minimally invasive surgery and robot-assisted surgery

• length of surgery

• conversion rates to open surgery

• disposable component costs for robot-assisted surgery.

In the long-term scenario analysis, the EAG found that if people who had robot-assisted surgery gained at least 0.1 QALYs (equivalent to 36.5 days in full health) compared with other surgery, then robot-assisted surgery could plausibly be cost effective. Experts generally agreed that this was feasible over a longer-term time horizon than 1 year. Some experts said that it may be more feasible in some specialties and procedures than others. This is because the gains would likely come from reduced severity of disease or reduced disease progression. Long-term clinical benefits, reduced operation times and high use of the robot when it has been purchased outright or leased would also make long-term cost-effectiveness more likely. The EAG said that benefits may have been underestimated if data used to populate the model was more representative of surgeons and centres that were still on the learning curve, rather than those that were operating at the end of their learning curve.

The committee concluded that there was enough evidence of a potential benefit of robot-assisted surgery technologies for soft tissue procedures for them to be used in the NHS while further evidence is generated. The main evidence gaps for soft tissue procedures (excluding prostatectomy) are:

• Learning curve: The EAG noted that most studies did not clearly report the level of surgeon and centre experience of using the robotic system, or adjust for it. Learning curve was characterised or measured in different ways when studies did report it. Experts emphasised that differences in findings across a range of outcomes may be attributable to the surgeon and centre learning curve. The EAG noted that potential cost effectiveness may be underestimated if surgeons and centres were operating on the learning curve during the study.

• Resource use: There was little information on the costs of setting up and training staff for a robot-assisted surgery service for soft tissue procedures. This will likely vary depending on the mixture of procedures being done. Also, resource use during surgery including healthcare professional resource and consumables was difficult to quantify.

• Costing structures to procure a robotic system: There was little information on available costing structures to model cost effectiveness, other than for upfront cost for each system. Costing structures other than the 3 modelled in the EAG's analysis may be available to trusts. Costs of the robotic system and consumables drove incremental costs associated with robot-assisted surgery in the base-case analysis.

• Outcomes: The prioritised studies did not have evidence for some outcomes listed in the scope of this assessment. Little or no evidence was available for:

  • health-related quality of life

  • procedure-related discomfort and ergonomics for the surgeon

  • rates of minimally invasive surgery compared with open surgery after centres introduce robot-assisted surgery

  • volume of procedures and robot utilisation

  • hospital capacity and wait list reduction, and

  • long-term outcomes, including:

    • return to normal activities

    • survival and

    • need for revision surgery.
      
      Limited data on some of these outcomes may have limited the economic model. Also, rates of conversion to open surgery, complications, and outcomes including length of hospital stay and health-related quality of life were key drivers in the economic model. Robust evidence, adjusted for learning curve, is needed on these outcomes.

• Surgeon experience: Evidence on ergonomics for the surgeon and procedure-related discomfort (for example, using the SURG-TLX outcome measure) will help the committee understand how the technologies are benefitting surgeons.
  
  The committee concluded that further evidence should be generated to address the identified evidence gaps. Specialist committee members noted that it should be clear from the evidence which robotic system has been used. The committee noted that head-to-head studies may be difficult to do. Experts explained that whenever this is the case, other study designs, including non-inferiority studies, can be considered. The committee noted the potential of using real-world data for evidence generation. It also noted the potential benefits for the NHS of collecting long-term evidence across different robot-assisted surgery platforms at the system level, such as through a national registry.

Experts made the committee aware of 3 ongoing UK trials on robot-assisted surgery that may have relevant data to contribute to a future assessment:

• The REINFORCE trial, a real-world, in-situ trial evaluating the introduction and scale-up of robot-assisted surgical services in the NHS, and its impact on clinical and service delivery, effectiveness and cost. This is a stepped-wedge randomised trial with process evaluation and economic evaluation (NIHR131537).

• The MASTERY study measuring the quality of surgical care and setting benchmarks for training using Intuitive Data Recorder technology (ISRCTN 273555).

• The MAYFLY study assessing the clinical, economic and efficiency outcomes of using robot-assisted surgery for outpatient procedures in England (IRAS ID 327536).