How can robot-assisted surgery provide value for money?

Introduction

Over the past decades, the implementation of robot-assisted surgery has increased tremendously. In 2000, about 1000 robot-assisted procedures were performed worldwide, whereas in 2018, that number had increased to more than 1 million.1 Besides an increase in the number of robot-assisted procedures performed, an expansion in types of procedures is seen. The first robotic systems were mainly used in urology (prostatectomy) and gynecology (hysterectomy). Nowadays, we see a huge increase in application in other fields such as general, gastrointestinal, and thoracic surgery.2 With the ever increasing pressure on sustainable healthcare, it is important to know how a surgical robotic system can be used most cost-effectively.

Currently, the da Vinci Surgical System of Intuitive is with 5114 installed systems worldwide the best known and most used system.3 Ever since the introduction of da Vinci Surgical System there has been both enthusiasm and skepticism, followed by a vicarious debate on its value for money. Reported effects of robot-assisted surgery include less conversions, reduced blood loss, fewer perioperative complications, shorter hospital stays, faster recovery, and less positive tumor resection margins,4–6 but these advantages are not confirmed in all studies and for all procedures. Moreover, these advantages come at a high financial investment.1 Hospitals spend between $1000 and $4000 more per robot-assisted case compared with endoscopic minimally invasive or open procedures, in addition to the purchase and maintenance costs of the surgical robotic system.7 Barbash and Glied estimated that if robot-assisted procedures replaced all conventional procedures, an additional $2.5 billion would be spent in annual healthcare costs in the USA.7

Sustainability of healthcare is an important issue in all countries. The increasing number of innovation entering the market demands healthcare providers to prioritize their investments in order to manage increasing costs and deliver the best possible outcomes for patients. Therefore, identifying which innovations provide most value for money is becoming increasingly important. As evidence regarding the cost-effectiveness of robot-assisted surgery is still lacking, the use of robotic surgery is becoming an important societal issue. In order to provide value for money, the increased costs of robot-assisted surgery need to be outweighed by the benefits. It is expected that in the coming years more providers (eg, Stryker, Johnson & Johnson, Medtronic and smaller companies) will increase their market share or enter the market with new surgical platforms leading to a further increase in robot-assisted surgeries in daily practice. With more providers of surgical robotic systems on the market, competing with Intuitive for new customers, it is expected that less expensive systems will become available and that robot-assisted surgery will be applied for a wider range of procedures.7 However, the question remains whether such new systems will provide value for money.

As different surgical robotic systems may have many different features and can be used for multiple indications, we aimed to develop a generic online interactive tool that can be used to analyze the (health) effects needed to compensate for the additional costs of using a surgical robotic system, in order to become cost-effective. This tool can be used to explore under which circumstances a (new) surgical robotic system could provide value for money and to inform clinical research. In this paper, we will describe the tool and illustrate its application using a hypothetical new surgical robotic system. The tool enables adjustment of all input variables (ie, acquisition costs robotic system, number of procedures yearly performed) for different surgical robotic systems, and different hospital settings and indications.

Model construction

Model output

The model can be used in two ways. First, the model shows the additional costs of performing a robot-assisted procedure and the effect needed on each aspect to compensate for these additional costs, for example, the number of hospital days that need to be prevented per procedure to offset the extra costs of a procedure with a surgical robotic system. Second, when the effects of a robot-assisted procedure are known, for example, a reduction in procedure time, they can be entered into the model. The model will show if the additional costs for a robotic procedure are compensated by the benefits. In case the (expected) benefits do not (yet) compensate for the additional costs of a robot-assisted procedure, it shows how much extra effect on each aspect is needed to outweigh the additional costs.

Clinical example

To illustrate how the model can be used, we will describe a hypothetical example for a procedure performed with a surgical robotic system compared with laparoscopy from a healthcare perspective. We would like to emphasize that the examples we used are all hypothetical to explain how the tool can be used, and do not represent real examples from clinical practice. For the costs of the robot-assisted procedure we assumed a new surgical robotic system will be used with an acquisition price of €1 million and a maintenance charge of €80 000 per year. For the endoscopic procedure, acquisition cost for the endoscopy tower of €100 000 and a maintenance charge of €8000 per year were assumed. For both devices, we assumed a useful life of 7 years and an interest rate of 4.2%.5 An average of 200 procedures per year for both robot-assisted and endoscopic surgeries were used to calculate the fixed costs per procedure (table 1). We assumed variable costs of €1500 and €1000 for the robot-assisted and endoscopic procedures, respectively (table 1). The assumed costs of a conversion, complications and a positive surgical margin are presented in table 1. Since we used a hypothetical example to explain the model, the population is not relevant. When one uses the model to a specific procedure the population needs to be specified.

For this example, the total extra costs for a robot-assisted procedure compared with laparoscopy were €1615 (€2739 minus €1124). To compensate for these extra costs, robot-assisted surgery should result in either a shorter operating time of 116 min, or a reduction in conversions of 95%, or a reduction of 32% of Clavien-Dindo I/II complications, or prevent 10% of Clavien-Dindo III–V complications, or prevent 32% of lasting complications, or reduce positive margins with 32%, or a reduction of 2.42 hospital days, or gain 0.032 QALYs (table 2). This means that, for example, 0.032 QALYs should be gained with robotic surgery compared with laparoscopic surgery to compensate for the extra costs of €1615 per robot-assisted procedure (table 2), when assuming a willingness to pay of €50 000 per QALY (table 1).

A surgical robotic system may provide effect on more than one aspect at a time. Therefore, a combination of effects seen or expected for a procedure can be entered into the model to determine if these would compensate for the extra costs involved with a robot-assisted procedure or if extra benefit is still needed. Table 3 provides three scenarios for which the effects would outweigh the extra costs of the procedure when those effects can be achieved using the surgical robotic system. When a reduction of 25 min operating time, as well as a reduction of 10% for complications with Clavien-Dindo I/II, and 5% for a Clavien-Dindo III–V can be achieved using the surgical robotic system, the benefits of these effects will outweigh the extra costs of the procedure (scenario I). A reduction of one hospital day together with a QALY gain of 0.02 will also result in sufficient benefit to outweigh the extra costs of a robot-assisted procedure (scenario II). Scenario III shows an example in which a combination of effect on operating time, conversion, complications and positive margin could result in compensating the additional costs of a robotic procedure. Scenario IV provides an example in which besides positive consequences, also a negative consequence, a longer operating time of 15 min, for a robotic procedure is included.

The additional costs for a robot-assisted procedure are dependent on various inputs such as the acquisition costs of the surgical robotic system and the number of procedures performed per year (online supplemental appendix I figures 1 and 2). The model can be used to assess the impact of such changes on the value for money. Table 4 shows the influence of the additional costs of a robot-assisted procedure when the acquisition costs of the surgical robotic system were varied from €500 000 to €2 500 000 and the number of procedures was varied from 100 to 300. In case the costs of the surgical robotic system are reduced to €500 000, the additional costs per procedure for our example were €1086 as compared with laparoscopy. The effects needed for each individual aspect to compensate for these extra costs are presented in table 5. The results show that a 7% reduction in complications with Clavien-Dindo III–V is sufficient to compensate for these additional costs. On the other hand, when performing 100 robot-assisted procedures per year the additional costs per procedure resulted in €4598 as compared with laparoscopy. A reduction of 29% in complications with Clavien-Dindo III–V is then needed to compensate the additional costs of the procedure.

Discussion

As far as we are aware, we are the first offering a generic and flexible insight in the potential cost-effectiveness of (new) surgical robotic systems. This is particularly important in the current era where many new systems are being developed while the affordability of healthcare is increasingly under pressure. The interactive model described in this paper provides relevant stakeholders, such as surgeons, purchasers, and policymakers, the unique opportunity to gain insight in what is needed to provide value for money when using (new) surgical robotic systems. It informs them on the effect that is needed to compensate the additional costs of a (new) surgical robotic system.

Earlier studies showed that surgical costs and effects for robot-assisted procedures may vary between indications.10 11 Gkegkes et al, for example, found ranges from €2539 to €57 002 and from €7888 to €16 851 for robot-assisted and endoscopic procedures, respectively.10 The costs used in our analysis, €2739 for a robot-assisted procedure and €1124 for an endoscopic procedure, seem low compared with these costs. This can be explained by the fact that the current surgical robotic system is more expensive than the costs we assumed for the new hypothetical surgical robotic system in the present study. Furthermore, we only included the fixed and variable costs of the equipment used and did not include other costs such as operating theater costs and anesthesia.

Some potential limitations of our model should also be discussed. First, less blood loss is often mentioned as a benefit for robot-assisted procedures, which we did not take into account in our model as we only focused on patient-related outcomes. However, blood loss may impact complications, which were incorporated in the model. Second, we also did not include potential ergonomic benefits for the surgeon in the model. Studies have shown that the ergonomics for the surgeon associated with robotic surgery is better compared with laparoscopic surgery.12–14 Ergonomic problems may affect a surgeon’s physical workload and potentially may even lead to absenteeism. As the costs of absenteeism of a surgeon are estimated at €1129 per day,5 while the extra costs per procedure with the robot are €2345, a saving of more than 2 days of absenteeism per procedure performed is needed to compensate the extra costs. However, although it is very relevant, this issue is difficult to quantify. Quality of life of the surgeon can be one of the reasons, next to value for money, for using or purchasing a surgical robotic system. Third, we did not include learning curve effects and a potential increase in procedures that can be performed with future robotic systems that cannot be performed endoscopically yet. Especially for procedures that are performed with open surgery, robotic systems might affect the costs as well as the volume of procedures, since more patients could be treated on 1 hour. Furthermore, due to advances in technology, it is expected that in the future the duration of learning curves will reduce for robot-assisted surgery.15 Furthermore, we only focused on including costs from a healthcare perspective. Productivity gains for patients who can return to work more rapidly were not taken into account. Also, we did not look at the reimbursement hospitals will receive from insurers for the additional hospital spending for robotic surgery. Therefore, we cannot provide insight in the value for money of robotic surgery from a societal or hospital perspective.

We aimed to follow the international CHEERS reporting guideline. However, as we did not conduct a standard cost-effectiveness study, some aspects were not applicable for our model. One should be aware that when using our model to perform a cost-effectiveness analysis, it is advised to follow all components of the guideline.

The clinical implication of our model is that more studies into the effectiveness of new surgical robotic systems seem warranted as these systems might become cost-effective, but only if they indeed also have added clinical benefits and if future systems become less costly. The many hospitals that already adopted a surgical robotic system can use the model to explore how the system should be used to provide most value for money by changing the costs and effects to their specific setting. As there is a lot of variation between indications, our model can also easily be used to assess if the expected effects for an indication will provide sufficient benefit to provide value for money. In this way, the model can be used to prioritize research.

Hospitals may also have other reasons for investing in costly new advanced technology, for example, scientific purposes, patient demand, fear of missing out, and competition between hospitals.16 17 On the other hand, we believe that in the current era of rising healthcare costs, hospitals and other relevant stakeholders have to educate patients and society that new is not always better, and that money should be spent wisely. Our interactive model helps them in this discussion.

The utilization rate of a surgical robotic system showed a large influence on the extra cost per procedure. Increasing the number of procedures in our example, from 100 to 300 per year, appeared to reduce the extra costs per procedure with €3005 when compared with laparoscopy. This might lead to the clinical implication that the number of procedures with the robotic system should be as high as possible. However, the additional variable costs per robot-assisted procedure have to be compensated by a minimum health gain, otherwise the total costs per year will increase even though the fixed costs per procedure will decrease when more procedures are performed. For this reason, it is important to study specific indications and to assess whether the variable costs are indeed compensated by a health gain for that specific indication. Setting the fixed costs in the tool to zero can help explore whether this is the case.

In conclusion, our model and interactive tool provide a unique opportunity for all stakeholders such as researchers, clinicians and policymakers, to gain insight in the benefit needed to outweigh the additional costs of a (new) surgical robotic system or, when the benefits are known or can be estimated, to assess the value for money for a specific procedure. To achieve the maximum value for money, we recommend assessing for each indication whether the necessary effects seem feasible. The online tool (see also: https://sejal.shinyapps.io/supplement_robot-assisted_surgery_article/) provides this information, and can be easily used by all stakeholders.