According to our study, robotic surgery for performing hysterectomies for benign pathologies is generally associated with a significant additional cost per complication compared to laparoscopic surgery (ICER at €377,534).
Regarding cost in our study, robotic surgery represents a total cost of €6615 at 1 month compared with 3859 € for laparoscopic surgery, i.e. an additional cost of 2756 € in favour of robotic surgery, with an additional cost of €248,288 per major complication avoided +/- transfusion +/- periprocedural bleeding ≥ 500 mL.
In a 2013 study by Wright et al., analyzing data from the Perspective Premier database of benign hysterectomies performed in several hospitals in the United States from 2007 to 2010, robotic surgery represented higher median total cost than laparoscopic surgery ($8868 vs. $6679) for both fixed and variable costs (22).
In addition, a cost-volume analysis was carried out and found a decrease in the total cost of robotic surgery the greater the number of cases performed per physician.
In the same vein, the 2017 study by Dandolu et al., in the USA, using the TruvenHealth Market Scan private insurance database, analyzed the total costs of benign hysterectomies between 2008 and 2012, with a follow-up period of 90 post-operative days.
The total cost of the robotic procedure was $16,820 vs. $13,031 without robotic assistance, not valuing the robot for use in every procedure, but rather encouraging the careful identification of patients who could benefit from it (23).
However, some studies moderate this cost difference. For example, Winter et al., published in 2015 an analysis of costs of robotic versus standard laparoscopic hysterectomy in two US hospitals. The total cost of hospitalization included operating room time, reusable and disposable instruments in the operating room, staff, hospital room rate, pharmacy and laboratory costs. Robotic and laparoscopic procedures have been performed only by experienced surgeons (at least 30 procedures already performed). Excluding adjustment, the difference in cost between the two approaches was $1558.2, but after adjustment for hospital, age, BMI, previous abdominal or pelvic surgery and uterine weight, the cost of robotic surgery was higher but not statistically significant compared to laparoscopic surgery, with a mean difference of $283.1.
This study suggests that robotic-assisted hysterectomy may be similar in cost to standard laparoscopic hysterectomy when patient characteristics are considered (24).
In terms of operating time in our analysis, robotic surgery is significantly longer than laparoscopic surgery (153 minutes on average vs. 120 minutes, with an additional room occupancy time of 54 minutes in favour of robotic surgery), which represents an additional cost of €216 in terms of human resources and €178 in terms of cost of operating room.
In the study by Billfeldt et al., which prospectively collected from 2009 to 2015 data contained in the Swedish national GynOp registry, which covers 75% of all gynecological procedures in Sweden, and compared open, standard laparoscopic and robot-assisted hysterectomies for benign indications, the operating time was longer for conventional laparoscopic surgery than for robotic surgery (127 minutes vs. 104 minutes) (25).
Some studies showed a shorter learning curve for robotic surgery than for conventional laparoscopy, with around 50 robotic surgery procedures to be performed to gain competence (26).
Two studies (Martinez et al. 2014 and Lönnerfors et al. 2015) reported significant results in favor of robotic surgery, with mean operating times of respectively 154 and 76 minutes versus 185 and 104 minutes respectively for laparoscopic surgery (27) (28), while two other studies (Sarlos et al. 2012 and Paraiso et al. 2013) reported significant results in favour of laparoscopic surgery, with operative times respectively of 106 and 172 minutes for robotic surgery versus 75 and 102 minutes for laparoscopic surgery (29) (30).
This difference may be explained by the fact that the surgeons performing these procedures in the Sarlos et al. and Paraiso et al. were highly experienced laparoscopic surgeons at the beginning of their learning curve for robot-assisted surgery, whereas in the Lönnerfors et al. study, 4 of the 6 surgeons performing the procedures were experienced in robotic surgery. Thus, the length of the operation may represent a limiting factor in terms of the cost of robotic surgery compared to laparoscopy, but this limitation is bound to diminish as the surgeon's experience.
Initial length of stay in our study was not significantly different between robotic and laparoscopic surgery (2 days versus 3 days), with an additional cost of €310 in favor of laparoscopic surgery. These results are comparable with those reported in the literature.
In a 2015 study by Albright et al., conducting a systematic literature review with meta-analysis of randomized and quasi-randomized studies evaluating robotic vs. laparoscopic surgery for hysterectomies of benign indication, there was no significant difference in length of stay between robotic and laparoscopic surgery, with mean lengths of stay ranging from 1 to 3 days (31). In the previously cited study by Wright et al., results were more moderate and then laparoscopic surgery more frequently had a hospital stay > 2 days than in the robotic surgery group (22).
Finally, the robot is less prone to complications than laparoscopic surgery, which has a positive impact on the overall cost of this practice. In our study, the percentage of major complications, associated or not with transfusions and estimated blood loss ≥ 500mL, was lower in the robotic surgery group than in the laparoscopic surgery group, without this difference being statistically significant.
Most studies published in the literature report similar results with no significant difference between the two approaches in terms of post-operative complications, transfusions and intraoperative bleeding.
In a Lim et al. study comparing laparoscopic, open and robotic surgery approaches for hysterectomies for benign indications, the post-operative results of robotic surgery were better in terms of complications, with a post-operative 30-day complication rate of 6.3% versus 16.3% for laparoscopic surgery; the difference being statistically significant (6). This disparity in terms of post-operative complications in may be explained by the fact that the surgeons performing the procedures were experienced in robotic surgery with at least 60 hysterectomies performed. In a previously cited study by Wright et al. there was no significant difference in perioperative complications (5.5% for robot vs. 5.3% for laparoscopy), or transfusion rate (22).
The lack of significance can be explained, once again that surgeons were only at the beginning of their learning curve for robotic surgery techniques ; the authorization to use the Da Vinci® robot in gynecology preceded the start of the study by two years. Thus, the surgeon's experience in robotic surgery limits these complications, which has a positive impact on the robot's cost, leading us to expect a reduction in the robot's overall costs in the future.
This is the first French study to analyse the medico-economic robotic surgery compared with laparoscopic approaches.
The analysis of costs, including direct and healthcare costs, as well as clinical criteria with postoperative complications, transfusion, blood loss ≥ 500mL, operating time and hospital stay, is detailed and comparable with the literature.
The use of the Clavien-Dindo classification enabled a standardized, objective and reproducible assessment of post-operative complications. Group comparability was ensured using the inverse propensity score weighting method.
In this way, the analysis between the different approaches could be adjusted by taking into account all covariates that could affect the choice of approach and efficacy criteria, including BMI, history of abdominal surgery, anticoagulant treatment, anticoagulant therapy, ASA score and uterine weight. Finally, this study reports few missing data.
Our study also has limitations, the main one being the lack of experience in robotic surgery at the time of the study. Indeed, the robot was introduced to the department in 2016, the year in which the data collection began. As a result, the robotic procedures included in this study were carried out when the practitioners were only at the beginning of their learning curve. This may have a definite impact on operative and post-operative complications, and by extension the initial length of stay and potential re-hospitalization. In addition, this study takes into account major post-operative complications occurring the month following surgery.
However, a longer-term follow-up would enable us to evaluate clinical and economic consequences beyond the first postoperative month. Furthermore, this is a retrospective monocentric study, with the various limitations analysis methods, notably the lack of external validity of the results, only on a national scale. A multicentric analysis would be desirable to provide a more accurate reflection of the practice of robotic hysterectomy in France. Finally, a point not addressed in our analysis is the societal aspect.
Indeed, our study is carried out hospital perspective and focuses on the direct costs and healthcare costs associated with each approach, but without taking into account societal aspects such as time to return to normal activity, the impact this may have in terms of lost wages and productivity, as well as the impact on businesses. Robotic surgery enables shorter hospital stays and a faster return to normal activity, so taking this societal cost into account could reduce the overall extra cost of robotic surgery and improve its acceptability.
Finally, our study did not include an analysis of the patient's quality of life because of the retrospective nature of the study and the absence of a questionnaire given to the patient. Calculating an ICER that takes into account quality of life rather than complications could be more favourable.