Robotic-assisted versus minimally invasive mitral valve repair: a claims data analysis of clinical and cost outcomes
Highlight box
Key findings
• Robotic-assisted mitral valve repair (MVr) had higher operative day costs but comparable outcomes and total inpatient costs to minimally invasive MVr (Mini-MVr).
What is known and what is new?
• Minimally invasive and robotic MVr offer benefits over sternotomy.
• This study is the first in Japan to compare costs between Mini-MVr and robotic-assisted MVr.
What is the implication, and what should change now?
• Robotic-assisted MVr is feasible and safe; its use may be expanded in high-volume centers with cost-efficiency strategies.
Introduction
Mitral valve repair (MVr) for degenerative mitral valve regurgitation (MR) is recommended if a successful and durable repair is possible (1). The practice of MVr has undergone numerous advances, and access to the mitral valve has also seen developments from median sternotomy to a right minimal thoracotomy (Mini-MVr) and totally endoscopic robotic approaches using robotic telemanipulation.
Minimally invasive and robotic-assisted MVr have short-term advantages compared to sternotomy cases such as cosmetic advantages, shorter hospital stay, and faster return to normal activities (2,3). However, data comparing robotic to Mini-MVr are conflicting, with the Society of Thoracic Surgeons (STS) database showing disadvantages (3) and the Japan Cardiovascular Surgery Database showing advantages of robotic surgery, including the acute clinical outcomes (4).
Also, the cost associated with initiating a robotic cardiac service is substantial; investment in a surgical robot is followed by the ongoing financial commitment to device maintenance. The STS database (2) had reported there were equivalent in-hospital costs between robotic-assisted and sternotomy MVr, however, Mihaljevic et al. compared the economic benefits of robotic-assisted MVr, taking into account purchase costs, maintenance and income from return to paid employment as well as costs of postoperative care (5). Overall costs for robotic-assisted MVr were higher than alternative approaches; however, in high volume centers performing 55–100 cases per year, the cost-effectiveness of robotic system over time was equilibrated with alternatives. In fact, since robotic-assisted MVr using the da Vinci surgical system (Intuitive Surgical Inc., Sunnyvale, CA, USA) was covered by Japanese national health insurance in April 2018, there were only a few institutions in Japan performing over 55 cases per year of robotic-assisted MVr, as such procedures have not yet been widely adopted. Moreover, despite Mini-MVr and robotic-assisted MVr being reimbursed at the same rate under Japan’s insurance system, to the best of our knowledge, no studies have examined the cost differences between these approaches.
Our objective was to conduct a claims-based analysis in Japan to evaluate and compare the inpatient costs as well as acute clinical outcomes between Mini-MVr and robotic-assisted MVr. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1086/rc).
Methods
Data source
This was a retrospective study using a Japanese administrative claims database provided by Medical Data Vision Co., Ltd. (MDV; Tokyo, Japan). The MDV database is derived from hospitals accredited as providing acute care services in Japan and participating in the Diagnostic Procedure Combination/Per-Diem bundled payment system (termed DPC hospitals). As of April 1, 2023, the MDV database included anonymized data from 473 contracted hospitals, representing about 27% of the 1,761 DPC hospitals in Japan. The MDV database includes the following information on each patient: patient demographics, main diagnoses and comorbidities, procedures, reimbursements, and discharge summary data. The Department of Medical Statistics at Osaka Metropolitan University had purchased the deidentified individual data from MDV.
Ethical statement
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The Osaka Metropolitan University Institutional Review Board reviewed this study and determined that formal ethical approval was not required because the data were anonymized by the database before the initiation of the study. Informed consent was waived in this retrospective study.
Study population
The study population for the main analysis included all patients who underwent general anesthesia and were identified using the procedure code master (L008) from February 2019 to July 2023. Patients registered in the MDV database were selected using the following steps (Figure 1).
The patients’ background characteristics and postoperative outcomes were defined under the following International Classification of Diseases 10th revision (ICD-10) diagnosis (Table S1), DPC procedures codes (Table S2), and the discharge summary data. The rates of concomitant procedures and material usage were identified based on reimbursement records.
Statistical analysis
Categorical data were expressed as frequencies and proportions and compared using the chi-squared test or Fisher’s exact test as appropriate. Meanwhile, continuous data were expressed as median with interquartile ranges (IQR) or means; unpaired data were compared using the Mann-Whitney U test for non-parametric analysis. Hospitalization costs were aggregated based on the DPC procedure codes under the fee-for-service structure of DPC system. Since the concomitant Maze procedure significantly impacts operative day costs, it was included as a covariate and assessed using the Mann-Whitney U test.
The R software version 4.3.1 (The R Foundation for Statistical Computing, Vienna, Austria) was used to conduct statistical analyses. The statistical significance was defined as 2-sided P<0.05. All P values may not be interpreted as confirmatory but descriptive.
Results
Patient background and procedural data
The patients’ background characteristics and procedural data are presented in Table 1. The median age was 60 (IQR, 51–69) years in the Mini-MVr group and 60 (IQR, 52–67) years in the robotic-assisted MVr group. Patients were stratified by hospital bed volume into three categories: large (≥500 beds), middle (200–499 beds), and small (<199 beds). In the Mini-MVr group, 737 patients (51.5%) were treated at large-volume hospitals, whereas all 110 patients (100%) in the robotic-assisted MVr group underwent surgery at large-volume hospitals.
Table 1
| Variables | Mini-MVr (n=1,430) | Robotic-assisted MVr (n=110) | P |
|---|---|---|---|
| Age, years | 60 [51, 69] | 60 [52, 67] | 0.85 |
| Sex, male | 906 (63.4) | 74 (67.3) | 0.47 |
| BSA, m2 | 1.67 [1.51, 1.81] | 1.69 [1.52, 1.83] | 0.56 |
| NYHA over class 3 | 25 (1.7) | 0 (0.0) | 0.25 |
| Hypertension | 1,253 (87.6) | 91 (82.7) | 0.14 |
| Dyslipidemia | 601 (42.0) | 44 (40.0) | 0.69 |
| Diabetes mellitus | 1,302 (91.0) | 54 (49.1) | <0.001 |
| Chronic kidney disease over stage 3 | 137 (9.6) | 5 (4.5) | 0.09 |
| Chronic obstructive pulmonary disease | 52 (3.6) | 2 (1.8) | 0.43 |
| Ischemic heart disease | 1,213 (84.8) | 42 (38.2) | <0.001 |
| Stroke | 0 (0.0) | 0 (0.0) | >0.99 |
| Hospital bed volume | <0.001 | ||
| Large (≥500) | 737 (51.5) | 110 (100.0) | |
| Middle (200–499) | 290 (20.3) | 0 (0.0) | |
| Small (≤199) | 403 (28.2) | 0 (0.0) | |
| Concomitant procedures | |||
| Maze procedure | 170 (11.9) | 12 (10.9) | 0.88 |
| Blood transfusion | 473 (33.1) | 48 (43.6) | 0.03 |
| RBC transfusion | 388 (27.1) | 47 (42.7) | <0.001 |
| RBC units, means | 1.6 units | 2.5 units | <0.001 |
| FFP transfusion | 287 (20.1) | 44 (40.0) | <0.001 |
| FFP units, means | 1.0 units | 1.7 units | <0.001 |
| PC transfusion | 91 (6.4) | 10 (9.1) | 0.31 |
| Fibrin-based hemostatic agents | 933 (65.2) | 107 (97.3) | <0.001 |
| Cellulose-based hemostatic agents | 88 (6.2) | 3 (2.7) | 0.20 |
Continuous variables are presented as median [IQR] or mean, whereas categorical variables are presented as number (%). BSA, body surface area; FFP, fresh frozen plasma; IQR, interquartile range; MVr, mitral valve repair; NYHA, New York Heart Association; PC, platelet concentrates; RBC, red blood cell.
The Maze procedure was performed in 170 patients (11.9%) in the Mini-MVr group and 12 patients (10.9%) in the robotic-assisted MVr group, with no significant difference between the groups. No other concomitant procedures were performed.
The rates of blood transfusion on the operative day were 33.1% in the Mini-MVr group and 43.6% in the robotic-assisted MVr group. While the usage rates of fibrin-based hemostatic agents, such as the TachoSil Sheet, the Beriplast P (CSL Behring LLC, PA, USA), and the Bolheal Sealant (KM Biologics, Kumamoto, Japan) were significantly higher in the robotic-assisted MVr group, the usage rates of cellulose-based hemostatic agents, such as the Surgicel (Johnson & Johnson company, NJ, USA), were not significantly different between the groups.
Postoperative outcomes and the costs
In-hospital mortality occurred in 7 patients (0.5%) in the Mini-MVr group and in none in the robotic-assisted MVr group, with no statistically significant difference between groups. Although the length of hospital stay was statistically shorter in the robotic-assisted MVr group [10 days (IQR, 9–13)] compared to the Mini-MVr group [10 days (IQR, 7–14); P=0.04]. Although stroke incidence was extracted based on diagnostic codes, no cases were identified in either group.
The costs of the operative day were significantly higher in the robotic-assisted MVr group than in the Mini-MVr group (P<0.001), with or without Maze procedure. In contrast, there was no significant difference in the postoperative costs between the groups (P=0.54) (Figure 2, Table 2).
Table 2
| Variables | Mini-MVr (n=1,430) | Robotic-assisted MVr (n=110) | P |
|---|---|---|---|
| In-hospital death | 7 (0.5) | 0 (0.0) | >0.99 |
| Hospital stay, days | 10 [7, 14] | 10 [9, 13] | 0.04 |
| Costs in the operative day, ×106 (JPY) | 3.12 [2.95, 3.37] | 3.38 [3.27, 3.57] | <0.001 |
| Without Maze procedure | 3.07 [2.94, 3.28] | 3.36 [3.26, 3.47] | <0.001 |
| Costs in the postoperative days, ×106 (JPY) | 0.62 [0.41, 1.10] | 0.59 [0.50, 0.80] | 0.54 |
Continuous variables are presented as median [IQR], whereas categorical variables are presented as number (%). IQR, interquartile range; JPY, Japanese yen; Mini-MVr, minimally invasive mitral valve repair; MVr, mitral valve repair.
Subgroup analysis in large bed volume hospitals
A subgroup analysis was conducted, including only patients treated at large bed volume hospitals (n=847). Consistent with the primary analysis, the costs of the operative day were approximately 0.2 million yen higher in the robotic-assisted MVr group compared to the Mini-MVr group (P<0.001). In contrast, no significant difference was observed in the postoperative costs between the two groups (P=0.41).
Discussion
The present study used a nationwide claims database to compare the cost and safety of robotic-assisted MVr. The major findings of the current study are as follows: (I) the cost incurred on the day of operation was significantly higher for robotic-assisted MVr; (II) the use of blood transfusions and fibrin-based hemostatic agents on the operative day was significantly more frequent in robotic-assisted MVr; and (III) there was no significant difference in the postoperative costs.
The costs associated with robotic surgery can be deconstructed into four broad categories: capital costs, operative costs, postoperative costs, and post-hospital costs (6). This study primarily focused on operative and postoperative costs. A 2024 analysis using a nationwide database in the US found that the median cost of robotic-assisted MVr was greater than that of conventional surgery, mainly due to the technology involved with robotic-assisted treatment and the longer operative times (7). It is important to note that our analysis did not include the capital costs associated with robotic systems, such as the purchase price (typically reported to range from 1 to 2 million US dollars), annual maintenance fees (approximately 100,000–200,000 US dollars), or the per-case cost of instruments (estimated at 2,000–4,000 US dollars) (8,9). Although our study was limited to the procedural and material costs covered by the Japanese reimbursement system, even under these constraints, robotic-assisted MVr was found to be significantly more expensive than Mini-MVr on the operative day. On the other hand, when focused exclusively on facilities that perform the most MVr procedures annually, this cost difference was no longer statistically significant. This may reflect better patient selection protocols, more efficient operating room practices, and expedited care pathways at such institutions (7). In our study, subgroup analysis by hospital bed volume showed that robotic-assisted MVr was performed exclusively at high-volume hospitals. Nonetheless, the operative day cost remained higher even in these high-volume institutions. Further research should include a broader range of institutions to explore the potential “cost-volume relationship” in the context of robotic-assisted MVr.
In this study, the rate of blood transfusion on the operative day was significantly higher in the robot-assisted MVr group, with approximately 40% of patients receiving transfusions. This rate exceeds those reported in previous database studies and may have contributed to the higher operative day costs. A comparison of minimally invasive and robotic mitral surgery in the STS database showed a significantly higher transfusion rate in the robotic group (3). In contrast, the Japanese National Clinical Database reported no significant difference between the two groups (4). However, transfusion rates for robot mitral surgery in these databases were generally around 20–30%, which is notably lower than the rate observed in the present study. Given the limited number of robotic-assisted cases in the current database, selection bias is a likely contributing factor. It should be noted, however, that in this study, any transfusion administered on the operative day was counted, even if no intraoperative transfusion occurred. Therefore, the reported transfusion rate may be an overestimate. The same applies to the higher use of hemostatic agents in the robotic-assisted MVr group, except for cellulose-based products. While limited studies have reported differences in the usage rates of hemostatic agents, this is a unique finding characteristic of research utilizing DPC database.
The length of hospital stay was statistically shorter in the robotic-assisted MVr group compared to the Mini-MVr group, despite similar medians: 10 days (IQR 9–13) vs. 10 days (IQR 7–14), respectively (P=0.04). This reflects a narrower distribution and fewer prolonged hospitalizations in the robotic group. These findings are consistent with prior reports indicating that the average length of stay after robotic-assisted MVr ranges from 3 to 7 days, whereas for sternotomy and minimally invasive approaches, it generally ranges from 5 to 11 days (10).
Postoperative costs were similar, suggesting that differences in major complications may have had a minimal impact. Notably, in-hospital mortality was 0 in the robotic-assisted MVr group and 7 out of 1,430 cases (0.5%) in the Mini-MVr group, indicating that both approaches are safe and feasible. Furthermore, robotic surgery may offer additional societal benefits, such as enabling earlier return to daily life and reducing the need for postoperative rehabilitation or convalescent care, with one report noting a transfer rate to such facilities as low as 2.9% (11). While the present analysis found that the operative day cost was higher in the robotic-assisted MVr group, the overall findings suggest that robotic surgery has the potential to contribute not only to patient safety but also to long-term healthcare efficiency, particularly if implemented in well-equipped, high-volume centers.
This study has several limitations inherent to the use of a nationwide administrative claims database.
First, although the DPC system aggregates costs on a fee-for-service basis, it does not necessarily reflect the actual reimbursed medical costs. Additional fee adjustments, such as surcharges for emergency nighttime procedures, are not fully captured in this dataset.
Second, our analysis did not account for the capital expenditures required for robotic surgery, including the cost of system acquisition, ongoing maintenance, and instruments used per case. These expenses are substantial and would likely further increase the total cost differential in favor of Mini-MVr. The absence of these factors may lead to an underestimation of the true financial burden associated with robotic-assisted MVr.
Third, robotic-assisted MVr was performed in a limited number of participating hospitals, potentially introducing selection bias and limiting the generalizability of the findings. Furthermore, all robotic-assisted MVr cases were concentrated in high-volume hospitals, which may limit the applicability of our findings to lower-volume hospitals. In this study, high-volume hospitals were defined by the number of beds; however, procedural volume, particularly mitral valve surgery caseload, is a more established predictor of outcomes. The absence of data on actual procedural volume represents an additional limitation, as it precludes more precise adjustment for institutional experience. Differences in surgical experience or institutional protocols could not be adjusted for, potentially affecting efficiency and cost profiles. Robotic-assisted MVr also has a steep learning curve, often requiring 50–100 cases for proficiency (12,13). Therefore, this nationwide cohort likely includes patients treated during the early adoption phase of robotic-assisted MVr, as the procedure only became reimbursed in Japan in 2018. This factor may have influenced perioperative outcomes, such as transfusion rates and the use of hemostatic agents.
Fourth, patient selection was based on the procedural code (specifically, “thoracoscopic valvuloplasty-single valve”), which did not allow us to differentiate between direct-vision, endoscopic-assisted, and totally endoscopic approaches within the Mini-MVr group. This lack of granularity may have introduced unrecognized heterogeneity, making it difficult to interpret the comparative effectiveness reliably. In addition, the procedural code did not permit exclusion of concomitant Maze procedures. Although we adjusted for this as a covariate in the cost analysis, residual confounding may persist.
Fifth, the database lacks detailed clinical information, including specific intraoperative details such as annuloplasty devices, cardiopulmonary bypass time, aortic cross-clamp time, and conversion to sternotomy. Furthermore, key echocardiographic parameters, such as pre- and postoperative MR grade, mitral valve pathology, and the success rate of repair techniques, could not be assessed. Patient-centered outcomes, such as quality of life, functional recovery, and time to return to work, are not captured. These limitations restrict the ability to perform a comprehensive and clinically meaningful comparison. Similarly, patient background was assessed using ICD-10 codes, which may not reliably capture the true medical history.
Finally, the database does not include post-discharge outcomes such as quality of life, time to return to work, or long-term clinical outcomes, which are important considerations when evaluating the overall cost-effectiveness and societal impact of robotic-assisted surgery.
Conclusions
Based on a nationwide claims database analysis, robotic-assisted MVr was associated with higher operative day costs and increased use of transfusions and hemostatic agents; however, total in-hospital costs and clinical outcomes, including mortality and length of stay, were comparable to those of Mini-MVr. These findings support the feasibility and safety of robotic-assisted MVr and suggest its potential to enhance healthcare efficiency when performed at high-volume centers. Future studies should investigate the relationship between surgical volume and cost-effectiveness to optimize the broader implementation of robotic cardiac surgery.
Acknowledgments
We gratefully acknowledge the Medical Data Vision Co., Ltd. (MDV, Tokyo, Japan) for providing access to the DPC database used in this study. Also, we would like to thank the department of medical statistics at Osaka Metropolitan University for their support in data access and analysis.
Footnote
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1086/rc
Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1086/dss
Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1086/prf
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1086/coif). The authors have no conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The Osaka Metropolitan University Institutional Review Board reviewed this study and determined that formal ethical approval was not required, because the data were anonymized by the database before the initiation of the study. Informed consent was waived in this retrospective study.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
References
- Otto CM, Nishimura RA, Bonow RO, et al. 2020 ACC/AHA Guideline for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2021;143:e72-e227. [Crossref] [PubMed]
- Paul S, Isaacs AJ, Jalbert J, et al. A population-based analysis of robotic-assisted mitral valve repair. Ann Thorac Surg 2015;99:1546-53. [Crossref] [PubMed]
- Hawkins RB, Mehaffey JH, Mullen MG, et al. A propensity matched analysis of robotic, minimally invasive, and conventional mitral valve surgery. Heart 2018;104:1970-5. [Crossref] [PubMed]
- Watanabe G, Kumamaru H, Kinukawa N, et al. Successful launch of robot-assisted mitral valve repair in Japan under universal health coverage. J Thorac Cardiovasc Surg 2025;169:1242-50. [Crossref] [PubMed]
- Mihaljevic T, Koprivanac M, Kelava M, et al. Value of robotically assisted surgery for mitral valve disease. JAMA Surg 2014;149:679-86. [Crossref] [PubMed]
- Moss E, Halkos ME. Cost effectiveness of robotic mitral valve surgery. Ann Cardiothorac Surg 2017;6:33-7. [Crossref] [PubMed]
- Hadaya J, Chervu NL, Ebrahimian S, et al. Clinical Outcomes and Costs of Robotic-assisted vs Conventional Mitral Valve Repair: A National Analysis. Ann Thorac Surg 2025;119:1011-9. [Crossref] [PubMed]
- Childers CP, Maggard-Gibbons M. Estimation of the Acquisition and Operating Costs for Robotic Surgery. JAMA 2018;320:835-6. [Crossref] [PubMed]
- Barbash GI, Glied SA. New technology and health care costs--the case of robot-assisted surgery. N Engl J Med 2010;363:701-4. [Crossref] [PubMed]
- Fatehi Hassanabad A, Nagase FNI, Basha AM, et al. A Systematic Review and Meta-Analysis of Robot-Assisted Mitral Valve Repair. Innovations (Phila) 2022;17:471-81. [Crossref] [PubMed]
- Murphy DA, Moss E, Binongo J, et al. The Expanding Role of Endoscopic Robotics in Mitral Valve Surgery: 1,257 Consecutive Procedures. Ann Thorac Surg 2015;100:1675-81; discussion 1681-2. [Crossref] [PubMed]
- Toolan C, Palmer K, Al-Rawi O, et al. Robotic mitral valve surgery: a review and tips for safely negotiating the learning curve. J Thorac Dis 2021;13:1971-81. [Crossref] [PubMed]
- Sandoval E, Bhoera RA, Tomšič A, et al. Learning curve of robotic mitral repair: prospective two-centre study of proficiency and clinical outcomes. Eur J Cardiothorac Surg 2024;66:ezae426. [Crossref] [PubMed]
