Cost analyses in randomized trials on robot-assisted surgery: systematic review

Abstract

Background

Robot-assisted surgery (RAS) is increasingly being used, yet its cost-effectiveness remains debated. Cost analyses of RAS are therefore essential but remain challenging. This systematic review evaluated the quality of cost analyses in randomized clinical trials (RCTs) comparing RAS with alternative surgical approaches.

Methods

A systematic review was performed in PubMed, EMBASE, Cochrane Library, and Web of Science from inception up to August 2025. RCTs were included if they compared RAS with other approaches and conducted a cost analysis. Risk of bias was assessed using the revised Cochrane Risk-of-Bias tool. The methodological quality and comprehensiveness of cost analyses were evaluated with the Economic Evaluation Bias Assessment Tool (ECOBIAS) and Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist, respectively. Studies were evaluated for key structural and component-specific costs of RAS.

Results

Overall, 38 RCTs involving 5832 patients were included. Most studies focused on general surgical procedures (20 RCTs, 53%), followed by urology (7), gynaecology (7), and cardiothoracic surgery (4). RAS was compared with laparoscopic surgery in 23 RCTs, open surgery in 14 RCTs, and another robotic modality in one RCT. Regarding bias, 10 RCTs (26%) were considered high risk, and 24 (63%) had some concerns. On average, RCTs met 5 of 11 ECOBIAS criteria and 14 of 28 CHEERS items. Only 15 of 38 RCTs (39%) included key structural costs of RAS, such as robot acquisition and maintenance, whereas 12 of 38 RCTs (32%) provided a component-specific cost overview. RAS was more expensive in 33 of 36 RCTs (92%).

Conclusion

Randomized trials comparing RAS with other surgical approaches rarely perform adequate cost assessments and cost-effectiveness analyses. The substantial risk of bias, methodological heterogeneity, and partial cost reporting observed underline the need for uniform economic evaluation in RCTs on RAS. Registration number: CRD42024520677 (https://www.crd.york.ac.uk/prospero/).

This systematic review of 38 randomized trials reveals that cost analyses in robot-assisted surgery are often incomplete and of poor methodological quality. Most studies failed to report key cost components, showed considerable risk of bias, and lacked standardized approaches. These findings highlight the need for standardized economic evaluation in trials of robot-assisted surgery.

Introduction

The utilization of robot-assisted surgery (RAS) has increased rapidly, offering possible technical advantages, such as three-dimensional imaging, tremor filtering, enhanced dexterity, and improved ergonomics^1,2. These features have the potential to improve surgical precision; however, evidence suggests they do not consistently result in improved clinical outcomes^3–6. Reported benefits, such as reduced blood loss, shorter hospital stays, and lower complication rates vary depending on the procedure and clinical context, but are often accompanied by longer operative times and higher costs^7–9.

Cost analyses of surgical procedures are challenging because of the difficulty in capturing all relevant costs and variations across countries, hospitals, and surgeons’ learning curves. For RAS, additional complexity arises from the high number of disposable materials, differing acquisition strategies, and the rapid evolution of systems and techniques^10–12.

The Idea, Development, Exploration, Assessment and Long-term monitoring (IDEAL) framework highlights the importance of robust and standardized methodologies for evaluating innovative surgical technologies, providing essential guidance for addressing these challenges¹³. In addition, initiatives such as the RoboCOS study have proposed a core outcome set to standardize evaluations of patient- and system-level impacts of RAS, including cost considerations¹⁴.

This systematic review aims to evaluate and investigate the methodology of cost analyses performed in randomized clinical trials (RCTs) comparing RAS to other approaches. The emphasis was on identifying factors influencing the quality and comprehensiveness of the cost analysis, using checklists and standardized evaluation criteria to highlight areas for improvement.

Methods

Search strategy and eligibility criteria

The systematic review was performed according to the PRISMA guidelines¹⁵ and the Cochrane Handbook for Systematic Reviews of Interventions¹⁶. A literature search was performed from the inception of the databases up to 1 August 2025 using the PubMed, EMBASE (OVID), Cochrane Library Central, and Web of Science databases. The search strategy was developed with the assistance of an expert librarian (FJ), and keywords related to ‘randomized controlled trials’ and ‘robot-assisted surgery’ were used; the full search is provided in the Supplementary material. The protocol of this systematic review was prospectively registered in PROSPERO (CRD42024520677).

RCTs that compared RAS to open surgery, laparoscopic surgery, or an alternative robot-assisted approach were included, provided they incorporated any type of cost analysis. In addition, separate cost analyses conducted alongside an RCT were considered. For the identified RCTs that did not include an explicit cost analysis, a supplementary search was conducted. This search involved examining both the citations and references of these RCTs to identify any related publications that addressed cost analyses.

Papers for which no full text was available, those published in a language other than English, animal studies, and non-clinical/simulation studies (for example, inanimate biotissue models) were excluded. In addition, RCTs involving single-arm robotic systems, as well as those focusing on orthopaedic and ear, nose, and throat indications, were excluded to minimize heterogeneity, because the robotic systems and procedures in these fields differ substantially.

Study selection

The selection of the studies was facilitated using the Rayyan.ai tool for systematic reviews¹⁷. Two reviewers (SB and RA) independently screened the titles and abstracts of all retrieved references. Following duplicate removal by an expert librarian, potentially eligible articles were identified and assessed for inclusion by full-text review. Any discrepancies in the identification of eligible studies were resolved by discussion and consensus. If needed, a third reviewer (JH) was consulted to resolve any remaining disagreement.

Data extraction

Data were extracted using standardized data extraction forms. Data were extracted from the included RCTs and, in cases where a separate cost analysis was conducted, relevant data were obtained from both the cost analysis publication and the primary RCT publication. The following data were extracted: study characteristics (first author and year of publication, country, study period, surgical speciality, type of procedure, comparison, number of patients, primary outcomes, and main results); time horizon (the duration over which the study assessed outcomes and costs); source of costs (the sources used to collect cost data, such as hospital records or questionnaires); currency; total cost of the robot-assisted and compared procedures, and the difference (delta) between these total costs; the reported variable cost components included in the total cost calculation; and the micro-costing for the robotic system. Costs were converted into Euros (€) for January 2025, with currency conversion conducted via the European Central Bank¹⁸. Inflation adjustments were calculated using the Eurostat Harmonised Index of Consumer Prices¹⁹ based on year of publication to ensure that the values are comparable across time.

Outcomes of interest

The primary outcome was the methodological quality of cost analyses, assessed with the Economic Evaluation Bias Assessment (ECOBIAS)²⁰ tool and the Consolidated Health Economic Evaluation Reporting Standards (CHEERS)^21,22 checklist. The secondary outcome was the reported costs, compared between robot-assisted and alternative procedures using standardized cost ratios.

Quality assessment

The risk of bias of the RCTs was evaluated independently by two reviewers (SB and RA) using the revised Cochrane risk-of-bias tool for randomized trials (RoB 2)²³. Each bias domain was examined and subsequently categorized as having a low risk of bias, some concerns, or a high risk of bias. A third reviewer (JH) was consulted to resolve discrepancies.

ECOBIAS was used to evaluate the methodological facets of the economic evaluations along with the risk of bias²⁰. ECOBIAS comprises a 22-item checklist and was developed through the combination of biases identified in both model- and trial-based studies, with the aim of helping researchers reduce biases in planning and conducting economic evaluations. The ECOBIAS checklist comprises 11 general items for assessing bias in cost evaluations and 11 model-specific components, assessed as yes, no, partially, unclear, and not applicable.

The CHEERS reporting guideline was used to assess the comprehensiveness of reporting. The CHEERS reporting guideline comprises 28 items designed to optimize transparency in health economic evaluations^21,22. The recommendations were converted into questions (Table S1) and each item was assessed as yes (appropriately addressed), partially (addressed to some extent), or no (not addressed).

For the assessment, the study protocol was consulted; for RCTs with a separate cost analysis, both the primary RCT publication and protocol were used for assessment.

Statistical analysis

Categorical variables are presented as frequencies and percentages. Statistical analyses were performed using SPSS^® version 28 (IBM, Armonk, NY, USA). The cost comparison between the robot-assisted, laparoscopic, and open procedures was evaluated by calculating the cost ratio, defined as the cost of the robot-assisted procedure divided by that of the alternative procedure. This metric provides a standardized comparison of cost ratios across studies, accounting for variations in currency and year. A ratio of 1 indicates cost parity between procedures, a ratio > 1 suggests higher costs for the robot-assisted procedure, and a ratio < 1 indicates lower costs.

Results

Study selection

The initial search yielded 15 751 studies across the four databases. After excluding duplicates and screening titles and abstracts, 327 full-text studies were assessed. Of these, 38 RCTs were eligible, encompassing a total of 5832 patients. The study selection process is outlined in Fig. 1.

Fig. 1.

PRISMA flow diagram of study selection

Study characteristics

The characteristics of the included studies are summarized in Table 1. The RCTs were published between 2004 and 2024, and the global distribution of the RCTs is shown in Fig. 2. The surgical fields in the 38 RCTs included general surgery in 20 studies, specifically upper gastrointestinal surgery (4 studies), hepatopancreatobiliary surgery (4), colorectal surgery (6), and 6 hernia repair; the other specialties included urology (7), gynaecology (7), and cardiothoracic surgery (4). In 23 RCTs, RAS was compared to laparoscopic surgery^{24,25,31,32,36,37,40,41,43,45–47,49,51,52,54,58–60,63–65} and 13 RCTs compared RAS to open surgery^{26,27,29,30,33,35,44,55,56,61,62,67,68}. One study⁵³ compared two distinct robot-assisted procedures, namely the robotic intraperitoneal-only mesh procedure versus the robotic enhanced-view totally extraperitoneal procedure for ventral hernias.

Table 1.

General characteristics of the included studies

Study	Country	Type of procedure	Comparison	Total (n)	Robot (n)	Primary outcome	Main result‡	Cost favourable approach
Ma et al.24 (2020)*	China	Adrenalectomy	R versus L	140	70	Operative time	Less blood loss and lower operative time for R	L
Morino et al.25 (2004)*	Italy	Adrenalectomy	R versus L	22	10	Feasibility and clinical results	Less morbidity for L and superiority in terms of feasibility	L
Bochner et al.26 (2015)*	USA	Cystectomy	R versus O	124	60	Complications	Similar complication rates	O
Dixon et al.27 (2023)†	UK	Cystectomy	R versus O	305	157	Days alive (RCT)28, cost-effectiveness (CA)27	Increase in days alive for R	R (cost-effective)
Mastroianni et al.29 (2022)*	Italy	Cystectomy	R versus O	116	58	Transfusion rate	Reduced transfusion rate for R	O
Mastroianni et al.30 (2024)*	Italy	Cystectomy	R versus O	116	58	Transfusion rate	Reduced transfusion rate for R	O
Silay et al.31 (2020)*	Turkey	Pyeloplasty	R versus L	53	26	Anterior–posterior diameter	Comparable success and complications	L
Lönnerfors et al.32 (2015)*	Sweden	Hysterectomy	R versus L	122	61	Hospital cost	Higher cost for R	L
Lundin et al.33 (2020)†	Sweden	Hysterectomy	R versus O	50	25	Quality of life (RCT)34, total cost (CA)33	Faster recovery for R	O
Salehi et al.35 (2017)*	Sweden	Hysterectomy	R versus O	120	60	Para-aortic LN count	Non-inferiority in LN count and shorter LOS for R	R
Silva E Silva et al.36 (2018)*	Brazil	Hysterectomy	R versus L	89	44	Clinical outcomes	Equivalent morbidity	L
Vuorinen et al.37 (2017)†	Finland	Hysterectomy	R versus L	101	50	Operative time (RCT)38, cost (CA)37	Shorter operative time for R	L
Anger et al.39 (2014)*	USA	Sacrocolpopexy	R versus L	78	40	Total costs	Higher costs for R procedures	L
Paraiso et al.40 (2011)*	USA	Sacrocolpopexy	R versus L	78	40	Operative time	Longer operative time and increased pain for R	L
Flemming et al.41 (2023)†	France	Colectomy	R versus L	127	43	LOS (RCT)42, clinical and operative outcomes (PH)41	Shorter LOS for R, but comparable short-term outcomes	L
Park et al.43 (2012)*	Korea	Colectomy	R versus L	71	35	LOS	Similar LOS for R	L
Chang et al.44 (2023)*	China	Rectal resection	R versus O	171	86	Complications	Fewer complications for R	O
Feng et al.45 (2022)†	China	Rectal resection	R versus L	1240	620	Recurrence rate (RCT), Short-term outcomes (PR)31	Better oncological quality for R	L
Jayne et al.46 (2017)*	UK	Rectal resection	R versus L	471	237	Conversion	Similar conversion rate for R	L
Mäkelä-Kaikkonen et al.47 (2019)†	Finland	Rectopexy	R versus L	30	16	Postoperative changes (RCT)48, healthcare cost (CA)47	Similar outcomes	L (+ R cost-effective)
Costa et al.49 (2022)†	Brazil	Hernia repair	R versus L	37	18	Short- and long-term outcomes (RCT)50, cost (CA)49	Similar outcomes	L
Olavarria et al.51 (2020)*	USA	Hernia repair	R versus L	124	65	LOS	Similar LOS, increased operation duration for R	L
Petro et al.52 (2020)*	USA	Hernia repair	R versus L	81	39	Pain	Comparable pain and higher operative time for R	L
Petro et al.53 (2023)*	USA	Hernia repair	R versus eTEP	100	49	Pain	Similar pain	Comparable
Prabhu et al.54 (2020)*	USA	Hernia repair	R versus L	102	48	Postoperative outcomes	Similar clinical outcomes and higher operative time for R	L
Warren et al.55 (2024)*	USA	Hernia repair	R versus O	90	46	Postoperative outcomes	No difference in primary composite outcome	O
Goense et al.56 (2023)†	Netherlands	Oesophagectomy	R versus O	112	54	Complications (RCT)57, hospital cost (CA)56	Fewer postoperative complications for R	Comparable
Morino et al.58 (2006)*	Italy	Fundoplication	R versus L	50	25	In-hospital cost	Comparable feasibility and outcomes	L
Müller-Stich et al.59 (2007)*	Germany	Fundoplication	R versus L	40	20	Perioperative outcomes	Similar short-term outcomes	L
Lu et al.60 (2021)*	China	Gastrectomy	R versus L	300	150	Survival rate	Faster recover and reduced morbidity for R	L
Huang et al.61 (2019)*	China	Lobectomy	R versus O	113	58	Short-term outcomes	Less operative blood loss for R	O
Huang et al.62 (2021)*	China	Lobectomy	R versus O	148	76	Survival rate	Similar survival, reduced bleeding and pain for R	O
Jin et al.63 (2022)*	China	Lobectomy	R versus L	320	157	Overall survival	Similar perioperative outcomes, higher LN yield for R	L
Patel et al.64 (2023)*	Canada	Lobectomy	R versus L	186	81	Health utility score	Comparable Health utility score	R (cost-effective)
Grochola et al.65 (2019)*	Switzerland	Cholecystectomy	R versus L	60	30	Surgeon's stress load	Reduction of stress load for R	L
Chen et al.66 (2017)*	China	Pancreatectomy	R versus O	100	50	LOS	Shorter LOS for R	O
Klotz et al.67 (2024)*	Germany	Pancreatoduodenectomy	R versus O	81	41	Morbidity	Comparable morbidity	O
Liu et al.68 (2024)*	China	Pancreatoduodenectomy	R versus O	164	82	LOS	Shorter LOS for R	O

^*RCT conducting a cost analysis. †Separate cost analysis. ‡Main result in favour of robot-assisted surgery: lower R cost or cost-effectiveness; main result in favour of the comparative procedure: lower cost for the comparative procedure; no significant difference: comparable cost between robot-assisted surgery and the comparative procedure. R, robot-assisted surgery; L, laparoscopic surgery; O, open surgery; eTEP, enhanced-view totally extraperitoneal; RCT, randomized clinical trial; CA, cost analysis; PH, post hoc analysis; PR, preliminary results; LN, lymph node; LOS, length of hospital stay.

Fig. 2.

Global distribution of RCTs included in this study

RCTs, randomized clinical trials.

In 30 of the 38 included RCTs, the cost analysis was included within the same publication that reported the trial’s primary outcomes^{24–26,29–32,35,36,39,40,43,44,46,51–55,58–68}. In three RCTs^32,39,58, costs were actually the primary outcome of the conducted trial. In addition, 8 RCTs^{27,33,37,41,45,47,49,56} published a separate cost analysis, which was eventually included.

Risk of bias

The risk of bias assessment of RCT quality was visualized using Robvis⁶⁹ (Fig. 3). Of the 38 RCTs, four^40,41,51,64 were at low risk of bias and 10 RCTs^{25,29–31,33,35,54,55,61,67} were considered to have a high risk of bias. The remaining 24 RCTs were classified as having some concerns regarding bias.

Fig. 3.

Risk of bias assessment

The risk of bias for each of the studies included was determined using the revised Cochrane risk-of-bias tool for randomized trials in different domains. D1, bias arising from the randomization process; D2, bias due to deviations from the intended interventions; D3, bias due to missing outcome data; D4, bias in measurement of outcomes; D5, bias in selection of the reported result.

Outcome of individual RCTs

The most frequently used (5 of 38 studies) primary outcome was length of hospital stay (LOS)^{41,43,51,66,68}, followed by survival rate (4 of 38)^27,60,62,63, operative time (3 of 38)^24,37,40, complications (3 of 38)^26,44,56, and cost (3 of 37)^32,39,58 (Table 1). The primary outcome favoured RAS in 42% of RCTs (16 of 38) and favoured the comparator in 11% of RCTs (4 of 38). The remaining 47% RCTs (18 of 38) reported no significant differences.

Quality and comprehensiveness of the cost analyses

On average, studies only met 5 of the 11 general ECOBIAS items (45%; standard deviation (s.d.) 1.97)²¹. Among the identified sources of bias, the most prevalent were narrow perspective bias (limiting the analysis to a specific payer or hospital perspective instead of including a broader societal view; 27 of 38 studies^{24,25,29–31,35–37,41,43–45,52–55,58,59,61–68}), inappropriate discounting bias (incorrect application of discount rates over time; 36 of 38 studies^{24–26,29–33,35–37,39–41,43–46,49,51–56,58–68}), limited sensitivity analysis bias (insufficient assessment of uncertainty in key parameters; 25 of 38 studies^{24,26,29,31,32,35–37,41,43,44,49,52–55,58,59,61–67}), and bias related to internal consistency (lack of internal validation of model assumptions and calculations; nine of nine modelling studies^{26,27,32,33,35,37,47,56,64}). Assessment of intermittent data collection bias proved challenging, resulting in its classification as unclear in some cases. Detailed information on the ECOBIAS assessments is provided in Table S2.

Only one study²⁷ explicitly referred to the CHEERS guidelines^22,23 in their method. The average implementation rate of the CHEERS items was 14 of 28 (50%; s.d. 4.90), ranging from 5 of 28 items reported (18%) to 25 of 28 items reported (89%). Detailed information on the CHEERS assessment is provided in Table S3.

Of the 38 RCTs, eight^{27,33,36,37,47,49,56,64} mentioned costs in their title. In addition, 10 of 38 RCTs^{27,32,33,37,39,47,49,53,56,64} explicitly described aspects of the cost analysis methodology in the abstract. In all, 15 RCTs (39%)^{27,29,30,32,33,35,37,39,41,43,45,47,49,56,65} accurately addressed the measurement and valuation of resources and costs item from the CHEERS guidelines. These RCTs accounted for the purchase and maintenance of the robotic system and properly reported all cost variables included in their cost analysis. Only 12 of the 38 RCTs (32%)^{29,30,32,33,36,37,39,41,47,49,56,65} provided a summary of the main results from micro-costing, as recommended by the CHEERS checklist, clearly quantifying key cost components per patient. None of the RCTs described the items ‘approach to engagement with patients’ and ‘effect of engagement’.

Furthermore, the study findings, incorporation considerations, effects, and limitations regarding the cost analyses were comprehensively discussed in the discussion sections of 14 of 38 RCTs^{27,32,33,35–37,39,40,43,46,49,56,60}.

A speciality-stratified summary of CHEERS and ECOBIAS adherence showed that gynaecology trials had the highest overall adherence to both checklists, whereas cardiothoracic trials had the lowest. Full details are presented in Tables S4 and S5.

Cost outcomes and cost-effectiveness

Of the 38 RCTs included in this study, 36 reported the total cost of RAS (Table S6). Most (33 of 36, 92%) of these RCTs found that RAS was more expensive than the comparison approach. Only one study³⁵, a model-based study, reported lower costs for RAS compared with open surgery, primarily due to shorter LOS and reduced analgesia use. Two RCTs^53,56 reported comparable total costs for RAS and comparison approaches. Another RCT⁵³ compared two robotic devices, whereas an RCT investigating oesophagectomies⁵⁶ explained this equivalence because of lower complication rates in the RAS group, which offset the higher initial cost of the robotic procedures.

Furthermore, three RCTs^27,47,64 conducted cost-effectiveness analyses using quality-adjusted life years (QALYs) as metric. Using extrapolated utilities, Dixon et al.²⁷ found that RAS cystectomy had an incremental cost-effectiveness ratio of £100 008 (US $144 312) per QALY gained, with higher cost-effectiveness in subgroups defined by age, tumour stage, and performance status. Another study⁴⁷ found that although robotic ventral mesh rectopexy had higher initial costs, it became cost-effective over a 5-year period, with a favourable incremental cost-effectiveness ratio of €16 707 per QALY gained, although the authors of that study also noted that RAS was not cost-effective in the short term. A different study⁶⁴ found that robot-assisted lobectomy is cost-effective, with an incremental cost-effectiveness ratio of US $14 925.62 per QALY gained compared with video-assisted thoracic surgery lobectomy.

Cost variables and acquisition cost

The variables included in the total cost calculations were extracted for each study, with their frequencies presented in Fig. 4. Acquisition costs of the robotic system were explicitly included in 42% of RCTs (16 of 38)^{26,27,29,30,32,33,35,37,39,41,43,47,56,60,63}. Detailed cost variables used in each study are listed in Table S7.

Fig. 4.

Reported variables used for the total cost calculation

An estimate of the robotic system purchase cost per patient was mentioned in 11 of 38 RCTs. These estimates were derived from factors such as acquisition costs, maintenance, resale value, and system utilization^{27,29,30,32,33,35,39,46,49}. The amounts were converted to euro (as of January 2025) and can be found in Fig. S1, ranging from €985³³ to €8576³⁰. However, in three of these RCTs, the robotic system acquisition costs were excluded from the total cost calculations^25,46,65. Additional details, including the original cost estimates and their description, are available in Table S8.

Cost comparison and quality assessment findings

The cost ratios between the compared modalities in each study are illustrated in Fig. 5. Trials that did not report total costs in their analyses^27,64 or lacked comparative cost data between surgical modalities⁵³ were excluded from the cost ratio analysis. In one study⁶⁶, a separate cost analysis was conducted for radical cystectomy of the neobladder and ileal conduit; therefore, two different ratios were calculated. Another study³⁵ was the only study that reported that the total cost of RAS was lower than that of open surgery. The cost ratios for the remaining 35 studies exceeded 1. In Fig. 5, studies that explicitly stated inclusion of robotic system acquisition costs in their total cost calculations and those that did not are shown separately; cost ratios were similar between the two groups.

Fig. 5.

Overview of the calculated cost ratios

a Cost ratio for RAS versus open surgery. b Cost ratio for RAS versus laparoscopic surgery. Cost ratios < 1 favour RAS; cost ratios > 1 favour open (a) or laparoscopic (b) surgery. RAS, robot-assisted surgery.

Discussion

This systematic review included 38 RCTs that evaluated the cost of RAS versus alternative surgical modalities. The findings indicated substantial methodological variability, with a lack of comprehensiveness and moderate reporting quality, as assessed by CHEERS and ECOBIAS. Key cost factors, such as robotic system acquisition, maintenance, patient complications, and patient sick leave, were detailed in only 38% of RCTs, and only 32% provided a summary quantifying these costs.

The findings of the present study align with those of a 2018 systematic review that evaluated the cost methodology for studies on gynaecological RAS. In that 7-year-old study⁷⁰, similar issues with low methodological quality were identified, particularly the lack of detailed cost data and the inconsistent inclusion of key factors, such as acquisition and maintenance costs of robotic systems and the use of surgical equipment being rarely reported. However, their focus was limited to gynaecological RAS and included various study designs, such as retrospective and case-control studies, which limited direct comparability.

An important observation in this study was the lack of transparency regarding the inclusion of robotic system acquisition costs. It is expected that some RCTs did include robot acquisition costs but, contrary to the guidelines, did not report this explicitly. This under-reporting likely influenced the cost ratios calculated.

Furthermore, owing to the high heterogeneity in cost variables used to calculate the total costs, generating a forest plot was not feasible. The RCTs calculated the total costs of different types of operations, each varying in operative time, staff required, and LOS, making direct comparison difficult. In addition, many studies lack currency standardization and inflation adjustments, further complicating direct comparisons. These methodological inconsistencies undermine the potential to draw reliable conclusions regarding the cost-effectiveness of RAS. However, despite the low methodological quality of the included cost analyses, the results were analysed (Fig. 5) to identify inconsistencies and variations in cost reporting.

Of the 36 RCTs reporting total costs, 92% reported higher costs for RAS. The RCT on para-aortic lymphadenectomy³⁵ was the only study to report lower costs for RAS, attributing this to shorter hospital stays and reduced postoperative care despite higher procedural costs and longer operation times after RAS for endometrial cancer with lymphadenectomy. However, the findings of that study³⁵ relied on modelled estimates that lacked a clear explanation, thus limiting generalizability. Notably, only three studies performed a cost-effectiveness analysis^27,47,64 despite its critical role in clinical decision-making, raising concerns about potential selective reporting. Moreover, the quality of the analyses conducted on the included studies is debatable.

The economic impact of RAS is highly dependent on factors such as the surgical volume, hospital type, country, and procedure. High-volume centres achieve economies of scale by spreading fixed costs over a larger number of procedures, leading to lower per-procedure costs. Consequently, these factors affect the generalizability of cost analyses, necessitating careful consideration when extrapolating the findings to different settings. Although these factors influence the cost-effectiveness of RAS, they do not necessarily offset the cost disparity with the comparator approach. In addition, although there are efficiency gains with RAS over time through the learning curve, potential cost-mitigating factors, such as reduced operating times, improved resource utilization, and lower staffing requirements, may contribute to cost reduction⁷¹. However, modelling studies suggest that these factors alone are unlikely to resolve the cost disparity⁷². Nonetheless, achieving true cost-effectiveness requires more than just cost reductions, necessitating a demonstrable improvement in clinical outcomes. Although RAS offers technical advantages, current evidence does not yet show substantial differences in clinical effectiveness versus conventional approaches^1–6. Moreover, it is important to consider whether current cost-effectiveness studies truly capture the full value of RAS. In addition to clinical and economic outcomes, future evaluations may benefit from incorporating broader factors, such as surgeon ergonomics, long-term workforce sustainability, and the integration of high-level performance metrics and AI-driven insights. Although cost is frequently discussed in the context of RAS, it should be recognized as only one of many relevant outcomes. Furthermore, although this study selected only RCTs based on the expectation of higher methodological quality and greater adherence to cost-reporting guidelines, it was found that even these trials frequently failed to report key factors, such as surgical volume and other cost-related perspectives. In addition, the exclusive inclusion of RCTs in this systematic review, as well as in future analyses, limits the real-world applicability of the findings because these trials are often conducted in high-volume or otherwise idealized settings. This further underscores the need to adopt standardized reporting frameworks that include essential cost-related factors.

This systematic review has several limitations that should be considered. First, only articles published in English were included, which may have introduced a language bias by excluding potentially relevant studies published in other languages. Second, although research suggests that RAS may result in shorter hospital stays, fewer complications, and reduced rehabilitation needs^7,8, these factors are often not explicitly included in the total cost calculation. This limitation is frequently mentioned in the discussion section of included studies. Third, the terminology used to describe the costs in the total cost calculation was inconsistent. For instance, some RCTs referred to ‘consumables’ as ‘disposables’, whereas others used the same term to denote ‘reusable equipment’. Furthermore, some RCTs provided detailed descriptions of the variable costs, whereas others offered only a summary or no explanation. Consequently, it is likely that some RCTs included additional variable costs that were not reported. Many studies have not clarified the sources and perspectives of their cost calculations, often considering only the hospital’s viewpoint. Finally, critical micro-costing results underlying the total costs were frequently absent, resulting in a lack of transparency regarding these cost drivers in the analyses.

Notably, none of the studies addressed environmental considerations related to RAS. Environmental sustainability is an increasingly recognized factor in healthcare cost evaluations, because resource-intensive technologies such as RAS contribute to elevated greenhouse gas emissions, increased waste production, and higher energy consumption. These factors not only have ethical and ecological implications but also translate into direct and indirect financial burdens, including waste management costs, energy expenditure, and potential regulatory requirements for reductions in the carbon footprint. Nevertheless, with the possible benefits of RAS, this presents an ethical dilemma for surgeons^13,73. This area of research is still relatively novel, but it is expected that the environmental impact will receive increasing attention in future economic evaluations of surgical innovations.

The strength of the present systematic review lies in its unique evaluation of cost analysis methodology across various surgical specialities, incorporating a bias assessment, quality evaluation, and completeness assessment tool, while exclusively including RCTs.

Based on established guidelines, the literature suggests that several key components are essential for conducting a complete cost analysis of RAS^74,75. A comprehensive approach should include both hospital and societal perspectives, capturing direct costs such as equipment, hospitalization, and surgical expenses, as well as indirect costs, including social impact, fast recovery, and informal care costs. Explicitly defining these perspectives is crucial for transparency and comparability⁷⁶. In addition, an adequate time horizon should be applied, extending from the patient’s admission to their return to work^23,74,76. The purchase and maintenance costs of the robotic system, as well as intervention-specific costs for the comparator, such as the laparoscopic tower, should be included and accurately documented^76,77. Furthermore, the calculation methods and assumptions used to determine these costs, as well as transparency in reporting all calculated micro-costing, are essential for clarity and reproducibility, although collecting such detailed data can be both costly and time-consuming^23,76. Although adherence to the CHEERS checklist is essential to ensure high-quality economic evaluations, it is not always feasible to comply with all items. In particular, capturing long-term follow-up costs, societal productivity losses, or informal care is often impractical due to logistical and methodological constraints. These limitations should nonetheless be reported transparently, so that readers and health economists can interpret the findings with appropriate caution and adjust for missing data where relevant. As an alternative, authors may consider using model-based approaches to estimate these types of outcomes.

To ensure meaningful interpretation and application, it is critical to transparently document contextual information, including country, hospital type, surgical volume, and the experience of the surgeons and hospitals with robotic systems²³. Financial parameters, such as currency, inflation adjustments, and discount rates, must also be explicitly reported to enable direct extrapolation⁷⁶. Finally, integrating environmental impact assessments will be pivotal in providing a holistic view of the implications of adopting RAS technologies⁷³. Consideration of all these factors is essential for generating robust evidence to support well-informed clinical and policy decisions regarding the feasibility and cost-effectiveness of RAS.

This systematic review of randomized trials underscores the substantial methodological variability, generally low quality, and incomplete reporting in the cost analyses of RAS, limiting the reliability and extrapolation to healthcare policy, clinical decision-making, and cost-effectiveness assessments.

Future cost analyses should adhere to standardized methodologies, such as CHEERS and the IDEAL framework, while incorporating established guidelines, such as the HTA Core Model^®74, to ensure consistency and comparability. Collaboration with health economists or health technology assessment specialists is strongly recommended to enhance the reliability and reproducibility of economic evaluations in RAS.

Funding

The authors have no funding to declare.

Author contributions

Sterre Bosscha (Data curation, Formal analysis, Methodology, Writing—original draft), Rawin Amiri (Conceptualization, Formal analysis, Writing—original draft), Faridi Van Etten-Jamaludin (Methodology), Maroeska Rovers (Methodology, Formal analysis), Marc Besselink (Supervision, Writing—review & editing), and Jony van Hilst (Conceptualization, Methodology, Supervision, Writing—review & editing)

Disclosure

M.G.B. is a proctor for Intuitive Surgical robotic pancreatic surgery and has received grants from Intuitive Surgical for investigator-initiated research projects. The other authors declare no conflicts of interest.

Supplementary material

Supplementary material is available at BJS Open online.

Data availability

Data are available for bona fide researchers upon reasonable request.