Background
Radical cystectomy is a standard surgical technique for non-metastatic muscle-invasive bladder carcinoma [
1]. Open radical cystectomy (ORC) has been the gold standard treatment method, while laparoscopic radical cystectomy (LRC) has also been used. The safety and efficacy of LRC have been well-documented [
2]. Recently, robot-assisted radical cystectomy (RARC) has become increasingly common [
3,
4]. The safety and efficacy of RARC compared to those of ORC have also been reported [
4‐
6]. In addition, previous randomized controlled trials (RCTs) have reported no significant differences in 2-year progression-free survival rates [
7] and quality of life (QOL) scores between ORC and RARC [
8,
9]. Comparing the perioperative outcomes of ORC and RARC, Tang et al. [
10] conducted a meta-analysis of several RCTs and reported significantly lower estimated blood loss (EBL), lower transfusion rates, longer operative times, and larger quantities of anesthesia used with RARC. Although surgical outcomes are important, cost-effectiveness is also of great significance if RARC is to be widely adopted [
11]. In addition, as bladder carcinoma is reported to have the highest lifetime treatment costs per patient among all malignancies [
12], the cost-effectiveness of bladder carcinoma treatments needs to be evaluated. Smith et al. conducted a cost analysis between ORC and RARC [
11], and the results show a total cost advantage of $1630/case for RARC. In contrast, the cost-analysis by Bansal et al. showed a total cost advantage of approximately $1945/case for ORC [
13]. Michels et al. [
14] conducted a cost simulation of ORC and RARC and found that RARC costs €3365 more than ORC at 30 days. These studies indicated that operative time, length of stay (LOS), and the number of annual cases were key drivers of costs [
14]. However, consensus regarding the most cost-effective surgical approach is yet to be reached. Although some reviews have been published on cost comparisons between surgical procedures [
15,
16], they do not clarify the cost structure or focus on the total cost and not on segmental costs or cost-effective measures for robotic surgery. For hospitals, identification of the cost components that influence the total cost is crucial to make RARC more cost-effective. Therefore, we conducted a systematic review on the segmental costs of ORC, LRC, and RARC. This study aimed to provide cost-effectiveness data for ORC, LRC, and RARC and provide insights for the effective management of treatments and applicability of RARC for patients with bladder cancer. This study also aimed to clarify the current available knowledge to identify any gaps in order to promote future research.
Discussion
Only one of the included single-institutional studies comparing ORC and RARC was an RCT while the other four were retrospective studies. In addition, the “risk of bias” evaluation showed that some studies could have “High risk” of “Selective reporting” as these studies did not report parts of the primary outcomes. Therefore, although these studies are important due to the lack of large studies comparing surgical procedures, more evidence from high quality studies (e.g. RCT) is required. Nevertheless, the results from single-institutional studies can provide insights on cost structures through interpretation with the included large database studies, which reflect the general cost trends.
Two single-institutional retrospective studies reported total cost advantages for RARC over ORC [
21,
22]; another single-institutional retrospective study, one RCT, and four database studies showed total cost advantages for ORC [
13,
22‐
26]. Considering the quality of the studies, the results indicated that, in general, RARC was more likely to be expensive. Michels et al. conducted a cost simulation of ORC and RARC using data from a literature review and reported that RARC was more expensive [
14], similar to our study result. Leow et al. reported that the cost advantage in ORC was due to the additional costs of purchasing and maintaining robots and longer operative times for RARC [
24,
28]. However, on the other hand, Martin et al. reported that the cost advantage in RARC was due to the lower complication rates [
21]. These conflicting findings indicate that RARC cost-effectiveness was institution-dependent. Leow et al. reported that the surgical approach (robot-assisted vs open) was neither a major factor on cost variations nor associated with high costs [
28]. Therefore, focusing on the segmental costs, such as operating costs and robot costs, is necessary to figure out how each aspect contributes to the total cost. The results of segmental costs from this study are of great importance for improving cost-effectiveness of RARC compared to that of ORC, from the hospital’s perspective. Per-case robot costs were calculated by amortizing the robot-related costs and dividing the costs by the annual number of cases in the subject hospitals, including cases of other surgeries such as prostatectomy. According to a previous study, robot equipment costs were attributed to higher costs in RARC [
21]. Although robots were an expensive initial investment for an institution, the per-case robot-related costs accounted for only 4.8–15.8% of the total costs (Table
3) [
11,
13,
22] because the included single-institutional studies were conducted in high volume hospitals (288–400 cases per year). On the other hand, sensitivity analysis (Fig.
4) revealed that per-case robot-related costs differed greatly depending on the annual number of cases. High-volume centers were more likely to have lower per-case robot-related costs [
24], while these costs tended to be higher in low-volume institutions.
The complication costs were higher for ORC in three single-institutional retrospective studies [
13,
21,
22], accounting for 1.9% [
13] and 16.7% [
22] of the total costs. On the other hand, Bochner et al. showed in an RCT that there was no significant difference in perioperative complication rates between ORC and RARC, although the research did not report segmental costs for complications [
23]. A previous meta-analysis of four RCTs by Tang et al. revealed no significant differences between groups in the occurrence rates of patients with Clavien-Dindo grade 2–5 or 3–5 [
10]; therefore, complication costs may not differ between groups. Although complication costs differ with Clavien-Dindo grades, none of the included studies showed complication costs according to the Clavien grade at 90 days, which is a standard method of reporting postoperative complications. Further studies on complication costs are required with a high-level of evidence. Additionally, it is recommended that future studies focus on complication costs according to the Clavien-Dindo grade.
Complication costs can differ among countries. A multi-institutional study by Osawa et al. reported that causes of complications differed between the USA and Japan [
30]. International comparisons of complication costs also need to be conducted carefully.
Differences in transfusion rates between ORC and RARC have been reported in various studies [
10]. However, our results showed that even though transfusion rates were clinically essential, the difference did not largely affect the total costs.
The operating costs of RARC were higher in all studies due to longer operative times, which was similar to previous reports [
10]. Operating costs accounted for approximately 63.1–70.5% of the total RARC cost (Table
5 and Fig.
3). Most of the operating time costs were attributed to operating room occupation and surgeon fees which are dependent on the operative time. Therefore, if an institution succeeds in shortening the operating time, it would effectively reduce the total cost. Operative time has been reported to decrease significantly with increased surgeon experience [
31] and hospital volume [
29]. Leow et al. reported that although total costs were significantly higher in the RARC group, the difference did not exist between high-volume surgeons (≧7 cases per year in their study) and hospitals (≧19 cases per year) [
24]. Large institutions can benefit from shorter operative times, lower complication rates, cheaper per-case robot costs, and therefore, achieve more cost-effective RARC. Patient centralization to high-volume centers has been suggested as an effective way for cost-effective RARC surgeries [
24], which is supported by our results. However, further research is required to reveal the relationship between a surgeons’ learning curve and cost-effectiveness of RARC.
Most studies were analyzed using operative time [
13,
21,
22], except which used operating room occupancy time (utilization time) [
11]. Operating rooms are essential for hospital profitability and thus, longer operating room occupation is associated with higher costs. For accurate cost estimations, it is recommended that these two parameters should be recorded and analyzed.
Urinary diversion types chosen can also influence the total cost. The results by Lee et al. showed that RARC was more cost-efficient for ileal conduit ($4846), while the cost benefit diminished for cutaneous continent diversion ($609), and was absent for orthotopic neobladder (−$1966; Table
2) [
22]. This is one of the few studies that compare ORC with RARC by urinary diversion types. Current evidence on the impact of urinary diversion types on the costs is inadequate.
Recently, intracorporeal urinary diversions have become increasingly common [
32]. Only one article included in this study was conducted with an intracorporeal urinary diversion [
13]. Further studies are necessary to evaluate whether an intracorporeal or extracorporeal urinary diversion can influence operative time, and subsequently, the total costs.
Lymph node dissections differ depending on the surgeon and institution. Bochner et al. included patients who had undergone standard or extended lymph node dissections. In their RCT, the rates of patients who underwent extended dissections were matched between the ORC and RARC groups [
23]. Lymph node dissections should be included when comparing ORC and RARC costs because extended dissections can lead to longer operating times and higher costs.
Three retrospective studies reported lower LOS costs for RARC. LOS costs also accounted for a large proportion of the total costs, following operating costs. However, Tang et al. conducted a meta-analysis of four RCTs and found no significant difference in LOS for ORC and RARC [
10]. Of the studies included in this study, an RCT by Bochner et al. reported no significant difference of LOS between ORC and RARC. A database study by Leow et al. reported that while LOS differed significantly between ORC and RARC, LOS-related costs did not differ since most of the costs were due to surgery and intensive care unit admission [
24]. Therefore, it is likely that LOS costs do not necessarily have a large impact on the differences between ORC and RARC costs.
LOS differs greatly between countries. Sugihara et al. compared the LOS of radical cystectomy patients in Japan and the USA and reported shorter durations in the USA (8 [7–11] days vs 32 [21–44] days) [
33]. The effects of LOS changes should be considered with each country’s healthcare system [
34].
Few studies mention the perioperative protocols used in their studies (e.g. ERAS protocols), making it difficult to compare results between studies. Nabhani et al. reported that using the ERAS protocol led to a cost saving of $4488/procedure [
35]. Future studies are expected to be conducted under standardized protocols for more generalizable results.
One study measured QOL for ORC and RARC patients and found no significant difference between groups. This finding supports the results of Messer et al. and Khan et al. [
8,
9]. None of the included studies analyzed cost-effectiveness using quality measurements, such as QALY; therefore, it is recommended that future studies should focus on cost-effectiveness.
Comparison of ORC and LRC was done in only two studies that had differing conclusions on which procedure was more expensive. Further studies are required to clarify and confirm which procedure is more cost-effective.
This is the first systematic review on the segmental costs of radical cystectomy to identify which cost segments impact the total cost. Therefore, the results of this research are significant to understand cost structure and consider how RARC can be cost-effective. However, this research has several limitations. First, medical systems differ to certain extent between countries [
34]. Therefore, the results should be interpreted along with each country’s healthcare system. Correcting the segmental costs with references (e.g. NHS reference costs) will enable cost differences between institutions to be partially. Second, although clinical practice such as the use of surgical equipment could differ among institutions and surgeons, the information (e.g. number or quantity of equipment used) was not explained in detail. However, we included the database studies [
24‐
27], which allows for some generalizability of the surgeons and institutions. Finally, the study periods were up to 90 days postoperatively. Bladder carcinomas have the highest lifetime treatment cost per patient out of all malignancies [
12]. Therefore, future research on the lifetime costs would be valuable.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.