A Comparative Cost Analysis of Robot-Assisted Versus Traditional Laparoscopic Partial Nephrectomy

Introduction

N ephron-sparing surgery (NSS) is the clear standard of care for management of most small renal masses based on reduced risk of chronic kidney disease (CKD) and associated morbidity and mortality compared with radical nephrectomy. ^1,2 Laparoscopic partial nephrectomy (LPN) has been the preferred method of NSS at most tertiary centers based on equivalent cancer control and hastened convalescence compared with open surgery. ^{3 –5}

Unfortunately, partial nephrectomy has been dramatically underutilized in the United States, based on concern for complications, the technical challenge of laparoscopic suturing, and inadequate education regarding risks of CKD. ⁶

Robot-assisted partial nephrectomy (RPN) has emerged as an alternative to LPN that mitigates the technical demands of laparoscopy. The advantages of robotic systems, including improved dexterity, magnification, and ergonomics, apply naturally to tumor resection and renorrhaphy. With dissemination of robotic systems, minimally invasive surgeons have shifted to performance of RPN to harness these ostensible benefits. Despite this shift, however, there have been limited objective data regarding the advantages of robotics. Initial series of RPN and retrospective comparisons with LPN have demonstrated equivalent short-term oncologic results, similarly low complication rates, and possibly shorter warm ischemia time (WIT). ^{7 –10}

While clinical investigation of RPN continues, there has not been concomitant scrutiny of the costs associated with RPN robotic vs LPN. In our system of finite health care resources, it is critical that we evaluate the costs associated with novel technologies to place potential benefits in economic perspective. As such, in the present study, we sought to evaluate the relative costs of RPN and LPN at our institution.

Methods

Detailed cost information was obtained from our hospital's accounting department for 20 consecutive RPN and LPN patients for a single surgeon from 2009 to 2010. These included itemized actual costs for the intraoperative and postoperative period. Professional fee data for anesthesia, surgery, and pathology were also obtained.

Capital costs were estimated using the purchase costs and amortization of two da Vinci robotic systems (Intuitive Surgical, Sunnyvale, CA) in use from 2001 to 2009. Maintenance contract costs were also included. After 2009, two new units were purchased and the old units were retired; thus, capital cost analysis was restricted to data from 2001 to 2009, representing the actual operating life span of these units. The cost per case was estimated using total robotic surgical volume during this time (2113 cases). Of note, this represents the start-up period of robotic surgery at our institution.

A separate estimate of capital robotic costs was made assuming “ideal” utilization of a new da Vinci system for an 8-year period. This estimate was made using the purchase and maintenance cost of a robotic system without amortization, as were the purchase terms for our institution, as well as 300 cases per year across specialities. This is an estimate of robotic cases that could be realistically performed for one unit on an annual basis. ¹¹ Total estimated costs were compared between groups using both actual and theoretical capital costs.

Patient age, tumor size, operative time, and length of stay were also obtained and compared between groups. Statistical analysis was performed using STATA software (College Station, TX); continuous variables were compared using the Wilcoxon sum test, and categorical variables were compared using chi-square testing, with statistical significance defined as P<0.05.

Results

Table 1 shows patient age, tumor characteristics, and perioperative data for the two groups. There were no significant differences (P>0.05). Table 2 has an itemized comparison of professional fees, perioperative and hospitalization costs for RPN and LPN. There was a cost premium of $1066.09 for RPN compared with LPN using capital cost estimates from the two robotic systems in use from 2001 to 2009. The primary driver of higher costs with RPN was capital cost (+$1,907).

When “ideal” utilization of one robotic system for an 8-year period was incorporated into a separate analysis, the capital cost decreased from $1907 to $1175 per case, decreasing the cost premium for robotics (+$333.85).

Discussion

We report a cost premium of +$1066 for RPN compared with LPN during the start- up period of robotic surgery at our institution. This premium decreases to +$333 when assuming “ideal” utilization (300 cases/year/unit), because the capital costs are dispersed over a larger volume of cases. These divergent results underscore how the economic burden of robotics is largely determined by surgical volume; in 2010, there were>600 robotics cases performed with two units at our institution; thus, the estimate of 300 cases per system is reasonable for a high-volume center.

A cost premium for RPN was also reported by Mir and colleagues ¹¹ in a recent study. ¹¹ These authors performed a meta-analysis to determine perioperative parameters for RPN and LPN (eg, operative time, length of stay), and generated economic models using costs from their institution. They estimated capital costs by assuming 300 cases/year over a 7-year period, however, with amortization as was the financing arrangement at their institution. A premium of $1652 for RPN vs LPN was found. Our disparate findings result from differences in local cost structures and differences in financing and capital costs. Also, their perioperative parameters were obtained from a meta-analysis, while ours were obtained from a single center with a mature minimally invasive partial nephrectomy (MIPN) experience, which impacted absolute and relative costs as well. Both analyses nonetheless found an increased cost associated with RPN, although modest when assuming high-volume robotic surgery.

Increased costs associated with RPN are not inherently problematic, provided that robotics provides proportionate clinical advantages. While preliminary series of RPN have reported similar perioperative data to LPN, retrospective comparisons have suggested advantages for RPN in WIT. ^8,12,13 Unfortunately, these studies have been retrospective and have had limited power. Furthermore, it is not clear whether improvements in WIT within a relatively short window of warm ischemia (<20 minutes) are clinically significant in preventing acute kidney injury or CKD for most patients. ^14,15

Matched retrospective case control studies with LPN have been small and in the early experience with RPN. ¹⁶ Initial data suggest a benefit for RPN in operative time and blood loss for treatment of more complex masses, ⁸ and the learning curve for RPN may be shorter than for LPN. ^17,18 Clarification of these clinical data is needed.

There may be other clinical benefits for RPN that have associated cost advantages in the long term. Increased utilization of MIPN based on facilitated intracorporeal suturing and an improved learning curve with robotics may help to alter practice patterns such that partial nephrectomy is more commonly performed, reducing risk of CKD in this population. ⁶ Similarly, providers may attempt MIPN for more complex tumors that would otherwise be treated by laparoscopic radical nephrectomy; there is promising evidence that RPN enables relatively short WIT with a low complication rate in the treatment of hilar and other complex tumors. ^19,20

Our study revealed other interesting cost features of RPN and LPN. Surgical supply costs were equivalent between the two procedures after integrating the costs of disposable instrumentation for robotics. LPN surgical supply costs were initially higher, primarily related to use of LigaSure (Covidien, Boulder, CO) in these cases. Indeed, the cost of LigaSure can exceed $700 per instrument, which initially drove a discrepancy between the procedures. While we prefer LigaSure for effective sealing of blood vessels and lymphatics, tools such as harmonic scalpel or bipolar forceps are less expensive alternatives that can be considered for the purpose of reducing costs.

Notably, surgical supply costs for the two procedures became equivalent when integrating disposable instrument costs for the robot. Indeed, costs of disposable robotic instruments range from $2200 to $3200 for needle driver, bipolar forceps, and monopolar scissors, and when averaged over 10 uses contributed $810 per case. While robotic instruments are obviously needed for that procedure, we do advocate for both LPN and RPN a conservative approach to use of disposables; surgeons should be conscientious about minimizing waste and using only the tools and materials that confer a clinical benefit within their judgment, or based on data when available.

There were no significant differences in operative time between the two groups, and thus no cost differential based on operating room utilization. This may reflect the comfort of the study surgeon with both RPN and LPN, and similar disease features of the treated population. There were no significant differences in length of stay between the two groups, which was expected, considering these patients had similar incisions, postoperative pathways, and disease characteristics. Indeed, during the study period, patients underwent LPN or RPN based largely on availability of the robotic system, given the study author's comfort with LPN as the standard approach.

As mentioned above, one additional study in the literature has evaluated comparative costs of RPN and LPN. Mir and colleagues ¹¹ compared direct costs associated with open, laparoscopic and robotic assisted partial nephrectomy through a cost model incorporating perioperative data from a meta-analysis, and utilizing local costs. They determined that LPN was the most cost-effective procedure with cost advantages of $1116 and $1652 over open partial nephrectomy (OPN) and robot-assisted laparoscopic partial necphrectomy (RALPN), respectively. RPN had substantially higher surgical equipment costs. Sensitivity analysis demonstrated that it would be impractical to make RALPN cost equivalent to LPN regarding operative time or equipment costs (operative time<106 minutes, equipment cost $169/case). They did not modulate robotic surgical volume, which would substantially affect capital cost and might ameliorate the cost disadvantage of robotics in a scenario of maximal utilization.

Other studies have evaluated relative costs of treatment for renal masses. Park and coworkers ²¹ compared costs of LPN and OPN; the higher cost of OPN was driven primarily by increased length of stay, despite the higher equipment costs of laparoscopy. Indeed, this is a common tradeoff between open and minimally invasive surgery. Lotan and Caddedu ²² examined comparative costs of OPN, LPN, and percutaneous radiofrequency (PRF) ablation using actual costs of 46 patients at their institution. PRF was less costly based primarily on shorter hospitalization, while LPN and OPN were cost equivalent because the short stay of LPN compensated for higher surgical supply costs.

The differential oncologic control rates for these procedures are relevant, however, because there is a higher rate of need for re-treatment with PRF. ^23,24 This concern is balanced against the limited longevity and high comorbidity index of many of these patients. These issues of oncologic efficacy are unlikely to be relevant for RPN, because it is technically analogous to LPN which has superb long-term cure rates.

It is unlikely that most centers can achieve “ideal” utilization of a robotic system—ie,>300 cases/year—to mitigate the significant capital costs associated with robotics. Robotic systems and associated maintenance contracts represent a substantial capital investment; updated robotic systems were obtained by our institution after 2009, for which purchase costs were $1.7 million, with maintenance contracts of $140,000 per year. The fact of capital investment may create an impetus to use these systems; high volume may help to reduce attributable costs and mitigate the impact of depreciation of capital equipment. ²⁵ Indeed, there is evidence that an increased volume of robotic surgery may result based on robot acquisition alone. ²⁶ Radical prostatectomy surgery is a useful case study; there has been an aggressive transition from open to robot-assisted prostatectomy, despite lack of rigorous evidence for superiority of robotics and some reports of inferior outcomes. ^27,28 Market forces have clearly contributed to these phenomena, and it is important that we attempt to ground our practice patterns in clinical data.

There are limitations of our study that deserve mention. Our assumptions regarding capital costs and surgical volume profoundly impacted the outcome of our analysis. Furthermore, we used local cost structures that may not be generalizable to other centers. The operative times and use of resources reflect the experience of the author surgeon and the robotics team at our hospital, which also may not be generalizable. There is the possibility of sampling error, but we believe that 20 patients in each group is a reasonable number to ascertain differences in cost, and there were no obvious outliers in either group. While tumor size and patient age were comparable between groups, there may have been differences between RPN and LPN patients that impacted resource utilization. Use of robotics during the study period, however, was driven largely by availability of the robotic system, because the author surgeon had a mature skill set with LPN and had been performing RPN for approximately 2 years. As such, the role of selection bias is likely minimized.

Finally, long-term data pertaining to complications or future therapy were not included. Complications are rare in this population, however, and recurrence rates are extremely low for these low-stage tumors regardless of approach.

Of note, our analysis included direct costs but did not account for the “opportunity cost” of robotics. This entails the alternative treatments that might have been pursued with the time and resources dedicated to robotic surgery. This is difficult to quantify and is likely to vary between procedures. Nonetheless, opportunity cost should be integrated into future analyses to help understand the true cost of robotics. There has been a rapid movement toward robotics within most surgical disciplines—eg, urology, gynecology, general surgery—and it is necessary that we seek and critically evaluate both cost and clinical data to ensure these transitions are justified.

Conclusions

We report a variable cost premium for RPN compared with LPN depending of assumptions regarding surgical volume. The increased overall cost of RPN may be justified based on clinical benefits, but objective data are lacking. While there are clear subjective benefits of RPN for tumor resection and renorrhaphy, the objective benefits should be stringently evaluated before widespread adoption. It is critical that cost-effectiveness analysis accompanies the adoption of new technologies to ensure responsible and sustainable stewardship of health care resources. The wide dissemination of robotic technology is hopefully a boon for patients, but it is critical that market forces do not dictate the standard of care.