Optimizing the Financial Burden of the Approach to Robot-Assisted Radical Prostatectomy

Introduction

Prostate cancer is the second most frequently diagnosed cancer with an incidence of 1.1 million cases a year. ^{1 –4} The disease is a significant public health burden and major cause of mortality worldwide. The use of robot-assisted surgery (RAS) in the treatment of prostate cancer has significantly expanded since it was first introduced in 2000 owing to improved ergonomics, dexterity, three-dimensional vision compared with open and laparoscopic approaches. In 2009, 70% of radical prostatectomies in the United States were done with robotic assistance. ^5,6 Safety and efficacy of robot-assisted radical prostatectomy (RARP) have been confirmed in multiple studies. ^7,8

Nevertheless, despite the benefits offered for both the surgeon and the patient, one of the main criticisms for RARP has been the high cost associated with installation of surgical robot, as well as the running costs. ^9,10 The chief driver of costs is the cost of installing the robotic system. In addition, the use of instruments, drapes, and surgical packs significantly adds to the costs. ¹¹ Therefore, managing the financial burden of RARP is critical as it continues to gain applicability and become ubiquitous. Proponents of RAS have suggested that high volume can drive down the costs of RARP and make it more cost effective. Other ways to optimize the cost of RAS are to reduce the number of instruments and optimize the supplies utilized during the procedure. Ramirez and colleagues have suggested that only three instruments are required to complete RARP, and described a detailed step-by-step process on how to complete the procedure but did not provide detailed financial outcomes. ¹²

In our study, we sought to analyze the variable costs associated with RARP at a Comprehensive Cancer Center by a single high-volume robotic surgeon, and then describe how our pilot cost reduction plan reduced the affected cost associated with RARP.

Methods

We retrospectively reviewed the RARPs performed by a single surgeon between April 1, 2017 and March 31, 2018. Costs were presented in the form of direct costs to the department. A cost analysis was performed for these cases looking at the variable costs associated with RARP: anesthesia-related costs, operative time, and medical supplies. The calculated variable cost per case was the sum of anesthesia-related costs, operative time costs, and medical supply costs. Anesthesia-related cost (from patient anesthetized to awake) was billed in 15-minute increments. Operative time cost (from skin incision to skin closure) was billed in 15-minute increments. The operative time cost was derived from the console time costs, which are also billed in 15-minute increments. These two variable costs are primarily time driven, and were measured to see whether there was an effect on case time that minimized the impact of our pilot process. Medical supplies included the cost of sutures, clips, staplers, trocars, and robotic instruments. The cost of the robotic instruments was fixed. Data were captured by our institution's electronic record-keeping system. Initially, six robotic instruments were utilized: two needle drivers, one Maryland Grasper, one Cobra Grasper, one scissor, and one Cautery Hook. V-Lock sutures and vascular staplers (Ethicon-Endo-Surgery, Guaynabo, Puerto Rico) were employed along with disposable trocars. Professional costs for anesthesia, surgeon, and labor were not included in this analysis because these are fixed costs that are non-negotiable.

From August 1, 2018 to December 31, 2018, a prospective cost reduction pilot process was implemented by the same surgeon. This prospective analysis began on August 1st as the cost review for the prior fiscal year was completed in July 2018. A cost analysis was again performed for these cases and was compared with the retrospective cases. The cost reduction plan included using fewer instruments (only four robotic instruments: one needle driver, one Maryland Grasper, one Cobra Grasper, and one Curved scissor). V-Lock sutures were switched for standard monocryl and vicryl sutures, while staplers were no longer utilized. Trocars were switched from disposable to reusable trocars, and reprocessing fees were included in this change (Table 1). All of the cases were performed on the Da Vinci Xi^®.

Results

Forty RARPs were retrospectively reviewed between April 2017 and March 2018, and 32 RARPs were prospectively evaluated between August 2018 and December 2018 after implementation of the pilot cost reduction plan. No cases were excluded. There was no statistically significant difference in the perioperative characteristics of patients. There was no significant difference in operative time between the two groups (142 ± 30 vs 127 ± 26, p = 0.12) estimated blood loss (146 ± 96 mL vs 188 ± 160 mL, p = 0.33), or complications (13% vs 25%, p = 0.22). None of the patients developed high-grade complications (Table 2). There was no significant difference in complications (overall and categorized) between both groups (Table 3).

In our retrospective review, total variable costs (which are the sum of medical supply costs, technical anesthesia costs, and operative time costs) were $3935 ± $99 (95% CI $3903, $3966). When looking at the specific components of variable cost, the total medical supply cost was $2918 ± $72 (95% CI $2894, $2941). In medical supplies, an average of $1987 ± $131 (95% CI $1945, $2029) was spent on robotic instruments per procedure, $761 ± $149 (95% CI $713, $809) was spent on surgical supplies (sutures, staplers, clip, etc.), $170 ± $148 (95% CI $122, $217) was utilized on disposable trocars. Technical anesthesia costs were $406 ± $73 (95% CI $382, $429). Total operative time costs were $611 ± $138 (95% CI $568, $654).

In our prospective evaluation, total variable costs were $3230 ± $82 (95% CI $3200, $3259). Looking at the specific components of variable cost, the total medical supply cost was $2447 ± $64 (95% CI $2423.70, $2470.3). In medical supplies, $1699 ± $121 (95% CI $1655, $1742) was spent on robotic instruments. $688 ± $154 (95% CI $632, $743) was spent on surgical supplies. $60 ± $7 (95% CI $57, $62) was spent on reusable trocars. Technical anesthesia costs were $370 ± $54 (95% CI $350, $389). Operative time cost was $413 ± $63 (95% CI $395, $431).

When looking at the net savings per case, the savings in total variable costs was $705 (95% CI $661, $748, p < 0.0001). The savings in medical supply costs per case was $471 (95% CI $438, $503, p < 0.0001). The savings per case in robotic supply costs was $288 (95% CI $227, $348, p < 0.0001). The savings per case in surgical supplies was $73 (95% CI $1, $144, p = 0.0458). The savings in trocars was $110 (95% CI $57, $162, p < 0.0001). The savings per case in anesthesia technical costs was $36 (95% CI $5, $66, p = 0.0229). The savings per case in operative time costs was $198 (95% CI $145, $250, p < 0.0001) (Table 4).

Discussion

Avoiding interventions in low-risk prostate cancer patients who are candidates for active surveillance is a key without compromising oncologic outcomes. However for patients undergoing definitive management, the increasing ubiquity of RAS, optimizing the costs associated with the system, is critical to maintaining its value to the patients. Many studies have identified the drivers of cost in robotics as the cost of purchasing a robot, the corresponding instruments, and maintaining the system. ^{13 –15} While the upfront cost of the robot is high, savings can be generated through the approach's efficacy in reducing patient stay and better management of complications. Kukreja and colleagues showed that robot-assisted radical cystectomy is not cost effective but improves health-related quality of life outcomes and in those improved outcomes is comparable to an open approach. ¹⁵ Another factor to drive the costs down of the system is to increase its usage among surgical disciplines. ¹⁶ As more cases are performed, a better balance is expected in the cost of the console time and depreciation.

For this study, we excluded the costs associated with maintenance and the cost of purchasing the system. Our institution has chosen to utilize two of the Da Vinci Xi systems, and it is widely used by at least five surgical subspecialties. Therefore, our goal was to optimize the variable costs surrounding this modality: instrumentation and time. In our current report, we showed that by reducing the number of robotic instruments to four and being cognizant of the surgical supply selection we could achieve an ∼16% reduction in medical supply costs. There was no difference in the cost of medical supplies from the evaluative to the prospective period, suggesting that the savings were realized through a reduction in instrumentation. When looking at the total variable costs of the procedure, we were able to achieve ∼20% cost reduction per case. If our approach was to gain applicability at our institution, we estimate that there would be net savings of ∼$104,000 per year. Other reports have suggested that the cost of a RARP could be potentially decreased by ∼40%. ¹⁷ However, this approach would involve utilizing a barebones set of surgical tools of solely a Prograsp forceps, robotic needle driver, and monopolar scissors. Anecdotally, we tried this approach after our prospective evaluation period, and we found it to be cumbersome. In the goal of cost savings, surgeon efficacy, as well as training the fellows and residents should not be compromised. Our study showed that our approach actually had a corresponding decrease in operative time. While this decrease in operative time was not statistically significant, there was a statistically significant decrease in the cost of console time. While other factors can drive this decrease like increasing utilization of the robot, these two elements together represent another source of variable cost savings to an institution.

While reducing the cost of the robotic system is critical, it is equally important to maintain patient safety. There was no difference in the clinical demographics of our cohort or the estimated blood loss during the procedure. In addition, no patient has experienced a high-grade complication (complication involving readmission) as a result of our new approach, and there was no difference in the any complication rate between the two groups. However, not all surgeons should immediately look to decrease their costs. The ability to shift focus from technical ability to variable events such as cost is the result of having gained experience with the technique and surgical approach. In other types of robotic surgeries, the literature has suggested that after ∼30 robot-assisted cases the learning curve begins to flatten. ¹⁸ In this study, we looked at the most recent retrospective data, and then prospectively implemented a cost reduction plan to minimize the effect that the learning curve can have on RAS. In addition, the console surgeon in our report is an expert surgeon who has performed a high volume of robot-assisted procedures (>10,000 console hours). This might explain why there was no corresponding increase in operative time with the introduction of our new approach. Having an experienced surgeon and a dedicated team is critical to maintain the cost savings realized through instrument reduction. Our group has published work, showing that having an experienced team that can anticipate the surgeon's actions can reduce operative time and lead to increased efficiency in the operating room. ¹⁹

It is noteworthy that cost reductions may not be as obvious in lower volume centers, where the focus should be continuous improvement in oncologic and functional outcomes. On the contrary, established higher volume centers would benefit even from small changes as they add up because of the higher volume.

Our study has several limitations. It relies on a small sample size. Furthermore, the surgeries were all performed by a single, high-volume, experienced surgeon, which may limit the generalizability of our findings. Also, our approach may hinder trainee's education. Operative time was calculated from skin incision to closure; however the actual robotic use time (time from docking to dedocking) was not calculated. Although our study did not show a significant difference in complication rates, this may be attributed to the relatively small number of procedures included in the study. Our results are exploratory, and suggest that deeper exploration is needed into optimizing the cost of robot-assisted surgeries for our health care programs.

Being cognizant of the cost implications of RAS can help generate solutions to address the corresponding challenges. Our study suggests that decreasing instrumentation can have a significant impact on variable costs without any impact on operative time or patient safety.