BJUI. Assessing the cost effectiveness of robotics in urological surgery a systematic review

Transcription

1 BJUI BJU INTERNATIONAL Assessing the cost effectiveness of robotics in urological surgery a systematic review Kamran Ahmed, Amel Ibrahim, Tim T. Wang, Nuzhath Khan, Ben Challacombe, Muhammed Shamim Khan and Prokar Dasgupta MRC Centre for Transplantation, King s College London, King s Health Partners, Department of Urology, Guy s Hospital, London, UK Accepted for publication 1 November 211 Study Type Therapy (systematic review) Level of Evidence 1a OBJECTIVES Although robotic technology is becoming increasingly popular for urological procedures, barriers to its widespread dissemination include cost and the lack of long term outcomes. This systematic review analyzed studies comparing the use of robotic with laparoscopic and open urological surgery. These three procedures were assessed for cost efficiency in the form of direct as well as indirect costs that could arise from length of surgery, hospital stay, complications, learning curve and postoperative outcomes. What s known on the subject? and What does the study add? Research on the subject has shown that robotic surgery is more costly than both laparoscopic and open approaches due to the initial cost of purchase, annual maintenance and disposable instruments. However, both robotic and laparoscopic approaches have reduced blood loss and hospital stay and robotic procedures have better short term post-operative outcomes such as continence and sexual function. Some studies indicate that the robotic approach may have a shorter learning curve. However, factors such as reduced learning curve, shorter hospital stay and reduced length of surgery are currently unable to compensate for the excess costs of robotic surgery. This review concludes that robotic surgery should be targeted for cost efficiency in order to fully reap the benefits of this advanced technology. The excess cost of robotic surgery may be compensated by improved training of surgeons and therefore a shorter learning curve; and minimising costs of initial purchase and maintenance. The review finds that only a few studies gave an itemised breakdown of costs for each procedure, making accurate comparison of costs difficult. Furthermore, there is a lack of long term follow up of clinical outcomes, making it difficult to accurately assess long term post-operative outcomes. A breakdown of costs and studies of long term outcomes are needed to accurately assess the effectiveness of robotic surgery in urology. METHODS A systematic review was performed searching Medline, Embase and Web of Science databases. Two reviewers identified abstracts using online databases and independently reviewed full length papers suitable for inclusion in the study. RESULTS Laparoscopic and robot assisted radical prostatectomy are superior with respect to reduced hospital stay (range days and days, respectively) and blood loss (range ml and ml, respectively) when compared with the open approach (range 2 8 days and 115 ml). Robot assisted radical prostatectomy remains more expensive (total cost ranging from US $2 $39 215) than both laparoscopic (range US $74 $29 771) and open radical prostatectomy (range US $187 $31 518). This difference is due to the cost of robot purchase, maintenance and instruments. The reduced length of stay in hospital (range days) and length of surgery (range min) are unable to compensate for the excess costs. Robotic surgery may require a smaller learning curve (2 4 cases) although the evidence is inconclusive. CONCLUSIONS Robotic surgery provides similar postoperative outcomes to laparoscopic surgery but a reduced learning curve. Although costs are currently high, increased competition from manufacturers and wider dissemination of the technology could drive down costs. Further trials are needed to evaluate long term outcomes in order to evaluate fully the value of all three procedures in urological surgery. KEYWORDS robotics, urology, healthcare economics, training, outcomes BJU INTERNATIONAL 11, doi:1.1111/j x x

2 ECONOMIC IMPACT OF ROBOTIC SURGERY IN UROLOGY INTRODUCTION Robotic surgery is becoming increasingly common in pelvic urological procedures such as radical prostatectomy and cystectomy [1 ]. Barriers to widespread dissemination include cost and the lack of long term outcomes. Despite this, robotic technology is gradually becoming popular due to a reduced learning curve, shorter hospital stay, comparable clinical outcomes and demand from patients [1,2 ]. Therefore the overall economic impact of robotic technology is considerable. Surgical departments need to use this concept of healthcare economics to see if services can be optimized in a cost efficient manner [3 ]. Health economics, that aim to provide maximum value for money without impacting quality, need to be appropriately researched and implemented for emerging technology such as robotics [4 ]. Urological conditions such as prostate cancer are often detected at an early stage (due to PSA screening) and radical prostatectomy can be an effective treatment option. Open radical prostatectomy may be associated with higher complications, increased hospital stay and postoperative morbidity [5 ]. Laparoscopic radical prostatectomy has been widely used to overcome these challenges. However, this technique has a steep learning curve and technological drawbacks including twodimensional images, limited degrees of freedom for movement of instruments and increased operative times [6 ]. Since 2, the da Vinci system (Intuitive Surgical, Sunnyvale, CA, USA) has been increasingly employed to provide an alternative for certain procedures such as radical prostatectomy. It may reduce hospital stay, improve postoperative outcomes and decrease the learning curve [ 2,7 ]. The major challenge at present is high operative costs with some studies claiming that robot assisted surgery adds US $1 25 to a procedure [8 ]. Robotic technology in surgery has enjoyed rapid expansion since its inception. There still remains a great deal of uncertainty over its viability with sceptics citing cost efficiency as its major shortfall. Indeed, the recent global economic difficulties have meant a desire by governments to implement deficit-reducing measures which have placed health organizations under increased pressure to justify its expenditure. Some argue that robotics may not be a financially viable option in comparison with laparoscopic and open techniques [9 ]. This article reviews systematically the evidence to analyze the concept and trends of cost effectiveness in relation to robotic urological surgery. METHODS This systematic review was conducted according to the PRISMA guidelines (Preferred Reporting Items for Systematic reviews and Meta-Analyses) [1 ]. Studies comparing cost of urological robotic procedures with other types of operation were included. Case reports, case series and empirical studies which did not report cost, procedural outcomes, technical and clinical challenges and training issues related to robotic urological surgery were excluded. We also excluded reviews, editorials, letters and studies not directly related to robotic surgery. Relevant studies were identified by searching the following databases: (i) Ovid Medline (195 May 21), (ii) Ovid MEDLINE(R) In-Process & Other Non-Indexed Citations 195 May 21), (iii) Ovid EMBASE (198 21) and (iv) Ovid PsychINFO ( ). We also searched the Cochrane database to look for any reviews on this subject. The search was performed using free text terms describing cost effectiveness in urological surgery in relation to the outcomes. The search terms included a combination of: robotics or minimally invasive or surgery and urology or cost or healthcare economics or robot or learning curve. Reference checks of published review articles were performed to supplement the aforementioned searches. Reviewers (KA and AI) independently identified potentially relevant articles. The full text of the article was obtained and further screened for inclusion if it had at least one of the following categories of information: robotic urological surgery, surgical outcomes (including cost, efficacy and safety), guidelines, health care economics or disease impact (use of resources, re-admissions and quality of life). Conflicts between the reviewers were subsequently discussed until there was 1% agreement on the final studies to be included. Two independent investigators (KA and AI) assessed each study for inclusion and quality assessment. The inter-rater agreement for inclusion was satisfactory with inter-rater reliability of.93. An electronic data collection form (Microsoft Excel 27) was used to extract data according to the outcome measures. Where applicable, for each item the statistical method used in the study was recorded. Disagreement in the assessment and data extraction were resolved by consensus. Due to heterogeneous study designs and a learning curve effect, direct comparisons or meta-analysis of data were not feasible. However, the framework from similar studies has been summarized. In the body of this article we review comparative studies to compare robotic surgery with laparoscopic and open approaches for the main urological procedures. In our analysis of each individual study, we discuss the following: (i) postoperative outcomes, (ii) training and learning curve and (iii) cost efficiency The quality of each study included in the analysis was assessed using variables based on existing stratification tools such as the Jadad s, MINORS (Methodological index for non-randomized studies) and Oxford EBM and the Grade system [11 13 ] ( Table 1 ). Six dimensions were used to evaluate the quality of the individual studies: (i) Title: whether the study clearly identifies itself as comparative of costs associated with robotic, laparoscopic and/or open urological surgery, (ii) Abstract: whether there is a structured format which outlines comparative groups and summarizes results, (iii) Introduction: whether the clinical problem is described and background information given, (iv) Methods: whether it explains measures taken to optimize study quality, defines groups and outcomes measured, (v) Results: whether outcomes are reported accurately and clearly and (vi) Discussion: if key findings are analyzed and limitations mentioned. 212 BJU INTERNATIONAL 1545

3 AHMED ET AL. RESULTS Six hundred and nine potentially relevant publications were identified by the search. Five hundred and eighty were excluded following the abstract review. Of the remaining 29 studies, we excluded a further 21 after reviewing the full text. Five additional studies were identified after reviewing references from relevant studies. Thirteen studies were finally included in the systematic review (Fig. 1 ). The selected studies consisted of case series and a limited number of case-control studies ( Table 2 ). The evidence was primarily level 2 in the form of systematic reviews as well as non-randomized comparative studies based on the Oxford Centre for Evidencebased Medicine [14 ]. Information from these studies was categorized under the headings of (i) postoperative outcomes, (ii) training and learning curve and (iii) cost efficiency (Figs 2 and 3 ). POSTOPERATIVE OUTCOMES A systematic review by Ficarra [15 ] analyzed studies comparing robot assisted with open radical prostatectomy. They found that robot-assisted prostatectomy had a shorter hospital stay (1 17 days, median = 1.5 days), limited blood loss ( ml, median = 329 ml), short learning curve (4 6 cases) and promising outcomes at follow-up in terms of erectile function (2 79%, median = 72%), urinary continence (82 96%, median = 91.6%) as well as similar positive surgical margins (2 59% median = 13%) to laparoscopic and open approaches [15 ]. Although promising, the authors acknowledged that further data were needed to establish significance of long term oncological and sexual potency outcomes. A later systematic review by Ficarra [16 ] analyzed studies which compared robotic ( n = 14 studies), laparoscopic ( n = 27 studies) and open prostatectomy ( n = 33 studies). They found that length of surgery was increased by both laparoscopic (weighted mean difference (WMD) 71.2 min, 95% confidence interval (CI) 97.35, 45.5, P <.1) and robotic approaches (median 231 min, range min) when compared with open (median 24 min, range min). Laparoscopic prostatectomy appeared TABLE 1 Quality of the included studies Hu [23 ] Lowrance [22 ] [27 ] [24 ] [2 ] Lotan [41 ] [33 ] [18 ] [32 ] [35 ] [2 ] [31 ] Lotan [19 ] Authors Quality criteria Title Identify the study as a comparative analysis of cost between open vrobotic prostatectomy/cystectomy Abstract Uses a structured format Clearly states the objectives Outlines comparative groups and which costs were analyzed Summarizes the main results Introduction Describes the clinical problem, rationale for comparative analysis Methods Provides a specific timeline over which data were collected Clearly outlines comparative groups and numbers in each Describes how cost data were obtained Subdivides costs into operative/nonoperative Further subdivides operative/nonoperative cost into specifics Explicitly states cost of robot / whether cost of the robot is included Accounts for learning curve by separately comparing cost of the initial cases vs later cases 1 Provides selected metrics of outcome (operative time, blood loss, length of hospital stay, positive margin %, PSA recurrence %) Describes a valid method and software of statistical analysis Results Accurately summarizes results in study Provides statistical analysis of comparison groups Graphical/Summary table representation of results Discussion Summarizes and interprets key findings Discusses results of study in view of totality of evidence available Discusses limitations of study Total: BJU INTERNATIONAL

4 ECONOMIC IMPACT OF ROBOTIC SURGERY IN UROLOGY FIG. 1. Search study for study selection. Titles/abstracts reviewed (based on search terms) 69 articles identified and screened from Medline EMBASE PsycINFO 58 articles excluded after reviews based on abstract and title 29 Retrieved for full text evaluation 21 excluded after full text search (Articles using virtual reality, motion analysis devices) 8 articles identified for inclusion in the study Five articles identified from cross referencing suggested that operative time was highest for robotic surgery (262 min) compared with open retropubic (22 min) and perineal (196 min). However, blood loss was less in robotic prostatectomy (227 ml) compared with open perineal (115 ml) and retroperitoneal prostatectomy (78 ml). The work of [2 ] found that length of hospital stay was higher in open prostatectomy (2.51 days) compared with laparoscopic (1.76 days) and robotic surgery (1.56 days). They also showed that nerve sparing occurred in 85% of robot assisted procedures compared with 96% of laparoscopic and 9% of open ( P <.1). Lymph node dissection was most likely to be performed in open prostatectomy (1%) in contrast with laparoscopic (22%) and robotic approaches (11%) ( P <.1). This difference in lymphadenectomy and nerve sparing procedures was not correlated with medium term outcomes. superior in comparison with open with respect to blood loss (WMD , 95% CI of WMD 277.1, , P < 1), transfusion rates (relative risk (RR) 4.72, 95% CI of RR 2.2, 1.14, P <.1), catheterization times (WMD 6.18, 95% CI of WMD.46, 11.91, P =.3), length of hospital stay (WMD 2.46, 95% CI of WMD 1.54, 3.37, P <.1) and postoperative complications. Both laparoscopic and robotic approaches showed similar outcomes of potency and incontinence ( P =.16) (both superior to open approach). Oncological outcomes in the form of positive margins were similar for all three techniques. However, the authors report that the quality of the included studies was not excellent and thus recommended further prospective, multicentre studies of high quality to improve the accuracy of the data. Postoperative outcomes have also been shown to improve in a study by Menon [ 17 ]. They used a cohort of over 26 patients and showed that with experienced surgeons robotic prostatectomy yielded.8% frequency of incontinence at 12 months. In patients undergoing bilateral Veil nerve sparing procedures, 7% sexual potency was achieved at 12 months and Thirteen articles included in final analysis 1% in 48 months. However, only half of these patients described this return to sexual function as normal compared with pre-operatively. A review by [18 ] suggested that robot assisted prostatectomy may be associated with reduced incidence of sexual dysfunction (22 85%, median 61%) ( Table 2 ) compared with laparoscopic prostatectomy. They also show similar rates for positive surgical margins. Lotan [19 ] compared open, laparoscopic and robot assisted prostatectomy with respect to postoperative outcomes. They found that length of stay was the same for laparoscopic and robotic surgery (1 day) which were both less than open prostatectomy (3 days). Meanwhile length of surgery was reduced in robot assisted approaches (14 min) compared with laparoscopic (16 min) and open (2 min). The study by [2 ] reported that length of stay was the same for robot assisted as well as open perineal prostatectomy (1 day). It was increased for the open retropubic approach (2 days). They Lowrance [21,22 ] used the SEERmedicare dataset to analyze postoperative outcomes at 9 days as well as 365 days for genitourinary complications and bowel disorders. They compared patients receiving open radical prostatecetomy ( n = 4697 patients) with those undergoing minimally invasive prostatectomy ( n = 16). They did not find a significant difference in postoperative complications between the two groups (OR.93, 95% CI.77, 1.14, P =.49). They also did not find a difference in development of bladder neck obstruction or urethral contracture between the groups (OR.96, 95% CI.76, 1.22, P =.74). The authors however, grouped robot assisted prostatectomy with laparascopic prostatectomy. Median length of stay (2 vs 3 days, P <.1) was on average 35% shorter in minimally invasive procedures when patient and tumour characteristics are controlled for. They explain, in a later review, that the majority of minimally invasive procedures are robot assisted [21 ] and thus these data represents a comparison of robotic with open prostatectomy. Hu [23 ] also used the SEERS-medicare dataset to compare minimally invasive radical prostatectomy (MIRP) ( n = 1938) with the open approach ( n = 6899). They compared 3 day postoperative complications, anastomotic stricture up to 365 days post-procedure, as well as incontinence and erectile dysfunction more 212 BJU INTERNATIONAL 1547

5 AHMED ET AL. than 18 months postoperatively. They also identified the need for further postoperative cancer therapies to determine indirectly cancer control. Using propensity scoreadjusted analyses, Hu [23 ] found that MIRP was associated with reduced length of stay (median, 2. vs 3. days, P <.1) and less need for blood transfusion (2.7% vs 2.8%, P <.1) when compared with open radical prostatectomy. MIRP was also superior with respect to fewer postoperative respiratory complications (4.3% vs 6.6%, P =.4), miscellaneous surgical complications (4.3% vs 5.6%, P =.3) and anastomotic stricture (5.8% vs 14.%, P <.1) [23 ]. However, MIRP was associated with worse long term complications in the form of genitourinary complications (4.7% vs 2.1%, P =.1) such as incontinence (15.9 vs 12.2 per 1 person-years, P =.2) and erectile dysfunction (26.8 vs 19.2 per 1 person-years, P =.9). There was no difference in use of additional cancer therapies between the two approaches (8.2 vs 6.9 per 1 person-years, P =.35) [23 ]. Cystectomy is another procedure that we tried to review. Table 5 shows an overview of postoperative outcomes of cystectomy. [24 ] revealed that the average length of hospital stay was the same for open and robotic cystectomy (5 days) whilst length of surgery was longer for robotic cystectomy (246 min) compared with open (228 min). Nix [25 ] conducted a randomized controlled trail (RCT) which showed the robotic approach to be equivalent to the open with respect to radical cystectomy with pelvic lymph node clearance and negative margins. Ahlering [26 ] also found that robotic surgery had significantly better outcomes such as blood loss, length of stay and return of bowel function. [27 ] found that the median length of stay was lower for robotic cystectomy (5.5 days) than open cystectomy (8 days). The mean complication rate was slightly lower for robotic cystectomy (49.4%) compared with open cystectomy (61%) and the mean cost of complications was significantly lower for robotic cystectomy (US $1624) compared with open cystectomy (US $722). Whilst robotic surgery appears to offer benefits over open approaches in terms of reducing post-operative complications and length of stay (range days vs 2 8 days) (Table 2 ), there have not been any FIG. 2. An overview of cost comparison of robotic surgery with the other modalities Total Price Comparison of Open, Lotan (24) Non-Operating Cost Comparison of Open, Lotan (24) Equipment Cost Comparison of Open, Lotan (24) Anaesthetic Cost Comparison of Open, Lotan (24) Laboratory Cost Comparison of Open, Lotan (24) Lenght of Stay Comparison of Open, Lotan (24) 1548 Time in Min Operating Cost Comparison of Open, Lotan (24) Professionals Cost-Comparison of Open, Lotan (24) Operating Room Cost Comparison of Open, Lotan (24) Medication Cost Comparison of Open, Lotan (24) Room and Board Cost Comparison of Open, Lotan (24) Length of Surgery Comparison of Open, Lotan (24) 212 BJU INTERNATIONAL

6 ECONOMIC IMPACT OF ROBOTIC SURGERY IN UROLOGY FIG. 2. Continued No. of Days Volume in ml Lotan (24) Lotan (24) Lenght of Stay Comparison of Open, Blood Loss Comparison of Open, Lowrance Hu No. of Cases Open Robotic Laproscopic Number of Cases Comparison of Open, Lotan (24) KEY RARP robot assisted radical prostatectomy LRP laparoscopic radical prostatectomy ORP open radical prostatectomy MIRP: Minimally Invasive Radical Prostatectomy CAP: Cryosurgical Ablation RARC robot assisted radical cystectomy LRC laparoscopic radical cystectomy ORC open radical cystectomy RALP robot assisted laparoscopic pyeloplasty LP laparoscopic pyeloplasty RC - Robotic cystectomy OC - Open cystectomy IC - Ileal conduit CCD - Continent cutaneous diversion OC ON - Orthotopic neobladder TABLE 2 Cost related equipment and peri-operative outcomes of robotic prostatectomy Authors Lotan [19] [31 ] [2 ] [32 ] Lotan [42 ] [33 ] [2 ] Robot cost included Y/N Number of patients Peri-operative outcomes Type of surgery Total cost Operative charges Nonoperative charges Professionals Surgical equipment LRP 6 41 Y x x RARP 7 28 x x ORP x x RARP N x x x ORP Specialist setting x x x ORP Community setting x x x ORP N x x X LRP x x X RARP x x X ORP 1 74 x x 197 x 2471 LRP x x 6 x 2788 RARP 1 47 N x x 137 x 3441 CAP X x 58 x 572 LRP N X X RARP x X ORP x X LRP X X x x ORP 1 87 X x x x RARP 5 41 X x x x LRP X x x RARP Y x x 262 x ORP x x 161 x RARPS x Y x x x x 175 x [18] LRP x x x x x 533. x RALP 2 Y X x 1 X x X [35] LP 74 X x X 1 x X x Lowrance LRP x x x x 165 x x x [22] ORP x x x x 4858 x x x Hu [23] MIRP x x x x 1938 x x x RRP x x x x 6899 x x x Operating room 212 BJU INTERNATIONAL 1549

7 AHMED ET AL. RCTs comparing postoperative as well as long term functional outcomes following robotic surgery compared with laparoscopic approaches. TRAINING AND LEARNING CURVE There is no overall consensus on the learning curve of robotic prostatectomy. Some studies, suggest that robotic prostatectomy may have a shorter learning curve than laparoscopic [18 ] (Table 3 ). The da Vinci robot was embraced by urologists particularly in the USA as it allowed substantial magnification ( 1 2), 3-D HD vision, 6 degrees of movement of the instruments and modulation of movement (i.e. can downscale movement to eliminate a tremor or increase precision) [8 ]. The learning curve was also reduced as it allowed surgeons to transfer skills used during open surgery to the robotic approach in 1 12 cases vs the 8 1 needed for laparoscopic transition [8 ]. This is supported by [18 ] who suggested that the learning curve for robotic prostatectomy was 1 2 cases compared with 5 6 for laparoscopic. However, Steinberg [28 ] devised a theoretical model which approximated an average learning curve of 77 cases (range 24 36) at an average cost of US $ (range $95 $1.5 million). Limitations highlighted by the authors were a lack of differentiation between learning curves for individual surgeons vs teams working together and the use of charge data rather than direct costs. They did, however, analyze published series which gave a learning curve ranging from 13 to 2 cases at a cost ranging from US $ They also did not assess clinical outcomes such as erectile function. Ahlering and colleagues [26 ] showed that after a learning curve of 4 procedures, both open and robotic radical prostatectomy had equivalent outcomes but robotic surgery led to significantly reduced length of stay (1.2 (.75 4) vs 2.12 (2 8) days, P <.1) as well as blood loss (13 ml (25 4 ml) vs 418 ml (15 12 ml), P <.1). Further attempts to reduce the learning curve associated with robotic surgery have included the use of live animal models and virtual reality simulators [29 ]. Ethical issues surround the former in some countries and thus research is currently being undertaken to establish methods which can deliver surgical training in robotic surgery that meet ethical standards [1 ]. COST EFFICIENCY Lotan [19 ] devised a model using cost data from a large county hospital as well as Anaesthetic Medication Laboratory Room and board Length of stay (days) Nerve sparing % n Lymphadenectomy % n Conversion to open/ laparoscopic/robot Length of surgery (min) x 78 x x x x 2 x x 72 x x x x 14 x x 15 x x x x 16 x 578 x x x x x x x 62 x x x x x x x 62 x x x x x x x x x x x 2 x x x x x x x 1 x x x x x x x 1 x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x 111 x x x 1.6 x x x x x 111 x x x 2.73 x x x x x 111 x x x x x x x x x (49) 1.76 (.83) x x x (758) 1.56 (1.53) x x x (678) 2.51 (1.37) 9 1 x x x x x x x 1 5 x x x x x x x 6 12 x x x x x x x x x x x 1.2 x x x x x x x x x 8.7 x x x x x 2. x x x x x x x x x 3. x x x x x x x x x 2. x x x x x x x x x 3. x x x x x Blood loss (ml) BJU INTERNATIONAL

8 ECONOMIC IMPACT OF ROBOTIC SURGERY IN UROLOGY FIG. 3. Fallacies in assessment of robotic technology [7 ]. Technology Assessment Parameters Clinical Outcome ONLY fallacy Technology specific questions? Is robotics overcoming the problems related to reduced dexterity & impaired visual control? Reduction in post surgical pain levels? cost of each operation was over $1 more than the laparoscopic procedure whilst open prostatectomy was the cheapest. The operating room costs [ US $2798 (range ) ] contributed to the increased cost of robot assisted procedures as these took longer to perform than the other two types of procedure. The equipment (US $215 fixed cost) necessary to perform robotic prostatectomy also increased the cost of the procedure [2 ]. A limitation of this study was that postoperative complications and long term follow-up were not measured. The study also did not give any measure of cost effectiveness ratios and cost per quality of life. Fallacies in assessment of Robotic technology a literature search to assess robot purchase costs, length of stay and operating time. They compared robot-assisted, laparoscopic and open prostatectomy. The findings suggested that open radical prostatectomy was the least expensive procedure [ US $6473 (range $ ) ], whilst robotic prostatectomy was the most expensive [ US $1 269 (range $ ) ]. The cost of purchase and maintenance of the robot was $857 per day [ 19 ]. Length of surgery (14 vs 16 min) was shortest in robotic surgery as was length of stay (1.2 vs 2.5 days) ( Table 2 ) [3 ]. This study identified the price for the robot (US $1.4 million), maintenance cost ($15 /year) and equipment prices ($533) as areas to target for cost efficiency [ 19 ]. They stated that if the cost of robotic equipment was similar to laparoscopic, then due to the reduced length of operation as well as hospital stay, robotic surgery would be cost equivalent to open surgery [19 ]. Limitations of this study include that the cost of the learning curve was not assessed (it was eliminated as a variable by using experienced surgeons) and nor were long term complications or recurrence. They also assumed that a minimum of one robot-assisted procedure was performed per week. Learning curve FORALL fallacy Availability of long-term survival/quality of life parameters? Do we know factors affecting learning curve (for surgeons with open and laparoscopic experience)? [31 ] compared robot assisted prostatectomy with open prostatectomy in the general hospital and specialist urology centre settings. Their base case model revealed a cost premium of US $783 when comparing robotic with open surgery at the specialist level (the premium dropped to US $195 when comparing robotic with open surgery in the general setting). However, based on this model robot assisted prostatectomy could be as cost effective (US $8929 vs $8734) as open in the specialist setting if case volume was increased to 1/week. Robot assisted prostatectomy may be even less expensive (as cost of maintenance would decrease from US $12 to $5 for each case) in the general setting if length of stay was 1.5 days and case volume increased to 14/ week [ 31 ]. Again, a limitation of this model was that it requires a high case load which is not necessarily reflective of all centres. The cost of the learning curve was also not discussed. [2 ] compared robot assisted prostatectomy with laparoscopic and open prostatectomy (643 patients). They found that even when the cost of purchasing and maintaining the robot were excluded, the [32 ] suggested that costs associated with robotic and laparoscopic surgery were greater than those of using an open approach ( P <.5). This was supported by [2 ] who reported a difference of almost US $1 [mean cost is US $ ($ ) ] between robot assisted and open retropubic prostatectomy. However, [2 ] used charge rather than direct costs. The former are higher as profit margins are included. Like [32 ], others have reported that a similar increased cost of instruments accounts for most of the increased charges associated with robotic surgery. This is supported by [18 ] and [33 ] who showed that the increased cost of laparoscopic surgery could be largely attributed to the cost of the instruments. It is reasonable to conclude that the main barriers to cost efficiency in robotic surgery include the high initial cost, annual maintenance fee and limited number of uses per instrument. There is some evidence that the price of laparoscopic instruments is rising [3 ]. This may allow robotic surgery to become a viable alternative if simultaneously the above costs can be reduced. Nearly 6 cases of robotic cystectomy have been performed world wide [24 ] and the International Robotic Cystectomy Consortium has collected data on over 1 patients. [24 ] recently published a study comparing open radical cystectomy with robotic surgery. They found that after adjusting for variable operating room and hospitalization costs, the fixed costs of robotic surgery are US $1634 higher. [34 ], on the other hand, suggest that the higher fees for robotic cystectomy were 212 BJU INTERNATIONAL 1551

9 AHMED ET AL. TABLE 3 An overview of postoperative outcomes in prostate surgery Authors Lotan [19] [31 ] [2 ] [32 ] Lotan [42 ] [33 ] [2 ] [18] [35] Lowrance [22] Hu [23] Postoperative outcomes Mean blood Oncological Sexual Urinary Type of surgery Post-anaesthetic care costs transfusion (units) margins (%) dysfunction (%) symptoms (%) Other outcomes Learning curve (cases) LRP x x x x x x x RARP x x x x x x x ORP x x x x x x x RARP 295 x x x x x x ORP Specialist setting 419 x x x x x x ORP Community setting 419 x x x x x x ORP- Retropubic x x x x x x x ORP- Perineal x x x x x x RARP x x x x x x x ORP- Retropubic x x x x x x x ORP- Perineal x x x x x x x RARP x x x x x x x CAP x x x x x x x ORP x x x x x x LRP x x 11 3 x x x 1 RARP x x x x 13 2 ORP x x x x x x x LRP x x x x x x x RARP x x x x x x x LRP x x x x x x x RARP x x x x x x x ORP x x x x x x x RARP x x x 1 2 LRP x x x 5 6 RALP x x x x x 1 urinary x leak LP x x x x x x x LRP x x 11.6 x 35.4 x x ORP x x 9.2 x 4.3 x x MIRP x x x x x RRP x x x x x offset by the reduced length of stay. [27 ] found that the direct relative cost performance of robotic cystectomy was US $ higher than open cystectomy. However, the reduced median length of stay with robotic cystectomy (5.5 days) compared with open cystectomy (8 days), most significantly contributed to lowering costs and, coupled with lower complication rates, resulted in robotic cystectomy being more cost efficient for some procedures, such as continent cutaneous diversion and ileal conduit. Other operations where robotic surgery is gaining popularity are partial nephrectomy, pyeloplasty and paediatric urological procedures [7 ]. For instance, [35 ] compared laparoscopic with robotic pyeloplasty and found that the latter demonstrated no improvement in clinical outcomes but was 2.7 times more expensive. Based on their findings, the laparoscopic procedure must take up to 6.5 h longer in order for the cost to become comparable with the robotic pyeloplasty approach. However as with prostatectomy and cystectomy, long term follow-up is critical in evaluating robotic surgery. So far all the studies suggest that robotic prostatectomy is more expensive than laparoscopic and open equivalents ( Tables 2 and 4 ). Breakdown of these costs suggests that expense is incurred in the cost of purchasing the robot, maintenance and instruments. At present, reduced length of stay in hospital and length of surgery are unable to compensate for the excess costs of robotic surgery. There is a lack of studies describing cost effectiveness ratios or cost per quality of life. An attempt has been made by Qing-Wang [36 ] from the London School of Economics and King s College London to suggest a solution. As there have been very few RCTs comparing robotic pelvic surgery with other approaches with respect to costs, quality of life, immediate and long term outcomes, it is important to devise a method to ascertain BJU INTERNATIONAL

10 ECONOMIC IMPACT OF ROBOTIC SURGERY IN UROLOGY TABLE 4 Cost related equipment and peri-operative outcomes of robotic cystectomy Peri-operative outcomes Authors Type of surgery Total cost Robot cost included Y/N Operative charges Nonoperative charges Number of patients Professionals Surgical equipment Operating room [24] RARC Y x x 2 x x x ORC x x 2 x x x [34] OC IC Y x OC CCD x OC ON x RC IC x RC CCD x RC ON x the overall value of robotic surgery. Decision analytic modelling can provide this solution but there must be clear outcome measures and suitable comparators [36 ]. A decision tree can then be made to assist in analyzing and processing data. However, for models to be accurate the best evidence on clinical pathways, possibility of decline in cost of technology and long term outcomes need to be known. Two studies included in this review which utilized decision tree analysis are those of Lotan [19 ] and [ 27 ]. However these were limited by the lack of higher level studies, data on long term outcomes as well as quality of life indicators. DISCUSSION Both robotic and laparoscopic urological surgery appear to be superior to open surgery with respect to blood loss and hospital stay. However, robotic surgery is currently more expensive than the other two approaches due to capital cost, maintenance of the robot and limited life of the instruments. This cost may be offset by a shorter learning curve and competition for purchase contracts once the patents limited to the sole manufacturer expire. However this is not guaranteed. In addition, the operating theatre cost can add to the expense, although, experienced robotic teams are often quicker than their laparoscopic and open counterparts, thus compensating for the initial increase in operating room time and set-up costs. Although the cost analyses differ across various studies, hospital and health systems, the fact that robotic procedures are more expensive than open will not change, regardless of how the calculations are designed [9,37 ]. As questioned by Graefen, are these extra costs justified? [9 ] The answer is possibly yes. First, if long term advantages for robotic procedures are reported; but this is the not currently the case. Second, if an improved learning curve is established by long term, multicentre studies then more urologists will acquire high quality skills rapidly and patient outcomes will improve. Third, if the cost of the operating theatre equipment and robots can be reduced. Finally, a prospective analysis of outcomes needs to be carried out in RCTs. Most of the existing trials have not included long term follow-up of clinical outcomes, have failed to mention additional costs associated with conversion of laparoscopic to open procedures (occurs in up to 14%), the huge costs of individual disasters requiring prolonged ITU and hospital stays and the key issue of training ( Fig. 3 ). This combined with reduced length of stay, postoperative complications as well as easier training in robotic surgery may provide an advanced, cost effective surgical tool. Accurate comparison of cost was difficult to assess as few studies gave itemized details of expenditure for each operative technique. In order to compare the outcomes of robotic surgery against other techniques we looked at the oncological control, continence and sexual function as these are often considered the most important indicators of quality of life postoperatively after radical prostatectomy [38 ]. However, problems arose in gathering enough data for oncological control in robotic and laparoscopic procedures as long term results are currently lacking or preliminary. Surgical training constantly has to try and keep abreast of innovation. The reduction of working hours in Europe makes this even more difficult to achieve. Thus technological advances such as robotic surgery which have a faster learning curve are more likely to be adapted in the face of limited funding, time and resources. The cost of training a surgeon to perform laparoscopic prostatectomy or cystectomy needs to be compared directly with that of training in robotic surgery in order for the true overall costs to be evaluated [39 ]. We conclude that the three areas where robotic surgery can be targeted for cost efficiency are cost and maintenance of the robot as well as improving the training of surgeons [ 2 ]. If robotic surgery were to yield better quality of life after discharge from hospital, this can have an impact on cost per quality of life. While this may be of little importance to direct hospital costs, it is highly relevant to society as a whole. A patient who returns to work sooner adds to the cost effectiveness of his or her mode of surgery. Health economic modelling needs to take this into consideration for future studies into this growing area of surgery. FUTURE PROSPECTS The main obstacle to wider adoption of robotic surgery is cost. If the initial cost of the robot, maintenance and instruments were reduced, the da Vinci system could prove to be as cost effective as laparoscopic surgery [ 4 ]. The da Vinci system may also have a shorter learning curve which saves 212 BJU INTERNATIONAL 1553

11 AHMED ET AL. Anaesthetic Medication Laboratory Room and board Length of stay (days) Nerve sparing % n Lymphadenectomy % n Conversion to open/ laparoscopic/robot Length of surgery (min) x x x X 5 x x x 246 x x x x X 5 x x x 228 x 636 x x X 9 x x x 336. x x x X 8 x x x x x x X 7.8 x x x x 669 x x X 5.5 x x x 36. x x x X 5.8 x x x x x x X 5 x x x x Blood loss (ml) TABLE 5 An overview of postoperative outcomes in bladder surgery Author [24] [27 ] Postoperative outcomes Type of surgery Post-anaesthetic care costs ($) Mean blood transfusion (units) Oncological margins (%) Sexual dysfunction (%) Urinary symptoms (%) Other outcomes Learning curve (cases) RRC x.4 x x x x x ORC x 1.2 x x x x x OC IC x x x x x x Cost of robot and OC CCD x x x x x x maintenance OC ON x x x x x x per case is US RC IC x x x x x x $1 RC CCD x x x x x x RC ON x x x x x x money in terms of training and may reduce complications. Moreover, the patent on the da Vinci system is due to expire in a few years and there are plans to produce alternative technologies in future [41 ]. These have the advantages of being less bulky with improved precision and access. They are expected to provide value for money as there will be competition which may act to drive down prices. Further savings can be made as the price of instruments for robots reduces or if the frequency of their replacement is reduced. This is, however, only an assumption based on wider economic theories of competition in the free market, rather than specific evidence and thus is merely speculative. LIMITATIONS Although the quality scoring of the articles included in this review is within appropriate limits, there is a paucity of well conducted randomized trials. Of the material that exists, only a few papers provide an itemized breakdown of the costs involved in each of the types of surgical approaches. None of the studies provided long term follow-up and thus it was difficult to assess thoroughly postoperative outcomes. CONCLUSIONS The increasing cost of robotic surgery is associated with the initial purchase of the robot, its annual maintenance and cost of disposable equipment. The cost of robotic surgery in urology can be minimized by addressing issues such as capital costs, maintenance and the learning curve. Robotic procedures have better short term outcomes in terms of complications and hospital stay. Ever growing demands for robotic surgery should be accompanied with studies looking at the actual cost of the procedures as a whole which includes the patient journey within the hospital. Long term outcomes are needed to establish the effectiveness of robotic surgery in urology. ACKNOWLEDGEMENTS Prokar Dasgupta acknowledges financial support from the Department of Health via the National Institute for Health Research (NIHR) comprehensive Biomedical Research Centre award to Guy s & St Thomas NHS Foundation Trust in partnership with King s College London and King s College Hospital NHS Foundation Trust. He also acknowledges the support of the MRC Centre for Transplantation, London Deanery and Olympus. CONFLICT OF INTEREST None declared. REFERENCES 1 Ahmed K, Khan M, Vats A Current status of robotic assisted pelvic surgery and future developments. Int J Surg 29 ; 7 : BJU INTERNATIONAL

12 ECONOMIC IMPACT OF ROBOTIC SURGERY IN UROLOGY 2 C, Gupta A, Hotze T Cost comparison of robotic, laparoscopic, and open radical prostatectomy for prostate cancer. Eur Urol 2 1 ; 57 : Becker C. Nowhere to run, or hide. Current economic downturn may exacerbate hospitals weakness with buildings, technology costs, reimbursements. Mod Healthc 28 ; 38 : Haycox A. What is health economics? 2nd edn. London : Hayward Medical Communications, 29. Available at: painres/download/whatis/what_is_ health_econ.pdf. Accessed 11 February Hummel S, Paisley S, Morgan A, Currie E, Brewer N. Clinical and costeffectiveness of new and emerging technologies for early localised prostate cancer: a systematic review. Health Technol Assess 23 ; 7 : iii, ix-x, Hakimi A, Feder M, Ghavamian R. Minimally invasive approaches to prostate cancer: a review of the current literature. Urol J 27 ; 4 : Dasgupta P. Urologic Robotic Surgery in Clinical Practice, 1st edn. London : Springer, 28 : Rocco B, Djavan B. Robotic prostatectomy: facts or fiction? Lancet 27 ; 369 : Graefen M. Editorial comment on cost comparison of robotic, laparoscopic and open radical prostatectomy for prostate cancer. Eur Urol 21 ; 57 : Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 29 ; 339 : b Jadad AR, Moore RA, Carroll D Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996 ; 17 : EBM _ Level _ of _ Evidence. Oxford Centre for Evidence-based Medicine Levels of Evidence (March 29). 29. Available at: aspx?o=125. Accessed 11 February Slim K, Nini E, Forestier D, Kwiatkowski F, Panis Y, Chipponi J. Methodological index for nonrandomized studies (minors): development and validation of a new instrument. ANZ J Surg 23 ; 73 : EBM. Oxford Centre for Evidence-based Medicine. 29. Available at: cebm.net/index.aspx?o=125. Accessed 19 March Ficarra V, Cavalleri S, Novara G, Aragona M, Artibani W. Evidence from robot-assisted laparoscopic radical prostatectomy: a systematic review. Eur Urol 27 ; 51 : 45 55, discussion 6 16 Ficarra V, Novara G, Artibani W Retropubic, laparoscopic, and robotassisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol 29 ; 55 : Menon M, Shrivastava A, Kaul S Vattikuti Institute prostatectomy: contemporary technique and analysis of results. Eur Urol 27 ; 51 : , discussion C á ceres F, S á nchez C, Mart í nez- Pi ñ eiro L [Laparoscopic radical prostatectomy versus robotic ]. Arch Esp Urol 27 ; 6 : Lotan Y, Cadeddu J, Gettman M. The new economics of radical prostatectomy: cost comparison of open, laparoscopic and robot assisted techniques. J Urol 24 ; 172 : S, Atug F, Castle E, Davis R, Thomas R. Cost analysis of radical retropubic, perineal, and robotic prostatectomy. J Endourol 26 ; 2 : Lowrance WT, Tarin TV, Shariat SF. Evidence-based comparison of robotic and open radical prostatectomy. Scientific World Journal 21 ; 1 : Lowrance WT, Elkin EB, Jacks LM Comparative effectiveness of prostate cancer surgical treatments: a population based analysis of postoperative outcomes. J Urol 21 ; 183 : Hu JC, Gu X, Lipsitz SR Comparative effectiveness of minimally invasive vs open radical prostatectomy. JAMA 29 ; 32 : A, Kurpad R, Lal A, Nielsen M, Wallen E, Pruthi R. Cost analysis of robotic versus open radical cystectomy for bladder cancer. J Urol 21 ; 183 : Nix J, A, Kurpad R, Nielsen M, Wallen E, Pruthi R. Prospective randomized controlled trial of robotic versus open radical cystectomy for bladder cancer: perioperative and pathologic results. Eur Urol 21 ; 57 : Ahlering T, Woo D, Eichel L, D, Edwards R, Skarecky D. Robot-assisted versus open radical prostatectomy: a comparison of one surgeon s outcomes. Urology 24 ; 63 : R, Ng CK, Shariat SF The economics of robotic cystectomy: cost comparison of open versus robotic cystectomy. BJU Int 211 ; 18 : Steinberg PL, Merguerian PA, Bihrle W, Seigne JD. The cost of learning robotic-assisted prostatectomy. Urology 28 ; 72 : Pierorazio P, Allaf M. Minimally invasive surgical training: challenges and solutions. Urol Oncol 29 ; 27 : Lotan Y. Economics of robotics in urology. Curr Opin Urol 21 ; 2 : CJ, Jones P, Eisenstein E, Preminger G, Albala D. Local cost structures and the economics of robot assisted radical prostatectomy. J Urol 25 ; 174 : V, Nosnik I, Sun L Financial comparative analysis of minimally invasive surgery to open surgery for localized prostate cancer: a single-institution experience. Urology 27 ; 69 : JV, Leonhardt A, Hitendra R, Patel H. The cost of radical prostatectomy: retrospective comparison of open, laparoscopic, and robot-assisted approaches. J Robotic Surg 28 ; 2 : R, Ng C, M. Open versus robotic cystectomy: re-examining the cost comparison. J Urol 29 ; 181 : RE, Bhayani SB, Kavoussi LR. A prospective comparison of robotic and laparoscopic pyeloplasty. Ann Surg 26 ; 243 : Wang Q, Armstrong D, McGuire A. Health Economics in Robotic Surgery. In Dasgupta P. Urologic Robotic Surgery in Clinical Practice, 1st edn. London : Springer. 28 : Barbash GI, Glied SA. New technology and health care costs the case of robot-assisted surgery. N Engl J Med 2 1 ; 363 : Gettman M, Blute M. Critical comparison of laparoscopic, robotic, and 212 BJU INTERNATIONAL 1555