- Email: [email protected]
Use, Costs and Comparative Effectiveness of Robotic Assisted, Laparoscopic and Open Urological Surgery
Use, Costs and Comparative Effectiveness of Robotic Assisted, Laparoscopic and Open Urological Surgery Hua-yin Yu, Nathanael D. Hevelone, Stuart R. Lipsitz, Keith J. Kowalczyk and Jim C. Hu* From the Division of Urology (HY, KJK, JCH), and Center for Surgery and Public Health (NDH, SRL, JCH), Brigham and Women’s/Faulkner Hospital, Harvard Medical School, Boston, Massachusetts
Abbreviations and Acronyms DS ⫽ data suppressed per Nationwide Inpatient Sample for 0⬍n⬍11 LOS ⫽ length of hospital stay LS ⫽ laparoscopic surgery OS ⫽ open surgery NIS ⫽ Nationwide Inpatient Sample RALS ⫽ robotic assisted laparoscopic surgery Submitted for publication July 23, 2011. Nothing to disclose. Supported by the American Urological Association Foundation (HY) and the U.S. Department of Defense Physician Training Award W81XWH08-1-0283 (JCH). Supplementary material can be obtained at www.jurology.com. * Correspondence: Brigham and Women’s/ Faulkner Hospital, 1153 Centre St., Suite 4420, Boston, Massachusetts 02130 (telephone: 617983-4570; FAX: 617-983-7945; e-mail: jhu2@ partners.org).
Editor’s Note: This article is the fifth of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 1514 and 1515.
Purpose: Although robotic assisted laparoscopic surgery has been aggressively marketed and rapidly adopted, there are few comparative effectiveness studies that support its purported advantages compared to open and laparoscopic surgery. We used a population based approach to assess use, costs and outcomes of robotic assisted laparoscopic surgery vs laparoscopic surgery and open surgery for common robotic assisted urological procedures. Materials and Methods: From the Nationwide Inpatient Sample we identified the most common urological robotic assisted laparoscopic surgery procedures during the last quarter of 2008 as radical prostatectomy, nephrectomy, partial nephrectomy and pyeloplasty. Robotic assisted laparoscopic surgery, laparoscopic surgery and open surgery use, costs and inpatient outcomes were compared using propensity score methods. Results: Robotic assisted laparoscopic surgery was performed for 52.7% of radical prostatectomies, 27.3% of pyeloplasties, 11.5% of partial nephrectomies and 2.3% of nephrectomies. For radical prostatectomy robotic assisted laparoscopic surgery was more prevalent than open surgery among white patients in high volume, urban hospitals (all p ⱕ0.015). Geographic variations were found in the use of robotic assisted laparoscopic surgery vs open surgery. Robotic assisted laparoscopic surgery and laparoscopic surgery vs open surgery were associated with shorter length of stay for all procedures, with robotic assisted laparoscopic surgery being the shortest for radical prostatectomy and partial nephrectomy (all p ⬍0.001). For most procedures robotic assisted laparoscopic surgery and laparoscopic surgery vs open surgery resulted in fewer deaths, complications, transfusions and more routine discharges. However, robotic assisted laparoscopic surgery was more costly than laparoscopic surgery and open surgery for most procedures. Conclusions: While robotic assisted and laparoscopic surgery are associated with fewer deaths, complications, transfusions and shorter length of hospital stay compared to open surgery, robotic assisted laparoscopic surgery is more costly than laparoscopic and open surgery. Additional studies are needed to better delineate the comparative and cost-effectiveness of robotic assisted laparoscopic surgery relative to laparoscopic surgery and open surgery. Key Words: robotics, urologic surgical procedures, laparoscopy, treatment outcome, costs and cost analysis
1392
www.jurology.com
0022-5347/12/1874-1392/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
AND
RESEARCH, INC.
Vol. 187, 1392-1399, April 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.11.089
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
WITH more than 1,400 robotic surgical systems installed in United States hospitals, with some having up to 5 systems, and the number of robotic systems in other countries doubling from 200 to 400 between 2007 and 2009,1 robotic assisted laparoscopic surgery has been rapidly adopted without population based evidence demonstrating superior outcomes compared to laparoscopic surgery and open surgery. As a result the Institute of Medicine has prioritized RALS for comparative effectiveness research.2 Direct to consumer advertising has fueled patient demand for RALS,3 particularly for radical prostatectomy. However, men who underwent radical prostatectomy with RALS vs OS were more likely to be diagnosed with incontinence and erectile dysfunction, and more likely to experience treatment regret.4,5 Most existing studies that demonstrate better outcomes with RALS are single surgeon series, whereby investigators may receive educational or research funding from the device manufacturer (Intuitive Surgical®, Sunnyvale, California). Moreover estimates of RALS use are provided primarily by the manufacturer.1,6 In this study we characterize population based RALS use and patterns of care for urological procedures, and compare perioperative costs and outcomes with LS and OS.
MATERIALS AND METHODS Data Source Subjects were identified from the HCUP (Healthcare Cost and Utilization Project) NIS (Nationwide Inpatient Sample), sponsored by the Agency for Healthcare Research and Quality.7 NIS is a 20% stratified probability sample that encompasses approximately 8 million acute hospital stays from more than 1,000 hospitals in 42 states per year. It is the largest all-payer inpatient care observational cohort in the United States and represents approximately 90% of all hospitalizations.
Study Cohort During the last quarter of 2008 there were 2,093,300 hospitalizations within the NIS. Using NIS discharge weights these represent more than 9.8 million patients. We used the ICD-9 code 17.4x for RALS, initiated on October 1, 2008, to identify RALS procedures approved by the U.S. Food and Drug Administration. RALS were comprised of 64% urological, 32% gynecologic, 2% cardiac and 2% general surgical procedures. Procedures with a laparoscopic designation (ICD-9 54.21, 54.51) were classified as LS, while those without LS or RALS designations/codes were classified as OS. To adequately power analyses we analyzed urological procedures with 40 or more unweighted procedures including radical prostatectomy (ICD-9 60.5), nephrectomy (ICD-9 55.51, 55.52, 55.54), partial nephrectomy (ICD-9 55.4) and pyeloplasty (ICD-9 55.87).
Covariates For each procedure we examined hospital and patient level characteristics that may be associated with out-
1393
comes. Hospital characteristics included U.S. census region, urban vs rural location, teaching status and bed size. Hospital surgical volume was assessed by stratifying each procedure into high, intermediate and low volume tertiles to minimize cell counts of less than 11 (for which DS is required per NIS). Patient level characteristics include age, number of comorbidities, race, median income and primary payer (private vs government health plans).
Outcomes ICD-9 diagnosis and procedure codes were used to identify blood transfusions and complications (cardiac, respiratory, genitourinary, vascular, wound, miscellaneous medical and miscellaneous surgical).4 NIS specific outcomes included death, hospital LOS, discharge disposition (routine [home] vs other [rehabilitation, skilled nursing facility, etc]) and total costs. Costs were derived from total charges using the HCUP cost-to-charge ratio.
Statistical Analysis Use of RALS, LS and OS during the study period was characterized for each procedure. Stratification, clustering and survey weights were used in accordance with NIS sampling. Because characteristics of subjects undergoing RALS and LS differed from those undergoing OS, propensity scoring methods were used to adjust for potential bias associated with selection for open vs minimally invasive procedures.8 This approach controls for factors that may confound group assignment and outcomes by adjusting discharge weights, with the goal of balancing characteristics among groups. Adjustments were conducted using multivariate logistic regression models to calculate the propensity of undergoing RALS, LS or OS based on all covariates described, and weighted by the inverse propensity of being in one of the treatment groups.9 Propensity scoring models were constructed for each of the 4 procedures examined. Balance of covariates across procedures was verified following propensity adjustment. Linear and logistic regressions were used to examine univariable and multivariable effects of surgical approach on outcomes. LOS comparisons were modeled using log-normal linear regression. All analyses were performed with SAS® version 9.2 and all tests were considered statistically significant at p ⱕ0.05.
RESULTS Procedure Frequencies The relative use of RALS, LS and OS is shown in the figure. Radical prostatectomy was the only procedure in which RALS (52.7%) was more prevalent than OS (44.4%) and LS (2.8%) combined. OS was the predominant surgical approach for all other procedures. LS was least prevalent among all procedures except laparoscopic nephrectomy, in which RALS was least prevalent. Characteristics of Study Sample Patient and hospital characteristics are shown in table 1. Propensity adjustment balanced covariates for all procedures except for pyeloplasty age and hospital
1394
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
100% 90% 79.4%
79.0%
80% 70% 60%
66.1% 52.7%
50%
44.4%
40% 27.3%
30% 20% 10%
2.8%
0%
)
2 58
1,
2 n=
y( m to
l ca
di
Ra
c te ta os Pr
2.3%
8)
)
4 83
5 ,8
1, n= y(
=3
n y( om ct
st
la
op
el Py
18.3%
11.5% 9.5%
6.6%
Roboc Laparoscopic Open
re
ph
r Pa
a
e lN
y m to ec hr p Ne
) 16 ,2 14 = (n
spiratory complications (all p ⱕ0.032). For partial nephrectomy RALS was associated with the fewest genitourinary, wound and vascular complications (all p ⱕ0.015). For pyeloplasty LS was associated with the fewest miscellaneous surgical complications (p ⫽ 0.049). Health care costs were higher for RALS vs LS and OS (all p ⱕ0.022) for all procedures except for partial nephrectomy, where costs were similar (table 3). OS was least costly for radical prostatectomy, whereas costs were lowest with LS for nephrectomy and pyeloplasty (all p ⱕ0.022).
DISCUSSION Use of robotic assisted, laparoscopic and open urological procedures.
type due to low procedure numbers. White patients were more likely than nonwhites to undergo RALS or LS vs OS for radical prostatectomy (p ⫽ 0.002). Urban vs rural hospitals were more likely to perform RALS for radical prostatectomy, partial nephrectomy and pyeloplasty (all p ⱕ0.015). High and intermediate vs low volume hospitals were more likely to perform RALS for radical prostatectomy (p ⬍0.001). However, high volume hospitals were more likely to perform OS vs RALS for pyeloplasty (p ⫽ 0.033). Use of RALS for nephrectomy and partial nephrectomy was more prevalent in the Midwest, whereas hospitals in the South were more likely to perform OS for both procedures (both p ⱕ0.014). Outcomes Results from unadjusted and propensity adjusted analyses were largely similar and, therefore, outcomes from adjusted analyses are presented (table 2). RALS and LS vs OS for radical prostatectomy and nephrectomy were associated with lower in-hospital mortality (both p ⬍0.001). Similarly RALS and LS vs OS were associated with fewer transfusions for radical prostatectomy (p ⬍0.001) while LS vs RALS and OS was associated with fewer transfusions for nephrectomy and pyeloplasty (both p ⬍0.001). For all procedures RALS and LS vs OS were associated with shorter LOS (all p ⱕ0.002). Additionally, length of stay was shortest with RALS for radical prostatectomy and partial nephrectomy, whereas LS was associated with the shortest LOS for nephrectomy (all p ⬍0.001). RALS and LS vs OS for partial nephrectomy was associated with more routine home discharges (p ⫽ 0.006). RALS and LS vs OS for radical prostatectomy were associated with fewer respiratory, wound and vascular complications (all p ⱕ0.024). For nephrectomy RALS was associated with the fewest cardiac complications and LS was associated the fewest re-
Minimally invasive surgery often offers distinct, consistently reproducible advantages compared to open approaches including smaller incisions, reduced intraoperative blood loss, decreased postoperative pain, and shorter hospital LOS and convalescence.10 –12 Moreover advantages of RALS vs LS include a 3-dimensional view of the operative field, the absence of a fulcrum effect, 7 vs 4 df of movement with wristed instruments that facilitate intracorporeal suturing, elimination of surgeon tremor and ergonomic benefits. Thus, RALS purportedly facilitates the learning curve for open surgeons transitioning to minimally invasive surgery.11,13,14 Many patients intuitively perceive minimally invasive approaches as reducing complications compared with conventional OS, and prefer them due to smaller incisions requiring fewer analgesics and shorter hospital stays, even at a greater cost.15 However, rapid adoption, prolonged learning curves, and varying hospital accreditation practices for attaining LS and RALS privileges may result in unforeseen risks. For example, the rapid adoption of laparoscopic cholecystectomy in the 1990s resulted in a spike in biliary tract injuries from 1,500 to 4,000 per year.16 Studies comparing the 3 surgical approaches are sparse and mostly single surgeon series that are often conflicting even for radical prostatectomy, the most commonly performed RALS procedure.17 Population based comparisons allow characterization of use and outcomes of these surgical modalities in the absence of randomized controlled trials. Our study is the first population based comparative effectiveness study of RALS use, patterns of care and costs, and resulted in several important findings. While the majority of radical prostatectomies were performed via RALS, its use is not as high as estimates from the device manufacturer,6 but consistent with reported Medicare radical prostatectomy use of RALS and LS combined.4 We found racial, geographic and hospital based variations in the use of RALS, LS and OS for some but not all procedures. This is consistent with previous reports of limited access to
Table 1. NIS weighted unadjusted patient and hospital characteristics Radical Prostatectomy p Value
RALS
LS
OS
996 (8.7) 4,051 (35.2) 5,516 (47.9) 950 (8.3)
63 (10.2) 277 (44.8) 249 (40.3) 29 (4.7)
764 (7.9) 3,597 (37.1) 4,449 (45.9) 894 (9.2)
101 (30.8) 45 (13.7) 74 (22.4) 0.262 109 (33.1)
907 (34.9) 508 (19.6) 527 (20.3) 654 (25.2)
3,525 (31.2) 2,468 (21.9) 2,744 (24.3) 2,553 (22.6)
7,948 (69.0) 398 (64.5) 1,727 (15.0) 1,723 (28.0) 1,838 (16.0) 46 (7.5)
5,794 (59.7) 1,677 (17.3) 2,233 (23.0)
213 (64.8) 45 (13.7) 0.002 71 (21.6)
1,473 (56.8) 502 (19.3) 621 (23.9)
6,801 (60.2) 2,306 (20.4) 2,184 (19.3)
7,647 (66.4) 3,242 (28.2) 624 (5.4)
440 (71.3) 144 (23.4) 33 (5.3)
6,137 (63.3) 2,877 (29.6) 690 (7.1)
135 (41.0) 134 (40.7) 0.202 60 (18.3)
1,190 (45.9) 1,004 (38.7) 402 (15.5)
5,215 (46.2) 4,327 (38.3) 1,749 (15.5)
1,575 (13.9) 2,743 (24.3) 2,973 (26.3) 4,001 (35.4)
104 (17.5) 108 (18.1) 138 (23.1) 247 (41.3)
1,791 (19.2) 2,237 (23.9) 2,374 (25.4) 2,943 (31.5)
83 (25.7) 90 (27.9) 86 (26.6) 65 (19.9)
553 (21.8) 643 (25.4) 629 (24.8) 707 (27.9)
2,520 (22.7) 2,858 (25.7) 3,015 (27.1) 2,720 (24.5)
229 (2.0) 3,582 (31.1) 7,702 (66.9)
DS 685 (7.1) 100 (16)* 2,691 (27.7) 510 (83)* 6,328 (65.2)
25 (7.5) 77 (23.4) 0.015 227 (69.1)
111 (4.3) 646 (24.9) 1,839 (70.8)
504 (4.5) 3,296 (29.2) 7,491 (66.4)
851 (7.4) 1,637 (14.2) 9,025 (78.4)
31 (5.0) 234 (37.9) 353 (57.2)
727 (7.5) 2,236 (23.0) 6,741 (69.5)
15 (4.7) 74 (22.4) 0.106 240 (72.9)
187 (7.2) 426 (16.4) 1,984 (76.4)
910 (8.1) 2,028 (18.0) 8,353 (74.0)
2,617 (22.7) 4,587 (39.8) 4,309 (37.4)
51 (8.2) 314 (51.0) 252 (40.8)
4,691 (48.3) 95 (28.8) 2,556 (26.3) 173 (52.7) 2,457 (25.3) ⬍0.001 61 (18.5)
785 (30.2) 808 (31.1) 1,004 (38.7)
3,890 (34.5) 3,791 (33.6) 3,610 (32.0)
2,352 (20.4) 3,266 (28.4) 3,834 (33.3) 2,061 (17.9)
105 (17.0) 107 (17.4) 297 (48.2) 108 (17.4)
1,684 (17.4) 2,057 (21.2) 3,625 (37.4) 2,337 (24.1)
51 (15.5) 152 (46.3) 67 (20.3) 0.596 59 (18.0)
456 (17.6) 848 (32.7) 628 (24.2) 663 (25.6)
2,013 (17.8) 2,489 (22.0) 4,221 (37.4) 2,568 (22.7)
4,412 (38.3) 4,448 (38.6) 2,652 (23.0)
220 (35.7) 269 (43.7) 127 (20.7)
3,597 (37.1) 3,515 (36.2) 2,592 (26.7)
76 (23.1) 108 (32.8) 0.080 145 (44.1)
865 (33.3) 657 (25.3) 1,074 (41.4)
3,182 (28.2) 3,016 (26.7) 5,093 (45.1)
0.088
RALS
LS
Partial Nephrectomy OS
p Value
RALS
LS
OS
0.222
155 (34.8) 107 (24.1) 113 (25.5) 70 (15.7)
106 (29.0) 113 (31.0) 59 (16.3) 86 (23.7)
931 (30.5) 797 (26.1) 794 (26.0) 527 (17.3)
0.621
308 (69.3) 36 (8.1) 100 (22.6)
232 (63.8) 82 (22.4) 51 (13.9)
0.980
268 (60.3) 113 (25.5) 63 (14.3)
Pyeloplasty p Value
RALS
LS
OS
p Value
310 (72)* 80 (19)* 30 (7)* 0.419 DS
80 (76)* 20 (19)* 0 DS
845 (80.8) 69 (6.6) 88 (8.4) 44 (4.2) ⬍0.001
2,011 (66.0) 611 (20.0) 427 (14.0)
243 (56.3) 62 (14.5) 0.200 127 (29.3)
63 (60.9) 20 (19.3) 21 (19.8)
602 (57.5) 240 (22.9) 205 (19.6)
0.601
220 (60.4) 125 (34.3) 20 (5.4)
1,753 (57.5) 941 (30.9) 354 (11.6)
315 (72.9) 35 (8.1) 0.376 82 (19.0)
80 (76)* DS 20 (19)*
615 (58.7) 85 (8.2) 346 (33.1)
0.089
0.737
70 (15.7) 121 (27.3) 122 (27.5) 131 (29.6)
44 (12.2) 61 (17.0) 99 (27.4) 156 (43.5)
577 (19.3) 662 (22.2) 758 (25.4) 989 (33.1)
85 (19.8) 84 (19.7) 161 (37.7) 0.297 98 (22.9)
20 (19.6) 25 (24.2) 34 (32.9) 24 (23.3)
217 (21.3) 298 (29.2) 216 (21.2) 289 (28.4)
0.335
0.649
0 176 (39.6) 269 (60.5)
0 60 (16.6) 304 (83.5)
131 (4.3) 0 635 (20.8) 113 (26.1) 2,283 (74.9) ⬍0.001 319 (74.0)
0 30 (28.4) 75 (71.6)
69 (6.6) 197 (18.8) 781 (74.6) ⬍0.001
0.887
DS 70 (16)* 370 (83)*
20 (5.5) 93 (25.6) 251 (69.0)
250 (8.2) 445 (14.6) 2,354 (77.2)
19 (4.3) 110 (25.5) 0.138 303 (70.2)
DS 20 (19)* 80 (75)*
62 (6.0) 210 (20.1) 774 (73.9)
0.830
0.252
150 (33.9) 114 (25.7) 180 (40.4)
96 (26.3) 204 (56.0) 65 (17.7)
1,056 (34.6) 966 (31.7) 1,027 (33.7)
125 (29.0) 205 (47.5) 0.095 102 (23.5)
40 (38.4) 49 (47.0) 15 (14.6)
365 (34.9) 277 (26.5) 404 (38.6)
0.033
0.014
64 (14.5) 252 (56.8) 62 (14.0) 65 (14.7)
105 (28.8) 106 (29.1) 59 (16.1) 95 (26.0)
755 (24.8) 104 (24.0) 519 (17.0) 153 (35.5) 1,103 (36.2) 82 (19.1) 672 (22.0) ⬍0.001 93 (21.5)
15 (14.2) 26 (24.6) 25 (23.8) 39 (37.5)
170 (16.2) 228 (21.8) 321 (30.7) 327 (31.3)
0.379
0.288
132 (29.8) 150 (33.7) 162 (36.5)
97 (26.7) 103 (28.2) 165 (45.1)
856 (28.1) 898 (29.5) 1,294 (42.5)
60 (57)* 30 (29)* DS
718 (68.7) 192 (18.3) 136 (13.0)
0.464
1395
Weighted counts using NIS complex survey weights, numbers may not sum to group totals or percents may not add to 100 due to need for rounding. * Number rounded to nearest 10 to prevent calculation of suppressed data, percent based on rounded number.
293 (67.9) 69 (15.9) 0.813 70 (16.3)
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
No. age (%): 50 or Younger 51–60 61–70 71 or Older No. race (%): White Nonwhite Missing No. primary payer (%): Private Medicare Medicaid/other No. ZIP Code income (%): 1st (lowest quartile) 2nd 3rd 4th No. hospital type (%): Rural Urban nonteaching Urban teaching No. hospital bed size (%): Small Medium Large No. hospital vol (%): Low Intermediate High No. hospital region (%): Northeast Midwest South West No. comorbidity (%): None 1 Multiple
Nephrectomy
0.015 0 0 DS 0.717 DS DS 16 (3.9) 0.813 DS 0 DS Not 0 0 0 applicable 0.260 DS DS 38 (9.1) 0.922 DS 0 DS ⬍0.001 2.2 (1.8) 2.2 (1.2) 3.0 (3.0) 0.006 166 (97.4) 29 (100.0) 407 (97.5) DS 262 (14.8) 89 (5.1) 0 DS 35 (16.8) 11 (5.1) 0 DS 40 (17.0) 22 (9.5) 0 (0.8) ⬍0.001 (4.6) 0.321 (2.6) 0.382 (0.2) ⬍0.001 64 399 225 19 (8.7) (2.5)
(3.6)
0 41 12 0 (0.4) (5.1) (2.0) 40 525 203 0 Vascular Miscellaneous medical Miscellaneous surgical Death
17 0
DS DS (0.7) (1.1) (1.2) (0.2) 73 110 119 23 No. complications (%): Cardiac Respiratory Genitourinary Wound
Any complication 864 (8.4) 68 (14.5) 869 (10.1) 0.096 71 (29.9) 755 (33.8) 3,616 (37.4) 0.218 46 (18.3) 55 (26.7) 512 (28.9) No. blood transfusion (%) 168 (1.6) DS 452 (5.2) ⬍0.001 56 (23.5) 165 (7.4) 1,366 (14.1) ⬍0.001 20 (8.1) 13 (6.1) 137 (7.8) Mean days LOS (SD) 1.7 (3.3) 2.0 (3.2) 2.4 (4.0) ⬍0.001 5.2 (13.2) 4.2 (7.9) 5.9 (15.6) ⬍0.001 2.8 (2.3) 3.6 (3.3) 4.5 (6.8) No. routine discharge (%) 9,770 (94.9) 443 (94.2) 8,101 (94.1) 0.658 219 (92.0) 1901 (85.2) 7,990 (82.6) 0.076 248 (97.9) 205 (99.2) 1,608 (90.9)
0 0
DS 28 (11.2) 18 (7.3) 0
0 DS 0
11 15 DS DS
(6.1) 472 (4.9) ⬍0.001 (8.0) 1,196 (12.4) 0.032 (6.1) 791 (8.2) 0.219 (1.4) 277 (2.9) 0.147 135 178 135 32 0 29 (12.2) 15 (6.3) DS (1.1) 0.154 (2.2) 0.024 (1.1) 0.203 (0.6) ⬍0.001 96 192 90 55
RALS p Value OS LS RALS
Radical Prostatectomy
Table 2. Propensity adjusted outcomes
17 (0.8) 218 (2.3) 0.138 442 (19.8) 1,873 (19.4) 0.880 102 (4.6) 725 (7.5) 0.060 13 (0.6) 139 (1.4) ⬍0.001
0
DS
16 0
DS DS DS DS DS 0.608 0.115 ⬍0.001 ⬍0.001 68 141 82 34 (5.3) (7.3)
OS LS LS
Nephrectomy
OS
p Value
RALS
Partial Nephrectomy
(3.9) (8.0) (4.6) (2.0)
p Value
RALS
0 0
LS
Pyeloplasty
OS
(3.7)
0.573 0.193 0.242 Not applicable 0.496 0.913 0.049 Not applicable 0.505 ⬍0.001 0.002 0.371
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
p Value
1396
care for nonwhite patients.18 Higher volume hospitals were more likely to offer RALS and LS for radical prostatectomy. However, they were more likely to offer OS for pyeloplasty, which may be necessary for complex cases with greater disease severity, such as those with prior surgical intervention that preclude the use of RALS and LS. Length of stay was also shorter for RALS and LS than OS for all procedures, with RALS having the shortest LOS for radical prostatectomy and partial nephrectomy. Moreover discharge to home vs rehabilitation or nursing facilities was more likely for RALS and LS vs OS for partial nephrectomy, and both were associated with fewer complications for radical prostatectomy, nephrectomy and partial nephrectomy than OS. Prior studies have demonstrated comparable or better outcomes with RALS vs LS or OS.19,20 Moreover RALS and LS were associated with fewer transfusions and, correspondingly, fewer in-hospital deaths. These advantages may be secondary to the 10⫻ to 12⫻ magnification inherent to RALS and LS, and the carbon dioxide insufflation that physiologically reduces venous ooze.11 The higher transfusion rate of nephrectomy with RALS may be related to the relative infrequency of use compared with RALS for the other 3 procedures and to the overall nephrectomy volume, which may reflect a learning curve effect. Despite the advantages mentioned, RALS was more costly than LS or OS. Similarly, others have demonstrated RALS to have $1,065 to $2,315 more in direct costs for radical prostatectomy and $535 to $1,651 more for partial nephrectomy, despite the shorter LOS.21,22 A recent study estimated that the incremental cost (without factoring in the purchase of the robotic surgical system) of RALS averages approximately $1,600, or an additional 6% per case more than OS.1 The cost differential between RALS and OS in our study for radical prostatectomy ($1,111) was more modest than this estimate. This may be due to lower costs associated with higher RALS volume for radical prostatectomy compared with other procedures, consistent with a study demonstrating 68% lower charges for high vs low volume minimally invasive radical prostatectomy surgeons.23 The incremental cost doubles when factoring in the $1.75 million capital acquisition cost of the robotic surgical system. This projects to $2.5 billion in annual U.S. health care expenditures if RALS replaces conventional surgical approaches.1 However, RALS use was unavailable due to the absence of billing designations for RALS at the time of that study, and costs were derived primarily from single surgeon series. Moreover because medical devices are not subject to the same level of scrutiny compared to pharmaceuticals in demonstrating clinical effectiveness,24,25 the rapid adoption of new medical technol-
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
1397
Table 3. Propensity adjusted costs Median Costs (IQR)
Radical prostatectomy Nephrectomy Partial nephrectomy Pyeloplasty
RALS
LS
OS
p Value
$10,804 ($8,289–$13,640) $13,894 ($11,291–$21,956) $15,724 ($10,398–$18,812) $11,829 ($8,577–$17,072)
$10,082 ($8,084–$12,631) $11,153 ($8,927–$14,884) $12,401 ($9,845–$16,042) $ 8,291 ($7,749–$11,830)
$ 9,693 ($7,600–$12,113) $12,548 ($9,377–$18,788) $11,817 ($9,334–$15,384) $ 9,520 ($7,666–$12,864)
0.007 ⬍0.001 0.442 0.022
ogies significantly contributes to spiraling health care costs.26 Pearson and Bach proposed a tiered payment system,27 whereby full reimbursement be rendered for therapies demonstrating superior outcomes to existing standards. Under this paradigm, in the absence of demonstrable cost-effectiveness, fiscal responsibility for the incremental cost of RALS shifts from insurers to patients. Our study must be interpreted within the context of the study design. Administrative data are designed for billing purposes and may lack detailed clinical information. We were unable to characterize tumor characteristics, disease severity or body mass index, which may affect patient selection and outcomes. For instance, surgeries for higher stage and grade disease may be technically more challenging and likely to require open surgical approaches. Therefore, tumor characteristics may be potential confounders for complications and transfusion outcomes. Additionally, the NIS contains hospital rather than physician level characteristics and may underestimate the frequency of some outcomes. We were unable to adjust for potential confounding due to surgeon attributes. The NIS is also limited to the inpatient hospital setting, and we were unable to assess outpatient complications and costs or earlier return to activities of daily living/ work, which may offset the higher costs of RALS. We evaluated generalized complications for many procedures. However, claims data have a high degree of corroboration with chart abstraction and are valid for detecting adverse outcomes.28 While we attempted to
adjust for confounding, there was missing race data. Moreover the propensity approach was unable to balance all pyeloplasty covariates and, thus, conclusions regarding race and for pyeloplasty must be interpreted with caution. Our RALS hospital costs exclude surgeon fees and robotic system acquisition and maintenance costs, and likely underestimate the true costs of RALS. While Intuitive Surgical currently enjoys a monopoly over RALS, likely contributing to higher expenses, costs should attenuate with the entry of competitors.29 Estimates are further limited by the lack of indirect cost data following discharge, including subsequent admissions or financial losses due to convalescence. It is possible that the higher direct costs of RALS may be offset by shorter convalescence periods which have been associated with radical prostatectomy.30 Finally, this is an observational study and there may be unobserved factors for which we were unable to adjust. While randomized control trials balance potential confounders, direct to consumer advertising, patient and surgeon preferences, and intra-surgeon heterogeneity in technique and outcomes regardless of surgical approach are barriers to implementing this research design. In summary, RALS has been rapidly adopted with disparities in use by race, hospital setting and geography for certain procedures. Despite greater costs, RALS has advantages such as fewer complications, fewer transfusions and shorter LOS. However, prospective studies are needed to assess the comparative effectiveness of RALS vs conventional surgical approaches.
REFERENCES 1. Barbash GI and Glied SA: New technology and health care costs–the case of robot-assisted surgery. N Engl J Med 2010; 363: 701. 2. Initial national priorities for comparative effectiveness research, list of priorities: Institute of Medicine of the National Academies 2009. Available at www.iom.edu/⬃/media/Files/Report%20Files/2009/ ComparativeEffectivenessResearchPriorities/Stand%20 Alone%20List%20of%20100%20CER%20Priorities% 20-%20for%20web.ashx. Accessed September 20, 2011. 3. Mulhall JP, Rojaz-Cruz C and Muller A: An analysis of sexual health information on radical prostatectomy websites. BJU Int 2010; 105: 68.
4. Hu JC, Gu X, Lipsitz SR et al: Comparative effectiveness of minimally invasive vs open radical prostatectomy. JAMA 2009; 302: 1557. 5. Schroeck FR, Krupski TL, Sun L et al: Satisfaction and regret after open retropubic or robot-assisted laparoscopic radical prostatectomy. Eur Urol 2008; 54: 785. 6. Kolata G: Results unproven, robotic surgery wins converts. New York Times, February 13, 2010. Available at www.nytimes.com/2010/02/14/health/14 robot.html?_r⫽1&scp⫽4&sq⫽Jim%20Hu%20robot& st⫽cse#. Accessed September 20, 2011.
7. HCUP Databases: Overview of the Nationwide Inpatient Sample (NIS). Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality, Rockville, MD, June 2010. Available at www.hcup-us.ahrq.gov/nisoverview. jsp. Accessed September 20, 2011. 8. Rosenbaum PR and Rubin DB: Reducing bias in observational studies using subclassifications on the propensity score. J Am Stat Assoc 1984; 79: 516. 9. Robins JM, Hernan MA and Brumback B: Marginal structural models and causal inference in epidemiology. Epidemiology 2000; 11: 550.
1398
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
10. Menon M, Tewari A, Baize B et al: Prospective comparison of radical retropubic prostatectomy and robot-assisted anatomic prostatectomy: the Vattikuti Urology Institute experience. Urology 2002; 60: 864. 11. Smith JA Jr and Herrell SD: Robotic-assisted laparoscopic prostatectomy: do minimally invasive approaches offer significant advantages? J Clin Oncol 2005; 23: 8170. 12. Rudich SM, Marcovich R, Magee JC et al: Handassisted laparoscopic donor nephrectomy: comparable donor/recipient outcomes, costs, and decreased convalescence as compared to open donor nephrectomy. Transplant Proc 2001; 33: 1106. 13. Sanchez BR, Mohr CJ, Morton JM et al: Comparison of totally robotic laparoscopic Roux-en-Y gastric bypass and traditional laparoscopic Roux-en-Y gastric bypass. Surg Obes Relat Dis 2005; 1: 549.
16. Strasberg SM, Hertl M and Soper NJ: An analysis of the problem of biliary injury during laparoscopic cholecystectomy. J Am Coll Surg 1995; 180: 101. 17. Ficarra V, Novara G, Artibani W et al: Retropubic, laparoscopic, and robot-assisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol 2009; 55: 1037. 18. Goodman DC, Brownlee S, Chang CH et al: Regional and Racial Variation in Primary Care and the Quality of Care among Medicare Beneficiaries. September 9, 2010. Available at www. dartmouthatlas.org/downloads/reports/Primary_ care_report_090910.pdf. Accessed September 20, 2011.
23. Budaus L, Abdollah F, Sun M et al: The impact of surgical experience on total hospital charges for minimally invasive prostatectomy: a populationbased study. BJU Int 2011; 108: 888. 24. Garber AM: Modernizing device regulation. N Engl J Med 2010; 362: 1161. 25. Gieringer DH: The safety and efficacy of new drug approval. Cato J 1985; 5: 177. 26. Maxwell S, Zuckerman S and Berenson RA: Use of physicians’ services under Medicare’s resource-based payments. N Engl J Med 2007; 356: 1853.
19. Cho JE and Nezhat FR: Robotics and gynecologic oncology: review of the literature. J Minim Invasive Gynecol 2009; 16: 669.
27. Pearson SD and Bach PB: How Medicare could use comparative effectiveness research in deciding on new coverage and reimbursement. Health Aff (Millwood) 2010; 29: 1796.
20. Gaia G, Holloway RW, Santoro L et al: Roboticassisted hysterectomy for endometrial cancer compared with traditional laparoscopic and laparotomy approaches: a systematic review. Obstet Gynecol 2010; 116: 1422.
28. Lawthers AG, McCarthy EP, Davis RB et al: Identification of in-hospital complications from claims data. Is it valid? Med Care 2000; 38: 785.
14. Lim PC, Kang E and Park do H: Learning curve and surgical outcome for robotic-assisted hysterectomy with lymphadenectomy: case-matched controlled comparison with laparoscopy and laparotomy for treatment of endometrial cancer. J Minim Invasive Gynecol 2010; 17: 739.
21. Bolenz C, Gupta A, Hotze T et al: Cost comparison of robotic, laparoscopic, and open radical prostatectomy for prostate cancer. Eur Urol 2010; 57: 453.
15. Pappas TN and Jacobs DO: Laparoscopic resection for colon cancer–the end of the beginning? N Engl J Med 2004; 350: 2091.
22. Mir SA, Cadeddu JA, Sleeper JP et al: Cost comparison of robotic, laparoscopic, and open partial nephrectomy. J Endourol 2011; 25: 447.
29. Frank RG: The ongoing regulation of generic drugs. N Engl J Med 2007; 357: 1993. 30. Hohwu L, Akre O, Pedersen KV et al: Open retropubic prostatectomy versus robot-assisted laparoscopic prostatectomy: a comparison of length of sick leave. Scand J Urol Nephrol 2009; 43: 259.
EDITORIAL COMMENTS This thoughtful and well executed analysis of the Nationwide Inpatient Sample forces us to consider the costs and benefits of our early and vigorous adoption of laparoscopic and robotic technologies. As the results indicate, there are pros and cons to these new approaches that need to be weighed carefully when assessing the value of these novel interventions. While this study provides some important information that will help patients, providers and policy makers make these value judgments, it fails to capture 3 important elements. First, as the authors note, the costs estimates do not include the capital investment of purchasing a robotic system or the indirect economic benefits of patients’ early return to work and increased productivity. Second, the analysis fails to capture the human cost of the learning curve. In other words, as providers learn new surgical techniques, outcomes are often worse for
patients early in the learning curve as hypothesized by Hu (reference 4 in article) and others.1 Finally, and perhaps most importantly, the study fails to capture patient reported outcomes such as postoperative pain, return to baseline functional status and health related quality of life, all of which are highly germane to the procedures under study here. It is difficult, if not impossible, to assess the costeffectiveness of minimally invasive technologies without including these critical outcomes. Future comparative effectiveness studies of these techniques must include patient reported outcomes as the primary end point if they are to inform the debate regarding the value of our interventions. David F. Penson Vanderbilt University Medical Center Nashville, Tennessee
REFERENCE 1. Vickers AJ, Savage CJ, Hruza M et al: The surgical learning curve for laparoscopic radical prostatectomy: a retrospective cohort study. Lancet Oncol 2009; 10: 475.
The authors do a commendable job of presenting a balanced overview of the use and costs associated
with a broad array of open and minimally invasive procedures. Their results demonstrate what has
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
been evident for many years: robotic surgery is widely and increasingly used, even if definitive data demonstrating its superiority over open surgery are lacking. While the reasons for this rapid adoption are varied, 1 possibility is elucidated by their results. RALS is associated with shorter LOS and fewer blood transfusions. However, as the authors admit, this reduction is not easily explained since the NIS data set does not include disease specific characteristics (eg tumor stage). In addition, longer term complications after surgery (eg erectile dysfunction and urinary incontinence after radical
1399
prostatectomy) cannot be assessed. Adoption of a new technology, while possibly reducing the risk of certain complications, continues to be associated with increased cost.1 However, as some authors have suggested, it is possible that at the societal level the reduced time to return to work after RALS may help offset this cost (reference 30 in article). G. Joel DeCastro Columbia University Medical Center New York, New York
REFERENCE 1. Nguyen PL, Gu X, Lipsitz SR et al: Cost implications of the rapid adoption of newer technologies for treating prostate cancer. J Clin Oncol 2011; 29: 1517.
Abbreviations and Acronyms DS ⫽ data suppressed per Nationwide Inpatient Sample for 0⬍n⬍11 LOS ⫽ length of hospital stay LS ⫽ laparoscopic surgery OS ⫽ open surgery NIS ⫽ Nationwide Inpatient Sample RALS ⫽ robotic assisted laparoscopic surgery Submitted for publication July 23, 2011. Nothing to disclose. Supported by the American Urological Association Foundation (HY) and the U.S. Department of Defense Physician Training Award W81XWH08-1-0283 (JCH). Supplementary material can be obtained at www.jurology.com. * Correspondence: Brigham and Women’s/ Faulkner Hospital, 1153 Centre St., Suite 4420, Boston, Massachusetts 02130 (telephone: 617983-4570; FAX: 617-983-7945; e-mail: jhu2@ partners.org).
Editor’s Note: This article is the fifth of 5 published in this issue for which category 1 CME credits can be earned. Instructions for obtaining credits are given with the questions on pages 1514 and 1515.
Purpose: Although robotic assisted laparoscopic surgery has been aggressively marketed and rapidly adopted, there are few comparative effectiveness studies that support its purported advantages compared to open and laparoscopic surgery. We used a population based approach to assess use, costs and outcomes of robotic assisted laparoscopic surgery vs laparoscopic surgery and open surgery for common robotic assisted urological procedures. Materials and Methods: From the Nationwide Inpatient Sample we identified the most common urological robotic assisted laparoscopic surgery procedures during the last quarter of 2008 as radical prostatectomy, nephrectomy, partial nephrectomy and pyeloplasty. Robotic assisted laparoscopic surgery, laparoscopic surgery and open surgery use, costs and inpatient outcomes were compared using propensity score methods. Results: Robotic assisted laparoscopic surgery was performed for 52.7% of radical prostatectomies, 27.3% of pyeloplasties, 11.5% of partial nephrectomies and 2.3% of nephrectomies. For radical prostatectomy robotic assisted laparoscopic surgery was more prevalent than open surgery among white patients in high volume, urban hospitals (all p ⱕ0.015). Geographic variations were found in the use of robotic assisted laparoscopic surgery vs open surgery. Robotic assisted laparoscopic surgery and laparoscopic surgery vs open surgery were associated with shorter length of stay for all procedures, with robotic assisted laparoscopic surgery being the shortest for radical prostatectomy and partial nephrectomy (all p ⬍0.001). For most procedures robotic assisted laparoscopic surgery and laparoscopic surgery vs open surgery resulted in fewer deaths, complications, transfusions and more routine discharges. However, robotic assisted laparoscopic surgery was more costly than laparoscopic surgery and open surgery for most procedures. Conclusions: While robotic assisted and laparoscopic surgery are associated with fewer deaths, complications, transfusions and shorter length of hospital stay compared to open surgery, robotic assisted laparoscopic surgery is more costly than laparoscopic and open surgery. Additional studies are needed to better delineate the comparative and cost-effectiveness of robotic assisted laparoscopic surgery relative to laparoscopic surgery and open surgery. Key Words: robotics, urologic surgical procedures, laparoscopy, treatment outcome, costs and cost analysis
1392
www.jurology.com
0022-5347/12/1874-1392/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
AND
RESEARCH, INC.
Vol. 187, 1392-1399, April 2012 Printed in U.S.A. DOI:10.1016/j.juro.2011.11.089
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
WITH more than 1,400 robotic surgical systems installed in United States hospitals, with some having up to 5 systems, and the number of robotic systems in other countries doubling from 200 to 400 between 2007 and 2009,1 robotic assisted laparoscopic surgery has been rapidly adopted without population based evidence demonstrating superior outcomes compared to laparoscopic surgery and open surgery. As a result the Institute of Medicine has prioritized RALS for comparative effectiveness research.2 Direct to consumer advertising has fueled patient demand for RALS,3 particularly for radical prostatectomy. However, men who underwent radical prostatectomy with RALS vs OS were more likely to be diagnosed with incontinence and erectile dysfunction, and more likely to experience treatment regret.4,5 Most existing studies that demonstrate better outcomes with RALS are single surgeon series, whereby investigators may receive educational or research funding from the device manufacturer (Intuitive Surgical®, Sunnyvale, California). Moreover estimates of RALS use are provided primarily by the manufacturer.1,6 In this study we characterize population based RALS use and patterns of care for urological procedures, and compare perioperative costs and outcomes with LS and OS.
MATERIALS AND METHODS Data Source Subjects were identified from the HCUP (Healthcare Cost and Utilization Project) NIS (Nationwide Inpatient Sample), sponsored by the Agency for Healthcare Research and Quality.7 NIS is a 20% stratified probability sample that encompasses approximately 8 million acute hospital stays from more than 1,000 hospitals in 42 states per year. It is the largest all-payer inpatient care observational cohort in the United States and represents approximately 90% of all hospitalizations.
Study Cohort During the last quarter of 2008 there were 2,093,300 hospitalizations within the NIS. Using NIS discharge weights these represent more than 9.8 million patients. We used the ICD-9 code 17.4x for RALS, initiated on October 1, 2008, to identify RALS procedures approved by the U.S. Food and Drug Administration. RALS were comprised of 64% urological, 32% gynecologic, 2% cardiac and 2% general surgical procedures. Procedures with a laparoscopic designation (ICD-9 54.21, 54.51) were classified as LS, while those without LS or RALS designations/codes were classified as OS. To adequately power analyses we analyzed urological procedures with 40 or more unweighted procedures including radical prostatectomy (ICD-9 60.5), nephrectomy (ICD-9 55.51, 55.52, 55.54), partial nephrectomy (ICD-9 55.4) and pyeloplasty (ICD-9 55.87).
Covariates For each procedure we examined hospital and patient level characteristics that may be associated with out-
1393
comes. Hospital characteristics included U.S. census region, urban vs rural location, teaching status and bed size. Hospital surgical volume was assessed by stratifying each procedure into high, intermediate and low volume tertiles to minimize cell counts of less than 11 (for which DS is required per NIS). Patient level characteristics include age, number of comorbidities, race, median income and primary payer (private vs government health plans).
Outcomes ICD-9 diagnosis and procedure codes were used to identify blood transfusions and complications (cardiac, respiratory, genitourinary, vascular, wound, miscellaneous medical and miscellaneous surgical).4 NIS specific outcomes included death, hospital LOS, discharge disposition (routine [home] vs other [rehabilitation, skilled nursing facility, etc]) and total costs. Costs were derived from total charges using the HCUP cost-to-charge ratio.
Statistical Analysis Use of RALS, LS and OS during the study period was characterized for each procedure. Stratification, clustering and survey weights were used in accordance with NIS sampling. Because characteristics of subjects undergoing RALS and LS differed from those undergoing OS, propensity scoring methods were used to adjust for potential bias associated with selection for open vs minimally invasive procedures.8 This approach controls for factors that may confound group assignment and outcomes by adjusting discharge weights, with the goal of balancing characteristics among groups. Adjustments were conducted using multivariate logistic regression models to calculate the propensity of undergoing RALS, LS or OS based on all covariates described, and weighted by the inverse propensity of being in one of the treatment groups.9 Propensity scoring models were constructed for each of the 4 procedures examined. Balance of covariates across procedures was verified following propensity adjustment. Linear and logistic regressions were used to examine univariable and multivariable effects of surgical approach on outcomes. LOS comparisons were modeled using log-normal linear regression. All analyses were performed with SAS® version 9.2 and all tests were considered statistically significant at p ⱕ0.05.
RESULTS Procedure Frequencies The relative use of RALS, LS and OS is shown in the figure. Radical prostatectomy was the only procedure in which RALS (52.7%) was more prevalent than OS (44.4%) and LS (2.8%) combined. OS was the predominant surgical approach for all other procedures. LS was least prevalent among all procedures except laparoscopic nephrectomy, in which RALS was least prevalent. Characteristics of Study Sample Patient and hospital characteristics are shown in table 1. Propensity adjustment balanced covariates for all procedures except for pyeloplasty age and hospital
1394
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
100% 90% 79.4%
79.0%
80% 70% 60%
66.1% 52.7%
50%
44.4%
40% 27.3%
30% 20% 10%
2.8%
0%
)
2 58
1,
2 n=
y( m to
l ca
di
Ra
c te ta os Pr
2.3%
8)
)
4 83
5 ,8
1, n= y(
=3
n y( om ct
st
la
op
el Py
18.3%
11.5% 9.5%
6.6%
Roboc Laparoscopic Open
re
ph
r Pa
a
e lN
y m to ec hr p Ne
) 16 ,2 14 = (n
spiratory complications (all p ⱕ0.032). For partial nephrectomy RALS was associated with the fewest genitourinary, wound and vascular complications (all p ⱕ0.015). For pyeloplasty LS was associated with the fewest miscellaneous surgical complications (p ⫽ 0.049). Health care costs were higher for RALS vs LS and OS (all p ⱕ0.022) for all procedures except for partial nephrectomy, where costs were similar (table 3). OS was least costly for radical prostatectomy, whereas costs were lowest with LS for nephrectomy and pyeloplasty (all p ⱕ0.022).
DISCUSSION Use of robotic assisted, laparoscopic and open urological procedures.
type due to low procedure numbers. White patients were more likely than nonwhites to undergo RALS or LS vs OS for radical prostatectomy (p ⫽ 0.002). Urban vs rural hospitals were more likely to perform RALS for radical prostatectomy, partial nephrectomy and pyeloplasty (all p ⱕ0.015). High and intermediate vs low volume hospitals were more likely to perform RALS for radical prostatectomy (p ⬍0.001). However, high volume hospitals were more likely to perform OS vs RALS for pyeloplasty (p ⫽ 0.033). Use of RALS for nephrectomy and partial nephrectomy was more prevalent in the Midwest, whereas hospitals in the South were more likely to perform OS for both procedures (both p ⱕ0.014). Outcomes Results from unadjusted and propensity adjusted analyses were largely similar and, therefore, outcomes from adjusted analyses are presented (table 2). RALS and LS vs OS for radical prostatectomy and nephrectomy were associated with lower in-hospital mortality (both p ⬍0.001). Similarly RALS and LS vs OS were associated with fewer transfusions for radical prostatectomy (p ⬍0.001) while LS vs RALS and OS was associated with fewer transfusions for nephrectomy and pyeloplasty (both p ⬍0.001). For all procedures RALS and LS vs OS were associated with shorter LOS (all p ⱕ0.002). Additionally, length of stay was shortest with RALS for radical prostatectomy and partial nephrectomy, whereas LS was associated with the shortest LOS for nephrectomy (all p ⬍0.001). RALS and LS vs OS for partial nephrectomy was associated with more routine home discharges (p ⫽ 0.006). RALS and LS vs OS for radical prostatectomy were associated with fewer respiratory, wound and vascular complications (all p ⱕ0.024). For nephrectomy RALS was associated with the fewest cardiac complications and LS was associated the fewest re-
Minimally invasive surgery often offers distinct, consistently reproducible advantages compared to open approaches including smaller incisions, reduced intraoperative blood loss, decreased postoperative pain, and shorter hospital LOS and convalescence.10 –12 Moreover advantages of RALS vs LS include a 3-dimensional view of the operative field, the absence of a fulcrum effect, 7 vs 4 df of movement with wristed instruments that facilitate intracorporeal suturing, elimination of surgeon tremor and ergonomic benefits. Thus, RALS purportedly facilitates the learning curve for open surgeons transitioning to minimally invasive surgery.11,13,14 Many patients intuitively perceive minimally invasive approaches as reducing complications compared with conventional OS, and prefer them due to smaller incisions requiring fewer analgesics and shorter hospital stays, even at a greater cost.15 However, rapid adoption, prolonged learning curves, and varying hospital accreditation practices for attaining LS and RALS privileges may result in unforeseen risks. For example, the rapid adoption of laparoscopic cholecystectomy in the 1990s resulted in a spike in biliary tract injuries from 1,500 to 4,000 per year.16 Studies comparing the 3 surgical approaches are sparse and mostly single surgeon series that are often conflicting even for radical prostatectomy, the most commonly performed RALS procedure.17 Population based comparisons allow characterization of use and outcomes of these surgical modalities in the absence of randomized controlled trials. Our study is the first population based comparative effectiveness study of RALS use, patterns of care and costs, and resulted in several important findings. While the majority of radical prostatectomies were performed via RALS, its use is not as high as estimates from the device manufacturer,6 but consistent with reported Medicare radical prostatectomy use of RALS and LS combined.4 We found racial, geographic and hospital based variations in the use of RALS, LS and OS for some but not all procedures. This is consistent with previous reports of limited access to
Table 1. NIS weighted unadjusted patient and hospital characteristics Radical Prostatectomy p Value
RALS
LS
OS
996 (8.7) 4,051 (35.2) 5,516 (47.9) 950 (8.3)
63 (10.2) 277 (44.8) 249 (40.3) 29 (4.7)
764 (7.9) 3,597 (37.1) 4,449 (45.9) 894 (9.2)
101 (30.8) 45 (13.7) 74 (22.4) 0.262 109 (33.1)
907 (34.9) 508 (19.6) 527 (20.3) 654 (25.2)
3,525 (31.2) 2,468 (21.9) 2,744 (24.3) 2,553 (22.6)
7,948 (69.0) 398 (64.5) 1,727 (15.0) 1,723 (28.0) 1,838 (16.0) 46 (7.5)
5,794 (59.7) 1,677 (17.3) 2,233 (23.0)
213 (64.8) 45 (13.7) 0.002 71 (21.6)
1,473 (56.8) 502 (19.3) 621 (23.9)
6,801 (60.2) 2,306 (20.4) 2,184 (19.3)
7,647 (66.4) 3,242 (28.2) 624 (5.4)
440 (71.3) 144 (23.4) 33 (5.3)
6,137 (63.3) 2,877 (29.6) 690 (7.1)
135 (41.0) 134 (40.7) 0.202 60 (18.3)
1,190 (45.9) 1,004 (38.7) 402 (15.5)
5,215 (46.2) 4,327 (38.3) 1,749 (15.5)
1,575 (13.9) 2,743 (24.3) 2,973 (26.3) 4,001 (35.4)
104 (17.5) 108 (18.1) 138 (23.1) 247 (41.3)
1,791 (19.2) 2,237 (23.9) 2,374 (25.4) 2,943 (31.5)
83 (25.7) 90 (27.9) 86 (26.6) 65 (19.9)
553 (21.8) 643 (25.4) 629 (24.8) 707 (27.9)
2,520 (22.7) 2,858 (25.7) 3,015 (27.1) 2,720 (24.5)
229 (2.0) 3,582 (31.1) 7,702 (66.9)
DS 685 (7.1) 100 (16)* 2,691 (27.7) 510 (83)* 6,328 (65.2)
25 (7.5) 77 (23.4) 0.015 227 (69.1)
111 (4.3) 646 (24.9) 1,839 (70.8)
504 (4.5) 3,296 (29.2) 7,491 (66.4)
851 (7.4) 1,637 (14.2) 9,025 (78.4)
31 (5.0) 234 (37.9) 353 (57.2)
727 (7.5) 2,236 (23.0) 6,741 (69.5)
15 (4.7) 74 (22.4) 0.106 240 (72.9)
187 (7.2) 426 (16.4) 1,984 (76.4)
910 (8.1) 2,028 (18.0) 8,353 (74.0)
2,617 (22.7) 4,587 (39.8) 4,309 (37.4)
51 (8.2) 314 (51.0) 252 (40.8)
4,691 (48.3) 95 (28.8) 2,556 (26.3) 173 (52.7) 2,457 (25.3) ⬍0.001 61 (18.5)
785 (30.2) 808 (31.1) 1,004 (38.7)
3,890 (34.5) 3,791 (33.6) 3,610 (32.0)
2,352 (20.4) 3,266 (28.4) 3,834 (33.3) 2,061 (17.9)
105 (17.0) 107 (17.4) 297 (48.2) 108 (17.4)
1,684 (17.4) 2,057 (21.2) 3,625 (37.4) 2,337 (24.1)
51 (15.5) 152 (46.3) 67 (20.3) 0.596 59 (18.0)
456 (17.6) 848 (32.7) 628 (24.2) 663 (25.6)
2,013 (17.8) 2,489 (22.0) 4,221 (37.4) 2,568 (22.7)
4,412 (38.3) 4,448 (38.6) 2,652 (23.0)
220 (35.7) 269 (43.7) 127 (20.7)
3,597 (37.1) 3,515 (36.2) 2,592 (26.7)
76 (23.1) 108 (32.8) 0.080 145 (44.1)
865 (33.3) 657 (25.3) 1,074 (41.4)
3,182 (28.2) 3,016 (26.7) 5,093 (45.1)
0.088
RALS
LS
Partial Nephrectomy OS
p Value
RALS
LS
OS
0.222
155 (34.8) 107 (24.1) 113 (25.5) 70 (15.7)
106 (29.0) 113 (31.0) 59 (16.3) 86 (23.7)
931 (30.5) 797 (26.1) 794 (26.0) 527 (17.3)
0.621
308 (69.3) 36 (8.1) 100 (22.6)
232 (63.8) 82 (22.4) 51 (13.9)
0.980
268 (60.3) 113 (25.5) 63 (14.3)
Pyeloplasty p Value
RALS
LS
OS
p Value
310 (72)* 80 (19)* 30 (7)* 0.419 DS
80 (76)* 20 (19)* 0 DS
845 (80.8) 69 (6.6) 88 (8.4) 44 (4.2) ⬍0.001
2,011 (66.0) 611 (20.0) 427 (14.0)
243 (56.3) 62 (14.5) 0.200 127 (29.3)
63 (60.9) 20 (19.3) 21 (19.8)
602 (57.5) 240 (22.9) 205 (19.6)
0.601
220 (60.4) 125 (34.3) 20 (5.4)
1,753 (57.5) 941 (30.9) 354 (11.6)
315 (72.9) 35 (8.1) 0.376 82 (19.0)
80 (76)* DS 20 (19)*
615 (58.7) 85 (8.2) 346 (33.1)
0.089
0.737
70 (15.7) 121 (27.3) 122 (27.5) 131 (29.6)
44 (12.2) 61 (17.0) 99 (27.4) 156 (43.5)
577 (19.3) 662 (22.2) 758 (25.4) 989 (33.1)
85 (19.8) 84 (19.7) 161 (37.7) 0.297 98 (22.9)
20 (19.6) 25 (24.2) 34 (32.9) 24 (23.3)
217 (21.3) 298 (29.2) 216 (21.2) 289 (28.4)
0.335
0.649
0 176 (39.6) 269 (60.5)
0 60 (16.6) 304 (83.5)
131 (4.3) 0 635 (20.8) 113 (26.1) 2,283 (74.9) ⬍0.001 319 (74.0)
0 30 (28.4) 75 (71.6)
69 (6.6) 197 (18.8) 781 (74.6) ⬍0.001
0.887
DS 70 (16)* 370 (83)*
20 (5.5) 93 (25.6) 251 (69.0)
250 (8.2) 445 (14.6) 2,354 (77.2)
19 (4.3) 110 (25.5) 0.138 303 (70.2)
DS 20 (19)* 80 (75)*
62 (6.0) 210 (20.1) 774 (73.9)
0.830
0.252
150 (33.9) 114 (25.7) 180 (40.4)
96 (26.3) 204 (56.0) 65 (17.7)
1,056 (34.6) 966 (31.7) 1,027 (33.7)
125 (29.0) 205 (47.5) 0.095 102 (23.5)
40 (38.4) 49 (47.0) 15 (14.6)
365 (34.9) 277 (26.5) 404 (38.6)
0.033
0.014
64 (14.5) 252 (56.8) 62 (14.0) 65 (14.7)
105 (28.8) 106 (29.1) 59 (16.1) 95 (26.0)
755 (24.8) 104 (24.0) 519 (17.0) 153 (35.5) 1,103 (36.2) 82 (19.1) 672 (22.0) ⬍0.001 93 (21.5)
15 (14.2) 26 (24.6) 25 (23.8) 39 (37.5)
170 (16.2) 228 (21.8) 321 (30.7) 327 (31.3)
0.379
0.288
132 (29.8) 150 (33.7) 162 (36.5)
97 (26.7) 103 (28.2) 165 (45.1)
856 (28.1) 898 (29.5) 1,294 (42.5)
60 (57)* 30 (29)* DS
718 (68.7) 192 (18.3) 136 (13.0)
0.464
1395
Weighted counts using NIS complex survey weights, numbers may not sum to group totals or percents may not add to 100 due to need for rounding. * Number rounded to nearest 10 to prevent calculation of suppressed data, percent based on rounded number.
293 (67.9) 69 (15.9) 0.813 70 (16.3)
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
No. age (%): 50 or Younger 51–60 61–70 71 or Older No. race (%): White Nonwhite Missing No. primary payer (%): Private Medicare Medicaid/other No. ZIP Code income (%): 1st (lowest quartile) 2nd 3rd 4th No. hospital type (%): Rural Urban nonteaching Urban teaching No. hospital bed size (%): Small Medium Large No. hospital vol (%): Low Intermediate High No. hospital region (%): Northeast Midwest South West No. comorbidity (%): None 1 Multiple
Nephrectomy
0.015 0 0 DS 0.717 DS DS 16 (3.9) 0.813 DS 0 DS Not 0 0 0 applicable 0.260 DS DS 38 (9.1) 0.922 DS 0 DS ⬍0.001 2.2 (1.8) 2.2 (1.2) 3.0 (3.0) 0.006 166 (97.4) 29 (100.0) 407 (97.5) DS 262 (14.8) 89 (5.1) 0 DS 35 (16.8) 11 (5.1) 0 DS 40 (17.0) 22 (9.5) 0 (0.8) ⬍0.001 (4.6) 0.321 (2.6) 0.382 (0.2) ⬍0.001 64 399 225 19 (8.7) (2.5)
(3.6)
0 41 12 0 (0.4) (5.1) (2.0) 40 525 203 0 Vascular Miscellaneous medical Miscellaneous surgical Death
17 0
DS DS (0.7) (1.1) (1.2) (0.2) 73 110 119 23 No. complications (%): Cardiac Respiratory Genitourinary Wound
Any complication 864 (8.4) 68 (14.5) 869 (10.1) 0.096 71 (29.9) 755 (33.8) 3,616 (37.4) 0.218 46 (18.3) 55 (26.7) 512 (28.9) No. blood transfusion (%) 168 (1.6) DS 452 (5.2) ⬍0.001 56 (23.5) 165 (7.4) 1,366 (14.1) ⬍0.001 20 (8.1) 13 (6.1) 137 (7.8) Mean days LOS (SD) 1.7 (3.3) 2.0 (3.2) 2.4 (4.0) ⬍0.001 5.2 (13.2) 4.2 (7.9) 5.9 (15.6) ⬍0.001 2.8 (2.3) 3.6 (3.3) 4.5 (6.8) No. routine discharge (%) 9,770 (94.9) 443 (94.2) 8,101 (94.1) 0.658 219 (92.0) 1901 (85.2) 7,990 (82.6) 0.076 248 (97.9) 205 (99.2) 1,608 (90.9)
0 0
DS 28 (11.2) 18 (7.3) 0
0 DS 0
11 15 DS DS
(6.1) 472 (4.9) ⬍0.001 (8.0) 1,196 (12.4) 0.032 (6.1) 791 (8.2) 0.219 (1.4) 277 (2.9) 0.147 135 178 135 32 0 29 (12.2) 15 (6.3) DS (1.1) 0.154 (2.2) 0.024 (1.1) 0.203 (0.6) ⬍0.001 96 192 90 55
RALS p Value OS LS RALS
Radical Prostatectomy
Table 2. Propensity adjusted outcomes
17 (0.8) 218 (2.3) 0.138 442 (19.8) 1,873 (19.4) 0.880 102 (4.6) 725 (7.5) 0.060 13 (0.6) 139 (1.4) ⬍0.001
0
DS
16 0
DS DS DS DS DS 0.608 0.115 ⬍0.001 ⬍0.001 68 141 82 34 (5.3) (7.3)
OS LS LS
Nephrectomy
OS
p Value
RALS
Partial Nephrectomy
(3.9) (8.0) (4.6) (2.0)
p Value
RALS
0 0
LS
Pyeloplasty
OS
(3.7)
0.573 0.193 0.242 Not applicable 0.496 0.913 0.049 Not applicable 0.505 ⬍0.001 0.002 0.371
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
p Value
1396
care for nonwhite patients.18 Higher volume hospitals were more likely to offer RALS and LS for radical prostatectomy. However, they were more likely to offer OS for pyeloplasty, which may be necessary for complex cases with greater disease severity, such as those with prior surgical intervention that preclude the use of RALS and LS. Length of stay was also shorter for RALS and LS than OS for all procedures, with RALS having the shortest LOS for radical prostatectomy and partial nephrectomy. Moreover discharge to home vs rehabilitation or nursing facilities was more likely for RALS and LS vs OS for partial nephrectomy, and both were associated with fewer complications for radical prostatectomy, nephrectomy and partial nephrectomy than OS. Prior studies have demonstrated comparable or better outcomes with RALS vs LS or OS.19,20 Moreover RALS and LS were associated with fewer transfusions and, correspondingly, fewer in-hospital deaths. These advantages may be secondary to the 10⫻ to 12⫻ magnification inherent to RALS and LS, and the carbon dioxide insufflation that physiologically reduces venous ooze.11 The higher transfusion rate of nephrectomy with RALS may be related to the relative infrequency of use compared with RALS for the other 3 procedures and to the overall nephrectomy volume, which may reflect a learning curve effect. Despite the advantages mentioned, RALS was more costly than LS or OS. Similarly, others have demonstrated RALS to have $1,065 to $2,315 more in direct costs for radical prostatectomy and $535 to $1,651 more for partial nephrectomy, despite the shorter LOS.21,22 A recent study estimated that the incremental cost (without factoring in the purchase of the robotic surgical system) of RALS averages approximately $1,600, or an additional 6% per case more than OS.1 The cost differential between RALS and OS in our study for radical prostatectomy ($1,111) was more modest than this estimate. This may be due to lower costs associated with higher RALS volume for radical prostatectomy compared with other procedures, consistent with a study demonstrating 68% lower charges for high vs low volume minimally invasive radical prostatectomy surgeons.23 The incremental cost doubles when factoring in the $1.75 million capital acquisition cost of the robotic surgical system. This projects to $2.5 billion in annual U.S. health care expenditures if RALS replaces conventional surgical approaches.1 However, RALS use was unavailable due to the absence of billing designations for RALS at the time of that study, and costs were derived primarily from single surgeon series. Moreover because medical devices are not subject to the same level of scrutiny compared to pharmaceuticals in demonstrating clinical effectiveness,24,25 the rapid adoption of new medical technol-
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
1397
Table 3. Propensity adjusted costs Median Costs (IQR)
Radical prostatectomy Nephrectomy Partial nephrectomy Pyeloplasty
RALS
LS
OS
p Value
$10,804 ($8,289–$13,640) $13,894 ($11,291–$21,956) $15,724 ($10,398–$18,812) $11,829 ($8,577–$17,072)
$10,082 ($8,084–$12,631) $11,153 ($8,927–$14,884) $12,401 ($9,845–$16,042) $ 8,291 ($7,749–$11,830)
$ 9,693 ($7,600–$12,113) $12,548 ($9,377–$18,788) $11,817 ($9,334–$15,384) $ 9,520 ($7,666–$12,864)
0.007 ⬍0.001 0.442 0.022
ogies significantly contributes to spiraling health care costs.26 Pearson and Bach proposed a tiered payment system,27 whereby full reimbursement be rendered for therapies demonstrating superior outcomes to existing standards. Under this paradigm, in the absence of demonstrable cost-effectiveness, fiscal responsibility for the incremental cost of RALS shifts from insurers to patients. Our study must be interpreted within the context of the study design. Administrative data are designed for billing purposes and may lack detailed clinical information. We were unable to characterize tumor characteristics, disease severity or body mass index, which may affect patient selection and outcomes. For instance, surgeries for higher stage and grade disease may be technically more challenging and likely to require open surgical approaches. Therefore, tumor characteristics may be potential confounders for complications and transfusion outcomes. Additionally, the NIS contains hospital rather than physician level characteristics and may underestimate the frequency of some outcomes. We were unable to adjust for potential confounding due to surgeon attributes. The NIS is also limited to the inpatient hospital setting, and we were unable to assess outpatient complications and costs or earlier return to activities of daily living/ work, which may offset the higher costs of RALS. We evaluated generalized complications for many procedures. However, claims data have a high degree of corroboration with chart abstraction and are valid for detecting adverse outcomes.28 While we attempted to
adjust for confounding, there was missing race data. Moreover the propensity approach was unable to balance all pyeloplasty covariates and, thus, conclusions regarding race and for pyeloplasty must be interpreted with caution. Our RALS hospital costs exclude surgeon fees and robotic system acquisition and maintenance costs, and likely underestimate the true costs of RALS. While Intuitive Surgical currently enjoys a monopoly over RALS, likely contributing to higher expenses, costs should attenuate with the entry of competitors.29 Estimates are further limited by the lack of indirect cost data following discharge, including subsequent admissions or financial losses due to convalescence. It is possible that the higher direct costs of RALS may be offset by shorter convalescence periods which have been associated with radical prostatectomy.30 Finally, this is an observational study and there may be unobserved factors for which we were unable to adjust. While randomized control trials balance potential confounders, direct to consumer advertising, patient and surgeon preferences, and intra-surgeon heterogeneity in technique and outcomes regardless of surgical approach are barriers to implementing this research design. In summary, RALS has been rapidly adopted with disparities in use by race, hospital setting and geography for certain procedures. Despite greater costs, RALS has advantages such as fewer complications, fewer transfusions and shorter LOS. However, prospective studies are needed to assess the comparative effectiveness of RALS vs conventional surgical approaches.
REFERENCES 1. Barbash GI and Glied SA: New technology and health care costs–the case of robot-assisted surgery. N Engl J Med 2010; 363: 701. 2. Initial national priorities for comparative effectiveness research, list of priorities: Institute of Medicine of the National Academies 2009. Available at www.iom.edu/⬃/media/Files/Report%20Files/2009/ ComparativeEffectivenessResearchPriorities/Stand%20 Alone%20List%20of%20100%20CER%20Priorities% 20-%20for%20web.ashx. Accessed September 20, 2011. 3. Mulhall JP, Rojaz-Cruz C and Muller A: An analysis of sexual health information on radical prostatectomy websites. BJU Int 2010; 105: 68.
4. Hu JC, Gu X, Lipsitz SR et al: Comparative effectiveness of minimally invasive vs open radical prostatectomy. JAMA 2009; 302: 1557. 5. Schroeck FR, Krupski TL, Sun L et al: Satisfaction and regret after open retropubic or robot-assisted laparoscopic radical prostatectomy. Eur Urol 2008; 54: 785. 6. Kolata G: Results unproven, robotic surgery wins converts. New York Times, February 13, 2010. Available at www.nytimes.com/2010/02/14/health/14 robot.html?_r⫽1&scp⫽4&sq⫽Jim%20Hu%20robot& st⫽cse#. Accessed September 20, 2011.
7. HCUP Databases: Overview of the Nationwide Inpatient Sample (NIS). Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality, Rockville, MD, June 2010. Available at www.hcup-us.ahrq.gov/nisoverview. jsp. Accessed September 20, 2011. 8. Rosenbaum PR and Rubin DB: Reducing bias in observational studies using subclassifications on the propensity score. J Am Stat Assoc 1984; 79: 516. 9. Robins JM, Hernan MA and Brumback B: Marginal structural models and causal inference in epidemiology. Epidemiology 2000; 11: 550.
1398
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
10. Menon M, Tewari A, Baize B et al: Prospective comparison of radical retropubic prostatectomy and robot-assisted anatomic prostatectomy: the Vattikuti Urology Institute experience. Urology 2002; 60: 864. 11. Smith JA Jr and Herrell SD: Robotic-assisted laparoscopic prostatectomy: do minimally invasive approaches offer significant advantages? J Clin Oncol 2005; 23: 8170. 12. Rudich SM, Marcovich R, Magee JC et al: Handassisted laparoscopic donor nephrectomy: comparable donor/recipient outcomes, costs, and decreased convalescence as compared to open donor nephrectomy. Transplant Proc 2001; 33: 1106. 13. Sanchez BR, Mohr CJ, Morton JM et al: Comparison of totally robotic laparoscopic Roux-en-Y gastric bypass and traditional laparoscopic Roux-en-Y gastric bypass. Surg Obes Relat Dis 2005; 1: 549.
16. Strasberg SM, Hertl M and Soper NJ: An analysis of the problem of biliary injury during laparoscopic cholecystectomy. J Am Coll Surg 1995; 180: 101. 17. Ficarra V, Novara G, Artibani W et al: Retropubic, laparoscopic, and robot-assisted radical prostatectomy: a systematic review and cumulative analysis of comparative studies. Eur Urol 2009; 55: 1037. 18. Goodman DC, Brownlee S, Chang CH et al: Regional and Racial Variation in Primary Care and the Quality of Care among Medicare Beneficiaries. September 9, 2010. Available at www. dartmouthatlas.org/downloads/reports/Primary_ care_report_090910.pdf. Accessed September 20, 2011.
23. Budaus L, Abdollah F, Sun M et al: The impact of surgical experience on total hospital charges for minimally invasive prostatectomy: a populationbased study. BJU Int 2011; 108: 888. 24. Garber AM: Modernizing device regulation. N Engl J Med 2010; 362: 1161. 25. Gieringer DH: The safety and efficacy of new drug approval. Cato J 1985; 5: 177. 26. Maxwell S, Zuckerman S and Berenson RA: Use of physicians’ services under Medicare’s resource-based payments. N Engl J Med 2007; 356: 1853.
19. Cho JE and Nezhat FR: Robotics and gynecologic oncology: review of the literature. J Minim Invasive Gynecol 2009; 16: 669.
27. Pearson SD and Bach PB: How Medicare could use comparative effectiveness research in deciding on new coverage and reimbursement. Health Aff (Millwood) 2010; 29: 1796.
20. Gaia G, Holloway RW, Santoro L et al: Roboticassisted hysterectomy for endometrial cancer compared with traditional laparoscopic and laparotomy approaches: a systematic review. Obstet Gynecol 2010; 116: 1422.
28. Lawthers AG, McCarthy EP, Davis RB et al: Identification of in-hospital complications from claims data. Is it valid? Med Care 2000; 38: 785.
14. Lim PC, Kang E and Park do H: Learning curve and surgical outcome for robotic-assisted hysterectomy with lymphadenectomy: case-matched controlled comparison with laparoscopy and laparotomy for treatment of endometrial cancer. J Minim Invasive Gynecol 2010; 17: 739.
21. Bolenz C, Gupta A, Hotze T et al: Cost comparison of robotic, laparoscopic, and open radical prostatectomy for prostate cancer. Eur Urol 2010; 57: 453.
15. Pappas TN and Jacobs DO: Laparoscopic resection for colon cancer–the end of the beginning? N Engl J Med 2004; 350: 2091.
22. Mir SA, Cadeddu JA, Sleeper JP et al: Cost comparison of robotic, laparoscopic, and open partial nephrectomy. J Endourol 2011; 25: 447.
29. Frank RG: The ongoing regulation of generic drugs. N Engl J Med 2007; 357: 1993. 30. Hohwu L, Akre O, Pedersen KV et al: Open retropubic prostatectomy versus robot-assisted laparoscopic prostatectomy: a comparison of length of sick leave. Scand J Urol Nephrol 2009; 43: 259.
EDITORIAL COMMENTS This thoughtful and well executed analysis of the Nationwide Inpatient Sample forces us to consider the costs and benefits of our early and vigorous adoption of laparoscopic and robotic technologies. As the results indicate, there are pros and cons to these new approaches that need to be weighed carefully when assessing the value of these novel interventions. While this study provides some important information that will help patients, providers and policy makers make these value judgments, it fails to capture 3 important elements. First, as the authors note, the costs estimates do not include the capital investment of purchasing a robotic system or the indirect economic benefits of patients’ early return to work and increased productivity. Second, the analysis fails to capture the human cost of the learning curve. In other words, as providers learn new surgical techniques, outcomes are often worse for
patients early in the learning curve as hypothesized by Hu (reference 4 in article) and others.1 Finally, and perhaps most importantly, the study fails to capture patient reported outcomes such as postoperative pain, return to baseline functional status and health related quality of life, all of which are highly germane to the procedures under study here. It is difficult, if not impossible, to assess the costeffectiveness of minimally invasive technologies without including these critical outcomes. Future comparative effectiveness studies of these techniques must include patient reported outcomes as the primary end point if they are to inform the debate regarding the value of our interventions. David F. Penson Vanderbilt University Medical Center Nashville, Tennessee
REFERENCE 1. Vickers AJ, Savage CJ, Hruza M et al: The surgical learning curve for laparoscopic radical prostatectomy: a retrospective cohort study. Lancet Oncol 2009; 10: 475.
The authors do a commendable job of presenting a balanced overview of the use and costs associated
with a broad array of open and minimally invasive procedures. Their results demonstrate what has
COMPARATIVE EFFECTIVENESS OF ROBOTIC UROLOGICAL SURGERY
been evident for many years: robotic surgery is widely and increasingly used, even if definitive data demonstrating its superiority over open surgery are lacking. While the reasons for this rapid adoption are varied, 1 possibility is elucidated by their results. RALS is associated with shorter LOS and fewer blood transfusions. However, as the authors admit, this reduction is not easily explained since the NIS data set does not include disease specific characteristics (eg tumor stage). In addition, longer term complications after surgery (eg erectile dysfunction and urinary incontinence after radical
1399
prostatectomy) cannot be assessed. Adoption of a new technology, while possibly reducing the risk of certain complications, continues to be associated with increased cost.1 However, as some authors have suggested, it is possible that at the societal level the reduced time to return to work after RALS may help offset this cost (reference 30 in article). G. Joel DeCastro Columbia University Medical Center New York, New York
REFERENCE 1. Nguyen PL, Gu X, Lipsitz SR et al: Cost implications of the rapid adoption of newer technologies for treating prostate cancer. J Clin Oncol 2011; 29: 1517.