Laparoscopy and Robotics

Complications of
Transperitoneal Laparoscopic
Nephrectomy: A Single-center Experience
Bum Soo Kim, Eun Sang Yoo, and Tae Gyun Kwon
OBJECTIVES To present the incidence of complications of transperitoneal laparoscopic nephrectomy per-
formed for various indications during a 6-year period.
METHODS From 2002 to 2007, 505 transperitoneal laparoscopic nephrectomies were performed, consisting
of 125 live donor, 212 radical, 80 simple, and 28 partial nephrectomies and 60 nephroureter-
ectomies. We retrospectively analyzed the factors related to perioperative complications, includ-
ing the type of operation, body mass index, history of abdominal surgery, and American Society
of Anesthesiologists score.
RESULTS The overall complication rate was 13.7% (69/505). Major complications requiring open con-
version or reoperation occurred in 15 patients (3.0%). The remaining 54 patients experienced
minor surgical or postoperative medical problems. The mortality rate in our series was 0%. The
complication rates by the type of operation were not significantly different. Patients with a
history of abdominal surgery demonstrated slightly greater complication rates (19.2% vs 12.6%,
P ⫽ .069). When stratified by a body mass index of ⬍25 kg/m2 and ⬍25 kg/m2, no statistically
significant difference was found in the complication rates (13.1% vs 15.6%, respectively, P ⫽
.067). Patients with greater American Society of Anesthesiologists scores had greater compli-
cation rates (P ⫽ .038). The intraoperative complication rates decreased as our experience with
laparoscopic surgery increased (P ⫽ .042); however, the total complication rates remained
constant throughout the study period.
CONCLUSIONS In consideration of the contributing factors, the complication rates of transperitoneal laparo-
scopic nephrectomy were not related to the type of operation, body mass index, or history of
abdominal operation but to the American Society for Anesthesiologists score. Complications
unique to laparoscopic nephrectomy exist but they decrease with experience. UROLOGY 73:
1283–1287, 2009. © 2009 Published by Elsevier Inc.

S
ince the first laparoscopic nephrectomy was per-
formed in 1991 by Clayman et al.,1 laparoscopy for
renal malignancies and live kidney donation has
rapidly evolved and is becoming a part of the operative
repertoire of an increasing number of urologists. Clear
advantages to laparoscopy compared with open nephrec-
tomy exist, including decreased postoperative pain, short-
ened hospital stay, rapid recovery, and improved cosme-
sis.2-5 However, laparoscopic nephrectomy is associated
with unique complications and remains challenging for
beginners. In the present study, we analyzed the inci-
dence and types of complications occurring with trans-
peritoneal laparoscopic nephrectomy performed at our
institution during a 6-year period. Various factors were
From the Department of Urology, Kyungpook National University Hospital, Daegu,
Korea
Reprint requests: Tae Gyun Kwon, M.D., Department of Urology, Kyungpook
National University Hospital, 200 Dongduk-ro, Jung-gu, Daegu 700-721 Korea.
E-mail: tgkwon@knu.ac.kr
Submitted: July 7, 2008, accepted (with revisions): January 12, 2009
assessed to identify the risk factors associated with the
perioperative complications.
MATERIAL AND METHODS
From 2002 to December 2007, a single surgeon performed 505
transperitoneal laparoscopic nephrectomies at our institution,
including 212 laparoscopic radical nephrectomies (LRNs), 125
laparoscopic donor nephrectomies (LDNs), 80 laparoscopic
simple nephrectomies (LSNs), 28 laparoscopic partial nephrec-
tomies (LPNs), and 60 laparoscopic nephroureterectomies
(LNUs) (Table 1). The average patient age was 49.6 years
(range 9-89). Of the 505 patients, 290 were male and 215 were
female. All surgeries, except for LDN, were performed using
pure laparoscopy, and every LDN was performed with a hand-
assisted procedure. In the case of pure laparoscopic nephrec-
tomy, the pneumoperitoneum was obtained using a Veress
needle, and 3 or 4 ports were used depending on the case. For
hand-assisted laparoscopic nephrectomy, a hand-assist device
was placed through a 7-cm midline abdominal incision, and a
10-mm camera port was placed periumbilically. An additional
10-mm port was placed at the flank, and another 5-mm trocar

© 2009 Published by Elsevier Inc. 0090-4295/09/$34.00 1283

doi:10.1016/j.urology.2009.01.017
Table 1. Patient characteristics and complications stratified by operation type
Characteristic LDN LSN LRN LNU LPN Total
Patients (n) 125 80 212 60 28 505
Age (y)
Mean 39.7 41.3 56.1 63.2 47.6 49.6
Range 18-64 9-72 28-89 30-81 26-68 9-89
Sex (n)
Male 64 29 137 40 20 290
Female 61 51 75 20 8 215
Side
Left 90 42 85 34 17 268
Right 35 38 127 26 11 237
Complications (n)
Intraoperative 8 4 9 6 1 28 (5.5)
Vascular injury 4 2 3 1 0 10 (2.0)
Bowel injury 1 0 0 2 0 3 (0.6)
Splenic injury 0 0 1 2 0 3 (0.6)
Diaphragmatic injury 0 0 1 1 0 2 (0.4)
Suspicious pancreatic injury 3 1 4 0 0 8 (1.6)
Liver injury 0 1 0 0 0 1 (0.2)
Renal parenchymal bleeding 0 0 0 0 1 1 (0.2)
Postoperative 10 4 17 0 3 34 (6.7)
Chylous ascites 6 0 12 0 0 18 (3.2)
Wound infection 2 0 1 0 0 3 (0.6)
Delayed bleeding 0 1 1 0 2 4 (0.8)
Pleural effusion 0 1 1 0 1 3 (0.6)
Ileus 1 2 2 0 0 5 (1.0)
Atelectasis 1 0 0 0 0 1 (0.2)
Medical 1 0 3 3 0 7 (1.4)
Infectious diarrhea 0 0 1 0 0 1 (0.2)
Azotemia 0 0 2 0 0 2 (0.4)
Hepatic dysfunction 1 0 0 0 0 1 (0.2)
Delirium 0 0 0 3 0 3 (0.6)
Total 19 (15.2) 8 (10.0) 29 (13.7) 9 (15.0) 4 (14.3) 69 (13.7)
LDN ⫽ laparoscopic donor nephrectomy; LSN ⫽ laparoscopic simple nephrectomy; LRN ⫽ laparoscopic radical nephrectomy; LNU ⫽
laparoscopic nephroureterectomy; LPN, laparoscopic partial nephrectomy.
Data in parentheses are percentages.

was placed in the subcostal area, if needed. We used 3 or 4
Hem-o-Lok clips, 10 or 15 mm, to control the renal artery. The
renal vein was controlled by 3 or 4 Hem-o-Lok clips, 15 mm, or
1 endovascular gastrointestinal anastomosis stapler. For LPN,
the renal pedicle was clamped using a laparoscopic bulldog
clamp. The specimen was removed using an Endo Catch re-
trieval bag (Autosuture, London, UK) through a 7-cm lower
abdominal incision in LRN and LSN. In LNU, a Gibson
incision was made to dissect the distal ureter in an open fashion
and remove the specimen. We retrospectively analyzed the
perioperative complications using various parameters, including
body mass index, previous history of abdominal surgery and
American Society of Anesthesiologists (ASA) score. Student’s
t test was used for statistical analysis, and statistical significance
was considered at P ⬍ .05.
RESULTS
A total of 69 (13.7%) complications occurred in 505
consecutive cases, including 28 (5.5%) intraoperative
(directly attributable to the technical execution at sur-
gery), 34 (6.7%) postoperative (directly or not directly
attributable to the technical execution and identified
after surgery), and 7 (1.4%) medical complications (Ta-
ble 1). Major complications, which required conversion
to an open procedure or reoperation, occurred in 15
patients (3.0%). The mortality rate was 0%. According
to the complication classification system suggested by
Clavien et al.,6 11 (2.2%), 43 (8.5%), and 15 (3%)
complications were grade I, II, III, respectively. Grade IV
complications did not occur in our series (Table 2).
The most common intraoperative complications were
vascular injuries inn 10 patients that could not be con-
trolled laparoscopically and required open conversion. Of
the 14 adjacent organ injuries, suspicious pancreatic in-
juries (n ⫽ 8) were the most common. Pancreatic injuries
were suspected when the postoperative drain amylase and
lipase levels were elevated. All were managed conserva-
tively. No severe parenchymal or ductal injury occurred.
One patient experienced a minor liver injury that was
managed conservatively. Splenic injuries occurred in 3
patients; 1 was treated conservatively and 2 required
open conversion to undergo splenectomy. Minor dia-
phragmatic injuries were found in 2 patients and were
repaired with laparoscopic suture. In these patients, post-
operative chest x-rays were obtained to rule out pneumo-
thorax. Three bowel injuries occurred. One was a simple
injury of the small bowel caused by electrocautery during
LNU and managed by laparoscopic suturing. The remain-
ing 2 cases were unrecognized duodenal injuries that

1284 UROLOGY 73 (6), 2009

Table 2. Complication rates stratified by variables studied
LDN LSN LRN LNU LPN
Variable (n ⫽ 125) (n ⫽ 80) (n ⫽ 212) (n ⫽ 60) (n ⫽ 28) Total (n) P Value
Grade .105
I 2 1 4 3 1 11/505 (2.2)
II 12 5 22 2 2 43/505 (8.5)
III 5 2 3 4 1 15/505 (3.0)
History of abdominal surgery .069
Patients (n) 14/98 7/70 21/179 8/55 4/25 54/427 (12.6)
Complication 5/27 1/10 8/33 1/5 0/3 15/78 (19.2)
BMI (kg/m2) .067
⬍25 14/107 5/68 23/137 6/52 3/26 51/390 (13.1)
ⱖ25 5/18 3/12 6/75 3/8 1/2 18/115 (15.6)
ASA score* .038
1 19/125 2/43 5/53 0/20 1/17 8/133 (6.0)
2 5/33 20/142 5/31 2/9 32/215 (14.9)
3 1/4 4/17 4/9 1/2 10/32 (28.1)
Total 19 8 29 9 4 69/505 (13.7)
BMI ⫽ body mass index; ASA ⫽ American Society of Anesthesiologists; other abbreviations as in Table 1.
Data in parentheses are percentages.
* Complication rates by ASA score compared, except for LDN group.

Table 3. Complication rates by year

2002 2003 2004 2005 2006 2007 P Value
Total operations (n) 26 59 87 129 106 98
Complications (n)
Intraoperative 3 (11.5) 7 (11.9) 5 (5.7) 7 (5.4) 3 (2.8) 3 (3.1) .042
Postoperative 1 (3.8) 2 (3.4) 5 (5.7) 7 (5.4) 9 (8.5) 10 (10.2)
Medical 0 (0) 0 (0) 1 (1.1) 1 (0.8) 3 (2.8) 2 (2.0)
Total 4 (15.4) 9 (15.3) 11 (12.6) 15 (11.6) 15 (14.2) 15 (15.3) .957
Data in parentheses are percentages.

required exploratory laparotomy. During LPN, 1 case of
persistent parenchymal bleeding occurred that could not
be controlled laparoscopically. Because the tumor was 4
cm and 50 minutes of warm ischemic time had already
passed, we converted to laparoscopic radical nephrec-
tomy instead of open partial nephrectomy.
Of the postoperative complications, chylous ascites (n ⫽
18) was the most common. Most patients were treated
conservatively with total parenteral nutrition for 3-7
days. However, 1 patient required surgical management
because of persistent chyle drainage despite conservative
management for 2 months. Delayed bleeding occurred in
4 patients (2 after LPN, 1 after LSN, and 1 after LRN).
Of the 2 patients who had undergone LRN, 1 required
open abdominal exploration, One LPN patient under-
went angioembolization, and the remaining 2 patients
were treated conservatively. The remaining minor post-
operative complications included 5 cases of persistent
ileus that lasted for ⬎5 days, 3 wound infections, 3
pleural effusions, and 1 atelectasis. All were managed
conservatively.
The complication rates according to the operation
type were not signficantly different statistically (P ⫽
.076): 13.7% (29/212) in LRN, 15.2% (19/125) in LDN,
10.0% (8/80) in LSN, 15.0% (9/60) in LNU, and 14.3%
(4/28) in LPN (Table 1). Comparing the incidence of
major (grade III) and minor (grade I and II) complica-
tions according to the Clavien classification, no statisti-
cally significant difference was found in each surgery
group (Table 2). For patients with previous abdominal
surgery, the complication rate (19.2%) was greater than
for patients who had not undergone previous abdominal
surgery (12.6%); however, the difference was not statis-
tically significant (P ⫽ .069; Table 2). All patients were
subdivided into 2 groups according to their BMI (nono-
bese, ⬍25 kg/m2; and overweight or obese, ⬎25 kg/m2)
for comparison of complication rates. Stratifying the data
by BMI for obesity, the comparison between overweight
or obese and nonobese patients demonstrated no signif-
icant difference in the complication rates (15.6% vs
13.1%, respectively; P ⫽ .067; Table 2). The complica-
tion rates of the laparoscopic nephrectomies, except for
LDN, had a close relationship with a high ASA score
(P ⫽ .038; Table 2). No statistically significant difference
was found in the total complication rates between the
early (2002-2003) and late (2006-2007) period of lapa-
roscopic surgery (P ⫽ .957). However, the intraoperative
complication rate decreased as the experience increased
(P ⫽ .042; Table 3).
COMMENT
Laparoscopic skills evolve with repetition, and a slow
learning curve exists for achieving these skills. Vallan-
cien et al.7 suggested in a review of laparoscopic urologic
complications that experience with a minimum of 50

UROLOGY 73 (6), 2009 1285

difficult cases is needed to acquire adequate laparoscopic
skills. In their study, the complication rate was 13.3% for
the first 100 cases and subsequently decreased to 3.6% for
the remaining cases. In our series, the intraoperative
complication rates decreased as the experience with lapa-
roscopic surgery increased; however, the total complica-
tion rate remained constant throughout the study period.
This result might have been because we performed more
complex laparoscopic nephrectomies in the latter period.
For example, chylous ascites occurred frequently in a
recent series because we performed excessive lymph node
dissection during LRN.
During laparoscopic surgery, fatal complications such
as major vascular injuries or adjacent organ injuries can
occur. In such cases, prompt and adequate management is
mandatory, and open conversion is often required. The
incidence of major vascular and adjacent organ injuries
has been reported at 1.7%-2.5% and 0.25%-0.4%, respec-
tively.8-10 The major cause of vascular injuries includes a
lack of meticulous dissection skill, malfunctions of clips
or staplers, accidental thermal injury, and unrecognized
aberrant vessels or severe adhesions around the hilar area.
In our series, 6 vascular complications occurred that were
caused by factors related to the endovascular gastrointes-
tinal anastomosis stapler. A clip in the line of the
stapler caused improper placing of staples and severe
bleeding. Therefore, minimizing the use of clips
around the renal hilum is important to reduce the risk
of stapler malfunction. With increased laparoscopic
experience, bleeding can often be controlled with var-
ious laparoscopic methods, including compression, su-
turing, clipping, and cauterization. However, cases of
hemodynamic instability or brisk bleeding that cannot
be controlled laparoscopically require conversion to an
open procedure.11 The incidence of open conversion
due to vascular injuries has been reported at 1.5%-
4%,11-13 similar to our results (2%).
Adjacent organ injuries are not uncommon complica-
tions during laparoscopic nephrectomy. The liver, spleen,
pancreas, and diaphragm are adherent to the kidney and
can easily be injured. To minimize the risk of injury, it is
important to recognize the anatomic position between
the kidney and other adjacent organs, as well as mini-
mizing spleen or liver retraction. Minor injuries of the
liver or spleen without severe bleeding can be managed
conservatively. In the case of massive bleeding, however,
it is mandatory not to hesitate to execute open conver-
sion. In our series, 1 liver, 3 splenic, 2 diaphragmatic, and
8 minor pancreatic injuries occurred. Most were managed
conservatively with a laparoscopic method. However, 2
(0.4%) of the splenic injuries required open splenectomy
because of severe bleeding. The incidence of splenic
injury requiring splenectomy was similar to that in other
series (0.7%).11
Bowel injuries usually occur during Veress needle ac-
cess, the first trocar insertion, paracolic gutter dissection,
or incidental electrocauterization during any procedure.
1286
The most fatal is the unrecognized bowel injury by inad-
vertent electrocautery out of the laparoscopic view.
Therefore, careful insertion of the Veress needle and the
first trocar is important. Moreover, careful attention is
always needed to avoid unrecognized electrocauterization
out of the laparoscopic view. Bishoff et al.14 reported 8
bowel injuries (0.8%) in 915 patients, including 4 unrec-
ognized injuries. Similarly, 3 bowel injuries (0.6%), in-
cluding 2 cases of unrecognized injury, were found in our
series.
Chylous ascites is a rare complication after major ret-
roperitoneal surgery such as nephrectomy. With the use
of laparoscopic nephrectomy increasing, not a few cases
of chylous ascites after laparoscopic nephrectomy have
been reported.15-18 It is unknown whether laparoscopic
lymphadenectomy tends to cause postoperative lym-
phatic leakage more frequently than does open surgery.
However, coagulation procedures during laparoscopic
surgery often result in incomplete lymphostasis. Lym-
phatic leakage can occur after a patient resumes oral
intake and lymphatic flow increases significantly. Postop-
erative chylous ascites are closely related to skeletoniza-
tion of the renal artery, vein, or aorta in LDN and
para-aortic lymph node dissection in LRN. Most patients
can be treated conservatively with total parenteral nutri-
tion, a medium-chain triglyceride diet, or a somatostatin
analog.16 However, prompt surgical repair is recommended
for refractory cases.16,17 Moreover, if severe symptoms, in-
cluding abdominal pain, nausea, vomiting, and severe
hypoproteinemia cannot be managed medically, early
surgical intervention is essential. In our series, most cases
of chylous ascites (17/18) were successfully managed con-
servatively. However, 1 patient required surgical repair
because of persistent chylous ascites despite conservative
management for 2 months.
The pneumoperitoneum used during laparoscopic sur-
gery can tamponade the veins and mask intraoperative
venous bleeding. This can be a cause of postoperative
delayed bleeding. Therefore, careful inspection of the
operative field and trocar site with the intra-abdominal
pressure lowered to 5 mm Hg is mandatory before exiting
the abdominal cavity. Serial hematocrit measurements
and physical examination can usually secure the diagno-
sis. Most postoperative bleeding can be managed conser-
vatively, except for severe and persistent bleeding. In our
series, 4 cases of delayed bleeding occurred and 2 were
managed conservatively. One LRN patient had bleeding
from an epigastric artery that required reoperation. The
remaining case of bleeding after LPN was managed by
selective angioembolization.
Abdominal surgery promotes the formation of adhe-
sions and can cause significant difficulties during reop-
eration. Furthermore, previous surgery at the same ana-
tomic site can be associated with high rates of morbidity.
However, it has been thought that previous laparoscopic
surgery leaves fewer adhesions compared with open sur-
gery. In our series, 73 patients had undergone previous
UROLOGY 73 (6), 2009
open abdominal surgery, including 25 appendectomies,
19 hysterectomies, 15 cholecystectomies, 6 gastrecto-
mies, 3 bowel perforation repairs, 2 pyeloplasties, 2 ure-
terolithotomies, and 1 nephrolithotomy. The complica-
tion rate for these patients was 19.2% (14/73). The
remaining 5 patients had undergone previous laparo-
scopic surgery, including 2 hysterectomies, 2 cholecystec-
tomies, and 1 renal cystectomy, and the complication
rate was 20% (1/5). In the previous urologic surgery
group (same anatomic site), only 1 postoperative com-
plication (pleural effusion) and no intraoperative com-
plications occurred. The complication rate was not af-
fected by the type of previous abdominal surgery.
It has been well documented in urologic reports that
obese patients have a greater rate of postoperative com-
plications. Appropriately, it has been considered that
obesity is closely related to technical difficulties and poor
outcomes after laparoscopic surgery; therefore, obesity is
regarded as a relative contraindication to laparoscopy.19
However, it is not clear that obesity is a risk factor for
complications during laparoscopic nephrectomy. For ex-
ample, Feder et al.20 reported fewer complications in
obese patients than in nonobese patients. In our study, no
significant difference was found between the nonobese
and overweight or obese groups.
Recent studies have shown that greater ASA scores
correlate closely with the complication rates of urologic
laparoscopic surgery.12,13 Our results also suggest that the
ASA score is a strong predictor for complications, except
for in the LDN group. This trend suggests an expected
baseline complication rate solely associated with patient
comorbidity.
CONCLUSIONS
Complications that are unique to laparoscopy exist, but
they should decrease with experience. In our analysis of
the contributing factors, the complication rates of lapa-
roscopic transperitoneal nephrectomy were not related to
the operation type, BMI, or history of abdominal surgery,
but were related to the ASA score. Despite a progressive
increase in the complexity of laparoscopic transperito-
neal nephrectomy with time, the overall rate of compli-
cations did not increase significantly and the intraoper-
ative complication rate decreased as the experience with
laparoscopic surgery increased. We believe that appropri-
ate understanding of the possible complications and man-
agement techniques is essential to reduce the risk of
complications and improve the safety of laparoscopic
transperitoneal nephrectomy.
UROLOGY 73 (6), 2009
References
1. Clayman RV, Kavoussi LR, Soper NJ, et al. Laparoscopic nephrec-
tomy: Initial case report. J Urol. 1991;146:278-282.
2. Dunn MD, Portis AJ, Shalhav AL, et al. Laparoscopic versus open
radical nephrectomy: a 9-year experience. J Urol. 2000;164:1153-
1159.
3. McDougall EM, Clayman RV, Elashry OM. Laparoscopic radical
nephrectomy for renal tumor: the Washington University experi-
ence. J Urol. 1996;155:1180-1185.
4. Ratner LE, Kavoussi LR, Sroka M, et al. Laparoscopic assisted live
donor nephrectomy—a comparison with the open approach. Trans-
plantation. 1997;63:229-233.
5. Shalhav AL, Dunn MD, Portis AJ, et al. Laparoscopic nephroure-
terectomy for upper tract transitional cell cancer: the Washington
University experience. J Urol. 2000;163:1100-1104.
6. Clavien PA, Sanabria JR, Strasberg SM. Proposed classification of
complications of surgery with examples of utility in cholecystec-
tomy. Surgery. 1992;111:518-526.
7. Vallancien G, Cathelineau X, Baumert H, et al. Complications of
transperitoneal laparoscopic surgery in urology: review of 1,311
procedures at a single center. J Urol. 2002;168:23-26.
8. Fahlenkamp D, Rassweiler J, Fornara P, et al. Complications of
laparoscopic procedures in urology: experience with 2407 proce-
dures at 4 German Centers. J Urol. 1999;162:765-770.
9. Thiel R, Adams JB, Schulam PG, et al. Venous dissection injuries
during laparoscopic urological surgery. J Urol. 1996;155:1874-1876.
10. Meraney AM, Samee AA, Gill IS. Vascular and bowel complica-
tions during retroperitoneal laparoscopic surgery. J Urol. 2002;168:
1941-1944.
11. Simon SD, Castle EP, Ferrigni RG, et al. Complications of lapa-
roscopic nephrectomy: the Mayo Clinic experience. J Urol. 2004;
171:1447-1450.
12. Permpongkosol S, Link RE, Su LM, et al. Complications of 2,775
urological laparoscopic procedures: 1993-2005. J Urol. 2007;177:
580-585.
13. Parsons JK, Varkarakis I, Rha KH, et al. Complications of abdom-
inal urologic laparoscopy: longitudinal five-year analysis. Urology.
2004;63:27-32.
14. Bishoff JT, Allaf ME, Kirkels W, et al. Laparoscopic bowel injury:
Incidence and clinical presentation. J Urol. 1999;161:887-890.
15. Shafizadeh SF, Daily PP, Baliga P, et al. Chylous ascites secondary
to laparoscopic donor nephrectomy. Urology. 2002;60:3456-3459.
16. Leibovitch I, Mor Y, Golomb J, et al. Chylous ascites after radical
nephrectomy and inferior vena cava thrombectomy: successful con-
servative management with somatostatin analogue. Eur Urol. 2002;
41:220-222.
17. Molina WR, Desai MM, Gill IS. Laparoscopic management of
chylous ascites after donor nephrectomy. J Urol. 2003;170:1938.
18. Nishizawa K, Ito N, Yamamoto S, et al. Successful laparoscopic
management of chylous ascites following laparoscopic radical ne-
phrectomy. Int J Urol. 2006;13:619-621.
19. Kapoor A, Nassir A, Chew B, et al. Comparison of laparoscopic
radical renal surgery in morbidly obese and non-obese patients. J
Endourol. 2004;18:657-660.
20. Feder MT, Patel MB, Melman A, et al. Comparison of open and
laparoscopic nephrectomy in obese and nonobese patients: out-
comes stratified by body mass index. J Urol. 2008;180:79-83.
1287