You are on page 1of 10

EUROPEAN UROLOGY 64 (2013) 588–597

available at www.sciencedirect.com
journal homepage: www.europeanurology.com

Platinum Priority – Review – Bladder Cancer


Editorial by Mathew C. Raynor and Raj S. Pruthi on pp. 598–599 of this issue

Definition, Incidence, Risk Factors, and Prevention of Paralytic


Ileus Following Radical Cystectomy: A Systematic Review

Jorge A. Ramirez a, Andrew G. McIntosh a, Robert Strehlow a, Valerie A. Lawrence b,


Dipen J. Parekh a, Robert S. Svatek a,*
a
Department of Urology, The University of Texas Health Science Center San Antonio, San Antonio, TX, USA; b Department of Medicine, The University of Texas
Health Science Center San Antonio, San Antonio, TX, USA

Article info Abstract

Article history: Context: Postoperative paralytic ileus (POI) has profound clinical consequences because
Accepted November 27, 2012 it represents a substantial burden on both patients and health care resources.
Published online ahead of Objective: To determine the knowledge base regarding POI in the radical cystectomy
(RC) population with an emphasis on preventive measures and risk factors.
print on December 5, 2012 Evidence acquisition: A systematic literature search of Medline (1966 to February 2011)
and a study review were conducted. Eligible studies explicitly reported the incidence of
Keywords: POI and/or at least two quantitative measures of gastrointestinal recovery.
Evidence synthesis: The search identified 727 relevant articles; 77 met eligibility criteria,
Cystectomy
comprising 13 793 patients. Of these, 21 used explicit definitions of POI, and they varied
Ileus widely. Across studies, the incidence of POI ranged from 1.58% to 23.5%. Possible risk
Review factors for POI included increasing age and body mass index. Seventeen studies reported
effects of an intervention on POI: 3 randomized controlled studies, 11 observational
cohort studies with concurrent comparison, and 3 observational cohort studies with
nonconcurrent comparison. Gum chewing was associated with shortened times to flatus
(2.4 vs 2.9 d; p < 0.0001) and bowel movement (BM) (3.2 vs 3.9 d; p < 0.001) in one
observational cohort study (n = 102); omission of a postoperative nasogastric tube (NGT)
was associated with shorter time to flatus (4.21 vs 5.33 d; p = 0.0001) and shorter length
of stay (14.4 vs 19.1 d; p = 0.001) in one observational cohort study (n = 430); and the
routine use of bowel preparation was associated with an increased incidence of POI (5%
vs 19%) in another series (n = 86). Additionally, readaptation of the dorsolateral perito-
neal layer was shown to shorten times to flatus ( p = 0.016) and times to BM ( p = 0.011) in
one randomized controlled study (n = 200).
Conclusions: The incidence/definition of POI after RC is highly variable. An improved
reporting strategy is needed to identify true incidence and risk factors, and to guide
future research for both potential preventive and therapeutic interventions.
# 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved.

* Corresponding author. Department of Urology, Division of Urologic Oncology, The University of


Texas Health Science Center San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229, USA.
Tel. +1 210 567 5676; Fax: +1 210 567 6868.
E-mail address: svatek@uthscsa.edu (R.S. Svatek).

1. Introduction high complication rate [1–3]. One of the most common


complications is postoperative paralytic ileus (POI) [2–5].
Despite major improvements in perioperative patient care, Transient cessation of bowel activity is expected following
radical cystectomy (RC) continues to be associated with a major abdominal surgery with small bowel motility and
0302-2838/$ – see back matter # 2012 European Association of Urology. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.eururo.2012.11.051
EUROPEAN UROLOGY 64 (2013) 588–597 589

gastric emptying returning quite quickly (within 24 and reported primary data. To be eligible for inclusion, studies
48 h, respectively). Large bowel recovery, however, is had to report one or more of the following: (1) incidence of
typically more protracted, taking up to 3–5 d to resume POI using an explicit definition or at least two quantitative
activity [6,7]. Despite the expectation of diminished measurements of GI recovery including LOS, time to bowel
intestinal motility following surgery, the period without sounds, time to flatus, time to bowel movement (BM), time to
signs of bowel function often extends beyond what is liquid diet/oral intake, time to regular diet; (2) potential risk
acceptable and is then diagnosed as POI. factors for the development of POI; and (3) interventions to
POI has been mainly studied in patients undergoing prevent or limit the incidence or duration of POI. Data from
general abdominal surgery. It has been shown to prolong cohorts represented in more than one publication were
the length of stay (LOS) by days to weeks and is associated only reported once. Exclusion criteria were studies with
with increased morbidity and costs [8–10]. Estimates of <10 patients, case reports, or those studies not reporting any
incidence vary in the general surgery literature, and clinical data on POI or GI recovery.
trials have been conducted to identify strategies to prevent Study quality and methodology for randomized con-
POI. A meta-analysis of the effect of postoperative gum trolled trials (RCTs) was assessed using the Jadad scale [17]
chewing revealed a decrease in the rate of POI and LOS [11]. and the Consolidated Standards for Reporting Trials
A Cochrane review of the effect of prokinetic agents on POI (CONSORT) statement [18]. In the Jadad scale, a maximum
found that certain medications reduce time to flatus and score of 5 points (2 for randomization, 2 for blinding, and 1
duration of hospital stay [12]. Epidural anesthesia, in for accountability of all patients) can be attained, and an
contrast to opiate use, was found to reduce POI but not overall score 3 indicates a high-quality study [17]. In the
length of hospital stay [13]. Alvimopan, a peripherally CONSORT scale, a 22-item checklist receives 0–2 points
acting m-opioid receptor antagonist, was found to reduce depending on the quality of the description (0 = no
time to flatus as well as accelerate time to hospital description, 1 = inadequate description, 2 = adequate de-
discharge [14]. Finally, early commencement of oral fluids scription). In observational cohort studies, quality was
has been shown to reduce time to bowel sounds and solid assessed using the Newcastle-Ottawa scale [19], where
diet intake and has demonstrated improvement toward studies are given up to 9 total points for meeting criteria in
shorter hospital stays [15]. selection, comparability, and exposure. We considered
To improve our understanding of POI as it relates studies as high quality if they received 8 or 9 points,
specifically to RC patients, we conducted a systematic medium quality if they received 6 or 7 points, and low
review of the evidence base on POI for patients undergoing quality if they received 5 points [19].
RC. We sought to capture the reported incidence of POI,
interventions to prevent POI, and risk factors for POI among 3. Evidence synthesis
RC populations. We also include standardized care path-
ways that, although they do not examine a specific 3.1. Overview
intervention, have similarly been used to optimize gastro-
intestinal (GI) recovery and/or the incidence of POI. Initially, our search yielded 727 potentially relevant
publications (Fig. 1). The titles of all 727 results were
2. Evidence acquisition reviewed, and if the publication could not be eliminated
based on title alone, the abstract was subsequently
We conducted a systematic review of Medline from 1966 to reviewed. The remaining 332 manuscripts were reviewed
February 2011 in adherence with the Preferred Reporting in full to determine eligibility. A total of 77 articles
Items for Systematic Reviews and Meta-analysis guidelines contained data in accordance with the eligibility criteria.
[16] using the following search criteria: From these 77 articles, we identified 13 793 patients who
underwent RC and who were noted to have experienced or
1. Medical Subject Headings (MeSH) terms: cystectomy, not experienced POI in the postoperative period. We
preoperative care or intraoperative care or postoperative identified studies that did not explicitly identify patients
care or perioperative care or postoperative period or with POI but did note various indicators of the return of
intraoperative complications or postoperative complications bowel function according to our criteria.
2. Ovid text word terms: preoperative care or intraoperative
care or postoperative care or perioperative care or 3.2. Incidence of postoperative paralytic ileus
postoperative period or intraoperative complications or
postoperative complications Of the 77 studies, 68 reported the incidence of POI. The rate
3. Linkage: Items 1 and 2, limited to human adults and of POI for cystectomy patients varied from 1.58% to 23.5%
publications in English with an average rate of POI of 9.86% for 13 793 patients.

We then conducted an additional search for additional 3.3. Definitions of postoperative paralytic ileus
articles using only the following text word terms: cystectomy
and ileus. We specifically used Ovid text word terms for the The definition for POI varied among reports. Only 21 of the
perioperative period to maximize the sensitivity of the search 77 studies meeting initial search criteria used explicit
strategy. Studies reviewed were limited to those that definitions of POI. Among RCTs, no definition of POI was
590 EUROPEAN UROLOGY 64 (2013) 588–597
[(Fig._1)TD$IG]

Initial Search: n = 727


(All potentially relevant articles reviewed Exclusions (n = 395):
by title initially; abstracts reviewed if
necessary to rule out relevance) 1. Case reports
2. Case series <10 patients
3. Not a study on radical
cystectomy
4. Functional/physiologic
data only; surgical atlas
n = 332; Remaining Articles
only
(All abstracts reviewed for relevance
and manuscript reviewed if needed to
rule out relevance)
Exclusion (n = 255):

1. Redundant data from


previously reviewed
study
2. Does not report on POI
or other criteria
n = 77; Manuscripts Reviewed in full
(Contain data meeting search criteria)

Fig. 1 – Flow diagram of the number of articles identified at each step of the systematic literature research. POI = postoperative paralytic ileus.

identified. Within observational cohort studies with con- be risk factors for the development of POI. Yamanaka and
current controls, definitions included no return of bowel colleagues [32], however, found no significant difference in
function after postoperative day (POD) 6 [20]; absence of POI among patients 80 yr of age (n = 72) compared with
flatus or stool on POD 6 [21,22]; absence of bowel function patients <80 yr (n = 557) in a cohort of RC patients. Chang
after POD 5 [23,24]; and postoperative nausea or vomiting and colleagues [2] demonstrated that ethnic minority, the
requiring cessation of oral intake, intravenous support, or existence of another major complication, and blood loss
NGT placement by POD 5 [25,26]. Within observational >600 ml were all significantly related to POI.
cohort studies with nonconcurrent controls, definitions
included absence of bowel function beyond POD 4 [27]; 3.5. Interventions to reduce postoperative paralytic ileus
return of bowel function after POD 5 [28]; absence of flatus
and/or bowel movement prolonging hospitalization beyond 3.5.1. Randomized controlled trials
discharge goal of POD 6–8 [2]; no evidence of bowel We identified three RCTs that studied a perioperative
function beyond the anticipated discharge goal of 6 d [29]; intervention on RC patients and reported data according to
lack of bowel activity >5 d following surgery [3]; and our criteria (Table 1). Both robot-assisted laparoscopic
inability to tolerate food by POD 5, the need to place a NGT, radical cystectomy (RALRC) and readaptation of the perito-
or the need to stop oral intake due to abdominal distension, neum were found to hasten the return of bowel function.
nausea, or emesis [30]. RALRC was found to have shorter times to flatus (2.3 vs 3.2 d;
p < 0.0001) and BM (3.2 vs 4.3 d; p = 0.0008) when compared
3.4. Risk factors for development of postoperative paralytic with open radical cystectomy (ORC) (n = 41) [33]. Read-
ileus aptation of the dorsolateral peritoneal layer was shown
(n = 200) to shorten times to flatus (reaching statistical
There are few observational cohort studies that identify significance on POD 3; p = 0.016) and times to BM (95% vs 84%
specific demographic or perioperative risk factors associat- had BM by POD 7; p = 0.011) [34]. Treatment with
ed with the development of POI in RC patients. Svatek and erythromycin, a prokinetic agent, showed no difference in
colleagues [29] identified increasing age and body mass bowel recovery compared with controls (n = 22) [35]. None of
index (BMI) as risk factors for POI in an observational cohort the three trials report explicitly on the incidence of POI.
study (25.6% of patients with a BMI 30 kg/m2 experienced
POI compared with 11.3% of patients with a BMI <30 kg/m2; 3.5.2. Observational cohort with concurrent comparison
p = 0.005). Lee and colleagues [31] observed that increasing We additionally identified 11 observational cohort studies
BMI is significantly associated with an increased incidence with concurrent comparison groups that reported data on
of POI. In addition to increasing age, Hollenbeck and POI or GI recovery (Table 2). Of the 11 studies, only 4
colleagues [3] noted a history of dyspnea, general anesthe- reported explicitly on the difference in POI rates between
sia (vs spinal anesthesia), and increasing operative time to cohorts.
EUROPEAN UROLOGY 64 (2013) 588–597 591

Table 1 – Randomized controlled trials

Study Definition No. of Intervention/ Results Conclusion Study


of ileus patients Comparison appraisal
*
Lightfoot Not defined 22; 11 received Postoperative erythromycin Median times, (+) vs ( ), Erythromycin is CONSORT score
et al. [35] erythromycin (+); therapy following ORC to normal diet (9 d vs 8 d; not useful in (30/44)
**
11 controls ( ) p = 0.60), bowel sounds reduction of POI Jadad score (5/5)
(2 d vs 3 d; p = 0.88), and
BM (6 d vs 5 d; p = 0.98)
not significantly different
*
Nix Not defined 41; 21 robotic Noninferiority study of RALRC Median times to: flatus Robotic approach CONSORT score
et al. [33] technique; compared with ORC (2.3 d vs 3.2 d; p < 0.0001) compares favorably (33/44)
**
20 open and to BM (3.2 d vs 4.3 d; with open approach Jadad score (2/5)
technique p = 0.0008) significantly for several perioperative
decreased in RALRC group measures including
recovery of bowel
function
*
Roth Not defined 200; 100 had the Bilateral readaptation of Times to flatus earlier in (+) Readaptation of the CONSORT score
et al. [34] peritoneum dorsolateral peritoneal layer group and reached statistical peritoneum following (32/44)
**
readapted by creating lateral flaps of significance ( p = 0.016) on radical cystectomy Jadad score (3/5)
bilaterally (+); peritoneum when accessing POD 3; 95% of (+) group vs hastens postoperative
100 controls the retroperitoneal space and 84% of ( ) group passed recovery of bowel
did not ( ) fixing them to the medial stool by POD 7 ( p = 0.011) function
peritoneal layer following ORC

BM = bowel movement; ORC = open radical cystectomy; POD = postoperative day; RALRC = robot-assisted laparoscopic radical cystectomy.
(+) received treatment; ( ) served as control.
*
CONSORT (Consolidated Standards of Reporting Trials) score: 22-item validated tool; maximum 44 points; score for each item: 0 = no description,
1 = inadequate description, 2 = adequate description.
**
Jadad score: Maximum score of 5 points (2 for randomization, 2 for blinding, and 1 for accountability of all patients) with an overall score 3 indicating a
high-quality study.

Both laparoscopic RC (LRC) and RALRC were associated they did demonstrate a higher risk of POI in those who had
with hastened return of bowel function when compared polyethylene glycol bowel preparation versus sodium
with ORC. Guillotreau and colleagues [23] demonstrated phosphate (40% vs 18%; p = 0.02; n = 101). Shafii and
that LRC was associated with a lower incidence of POI (11 vs colleagues [40] demonstrated a lower incidence of POI
30%; p = 0.043), shortened times to oral food intake (3.8 vs (5% vs 19%), shorter LOS (22.8 vs 31.6 d), and time to oral
6.4 d; p = 0.043), and shortened LOS (2.1 vs 4.4 d; p = 0.009) fluids (3.4 vs 5.8 d) when preoperative bowel preparation
when compared with ORC (n = 68). Although they did not was omitted altogether (n = 86).
report on the incidence of POI between groups, two studies With regard to the potential influence of urinary
[36,37] supported these findings, noting decreased times to diversion type on POI rates, Parekh and colleagues [41]
oral intake in the LRC group (3.3 vs 5.7 d; p < 0.05, and 2.8 vs found a higher incidence of POI in their ileal conduit
5 d; p = 0.004, respectively) compared with the ORC group versus ileal neobladder group (7.4% vs 2.6%). Conversely,
(n = 42 and n = 24, respectively). Additionally, one study Nieuwenhuijzen and colleagues [42] did not observe a
[38] demonstrated that RALRC was associated with shorter difference in POI incidence between ileal conduit, Indiana
times to regular diet (4 vs 5 d; p = 0.002) and LOS (5 vs 8 d; pouch, and neobladder in their series. The specific effect of
p = 0.007) when compared with ORC (n = 54). this variable on POI remains less clear.
Both prolonged use of NGT decompression and routine
bowel preparation were associated with exacerbated times 3.5.3. Observational cohort with nonconcurrent comparison
to bowel recovery. A large study by Inman and colleagues We identified three observational cohort studies with
[22] demonstrated that prolonged NGT tube decompression nonconcurrent comparisons, only one of which explicitly
was associated with prolonged time to GI recovery. In this reported the incidence of POI (Table 3). Both gum chewing
study, which included 430 patients, omission of routine and the use of RALRC were associated with earlier GI
NGT decompression after RC, compared with prolonged recovery. Kouba and colleagues [43] demonstrated an
nasogastric decompression, was associated with shorter association between postoperative gum chewing and
times to return of bowel sounds (1.99 vs 2.44 d; p = 0.0001), shorter times to flatus (2.4 vs 2.9 d; p < 0.0001) and BM
time to flatus (4.21 vs 5.33 d; p = 0.0001), and LOS (14.4 vs (3.2 vs 3.9 d; p < 0.001) when compared with those patients
19.1 d; p = 0.001). However, the incidence of POI was not not receiving gum following RC (n = 102). Additionally,
significantly different between groups. When combined Abraham and colleagues [44] (n = 34) found an association
with postoperative metoclopramide in a study by Donat and between RALRC, compared with LRC, and shorter times to
colleagues [39] (n = 81), NGT omission was again associated oral intake (2.3 vs 6.1; p = 0.012) and regular diet (3.7 vs 7.0;
with improved GI recovery according to similar parameters. p = 0.011). Combining these findings with that of the
Park and colleagues [21] did not show a significant decrease previous section, it appears that RALRC is associated with
in the incidence of POI with early NGT removal; however, hastened GI recovery when compared with both LRC and
592
Table 2 – Observational cohort with concurrent comparison

Study Definition of ileus No. of patients Intervention/comparison Results Conclusion Study appraisal
*
Park et al. [21] Absence of flatus 101; 20 had NGT removed Early NGT removal in patients POI in 23 patients (23%); rate of Early NGT removal is not associated Newcastle-Ottawa
or stool at POD 6 within 24 h of surgery (+); compared with prolonged NGT POI in (+) vs ( ) groups not with a decrease in POI; however, Scale (8/9)
(additionally, if 81 had NGT removed at decompression following open significantly different use of sodium phosphate for bowel
patients experienced first flatus ( ) radical cystectomy (25% vs 22%); risk of POI higher prep may reduce incidence of POI
nausea, emesis, or gas in those who had polyethylene
distention of abdominal glycol bowel prep vs sodium
x-ray, they were phosphate (40% vs 18%; p = 0.02)
considered to have ileus)
*
Inman et al. [22] Absence of flatus 430; 199 did not receive Effect on GI complications of Mean times to: return of BS NGT decompression after RC with Newcastle-Ottawa
or stool on POD 6 postoperative NGT (+); omission of NGT decompression (1.99 d vs 2.44 d; p = 0.0001), UD may prolong GI recovery, Scale (7/9)
221 received postoperative compared with patients who return of flatus (4.21 d vs increasing hospital stay and should
NGT ( ) received NGT decompression 5.33 d; p = 0.0001), and LOS not be used routinely
following ORC (14.4 d vs 19.1 d; p = 0.001)
shorter in ( ) group; POI

EUROPEAN UROLOGY 64 (2013) 588–597


incidence not significantly different
*
Donat et al. [39] Not defined 81; 27 received metoclopramide Effect of omission of NGT Times to return of normal bowel The combination of early NG Newcastle-Ottawa
combined with early NGT decompression and administration sounds (2.9 d vs 4.0 d; p = 0.0002) removal and metoclopramide may Scale (9/9)
removal (+); 54 received no of metoclopramide on GI recovery and resumption of solid diet hasten return of bowel function
metoclopramide and had compared with controls (6.7 d vs 7.9 d; p = 0.04) shorter
prolonged NGT decompression ( ) in the (+) group.
*
Shafii et al. [40] Not defined 86; 22 received no bowel Effect of omission of preoperative Incidence of POI lower in (+) Routine bowel preparation is Newcastle-Ottawa
preparation (+); 64 received bowel preparation on GI group (5% vs 19%) associated with increased hospital Scale (7/9)
standard bowel preparation ( ) complications compared with LOS (22.8 d vs 31.6 d) and time to stay and POI; it does not improve
controls who received preoperative fluids (3.4 d vs 5.8 d) both shorter surgical outcome
bowel preparation prior to ORC in (+) group
*
Wang et al. [47] Not defined 51; 31 had HALRC; 20 had LRC Clinical outcomes of HALRC LOS longer in LRC than HALRC LOS was significantly lower in Newcastle-Ottawa
compared with controls (19.7 d vs 24.7 d; p = 0.040); time HALRC patients; the choice of Scale (8/9)
undergoing LRC to oral intake not significantly urinary diversion in this group
different was main reason for this
*
Guillotreau Absence of normal 68; 38 had LRC; 30 had ORC Morbidity and mortality of LRC LRC group had lower POI incidence Laparoscopy decreases POI, Newcastle-Ottawa
et al. [23] bowel function compared with controls (11 vs 30%; p = 0.043); LRC group shortens time to fluid intake and Scale (8/9)
after POD 5 undergoing ORC had shorter times to fluid intake oral diet, and thus decreases the
(2.1 vs 4.4 d; p = 0.009), oral food time to resumption of normal
intake (3.8 vs 6.4 d; p = 0.006) and bowel function and LOS
shorter postoperative LOS (12.7 d vs
15.6 d; p = 0.033) than ORC group
*
Wang et al. [38] Not defined 54; 33 had RALRC; 21 had ORC Perioperative outcomes of RALRC RALRC group had shorter time to RALRC associated with hastened Newcastle-Ottawa
compared with controls regular diet (4 d vs 5 d; p = 0.002) return of bowel function Scale (8/9)
undergoing ORC and LOS (5 vs 8 d; p = 0.007) than
ORC group
*
Porpiglia Not defined 42; 20 had LRC; 22 had ORC Perioperative outcomes of LRC Overall POI incidence 4.2%; only LRC may be associated with early Newcastle-Ottawa
et al. [36] compared with controls time to oral intake significantly recovery of bowel function Scale (9/9)
undergoing ORC lower in the LRC group (3.3 vs 5.7 d;
p < 0.05)
*
Basillote Not defined 24; 13 had LRC; 11 had ORC Perioperative outcomes of LRC LRC group had shorter time to oral LRC associated with quicker return Newcastle-Ottawa
et al. [37] compared with controls diet (2.8 d vs 5 d; p = 0.004) and of bowel function and decreased Scale (7/9)
undergoing ORC LOS (5.1 vs 8.4 d; p = 0.0004) hospital stay
than ORC group
EUROPEAN UROLOGY 64 (2013) 588–597 593

Maximum score of 9 points (4 for selection, 2 for comparability, and 3 for outcome). Studies considered high quality if they received 8 or 9 points; medium quality if they received 6 or 7 points, and low quality if they
ORC, respectively. Metoclopramide was associated with a

GI = gastrointestinal; HALRC = hand-assisted laparoscopic radical cystectomy; LOS = length of stay; LRC = laparoscopic radical cystectomy; NGT = nasogastric tube; ORC = open radical cystectomy; NGT = nasogastric tube;
Newcastle-Ottawa

Newcastle-Ottawa
Study appraisal
reduction in nausea and vomiting but was not significantly
Scale (8/9) associated with the incidence of POI (n = 80) [45].

Scale (9/9)
3.5.4. Standardized care pathways
Standardized care pathways use perioperative standardized
*

order sets, protocols, and/or nursing care to decrease the


Ileal conduit diversion associated

variability associated with perioperative management.


No difference in POI incidence
between ileal conduit, Indiana
with higher incidence of POI

Maffezzini and colleagues [27] applied a care pathway to


a cohort of 107 patients to determine if specific consider-
Conclusion

pouch, 3.2% for orthotopic neobladder, pouch, and neobladder

ations would reduce POI. Components of their care pathway


include no preoperative fasting, no mechanical bowel
preparation, prevention of hypovolemia/hypothermia, pain
control, early postoperative nutrition, and no NGT postop-
eratively. They concluded that decreased median times to
BM and regular diet could be achieved with the care
pathway. They observed a POI incidence of 17.7%.
and 2.0% for sexually preserving RC
Higher incidence of POI in the ileal

Chang and colleagues [2] also examined the effects of a


conduit vs ileal neobladder group

Rates for early POI were 3.4% for

pathway designed to optimize perioperative outcomes in RC


ileal conduit, 2.0% for Indiana

patients. In addition to a standardized operative approach


and postoperative laboratory order sets, patients arrived the
Results

day of surgery after completing bowel preparation at home,


with neobladder

went directly to the surgical floor postoperatively, were not


(7.4% vs 2.6%)

allowed oral intake until bowel function resumed, and did


not receive routine NGT placement. The study concluded
that their complication rates were comparatively favorable
with other institutions when using such pathway. Their
observed incidence of POI was 17%.
Indiana pouch; 62 had orthotopic following the four most widely
Intervention/comparison

orthotopic diversion compared

The value of a fast-track program to optimize periopera-


with controls undergoing ileal

281; 118 had ileal conduit; 52 had Early and late complications

tive care for RC patients was described by Pruthi and


Morbidity and mortality of

colleagues [45]. POI incidence was not included in the


results independently, and instead it was grouped with
POD = postoperative day; POI = postoperative ileus; RC = radical cystectomy; UD = urinary diversion.
conduit diversion

nausea and vomiting as ‘‘GI complications.’’ As a whole,


used UD types

these complications combined for an incidence of 12% in the


final 100 patients studied, and the fast-track program was
deemed favorable in regard to the return of bowel function.

3.6. Discussion
preserving RC with neobladder
diversion; 81 had ileal conduit

neobladder; 50 had sexually

POI has long been recognized as a significant complication


No. of patients
198; 117 had orthotopic

of any major abdominal surgery. In a recent observational


cohort of >17 000 patients undergoing colectomy in the
United States [9], ‘‘ileus, digestive complications, or both’’
occurred in 17.4% of patients. More strikingly, the POI group
diversion

accounted for an average LOS of 4.9 additional days and


overall average additional cost of $8182 compared with the
(+) received treatment; ( ) served as control.

non-POI group. Goldstein [10] found that the annual cost of


treating POI in the United States reaches up to $1.42 billion.
Definition of ileus

Patients undergoing RC are uniquely different among


Inability to resume

demographically broad and pathologically diverse abdomi-


nal surgery patients. Unlike most bowel/abdominal surgery,
oral feeding

Not defined
for >1 wk

RC deperitonealizes the pelvis for extended periods of time


to perform an adequate pelvic lymphadenectomy; patients
Table 2 (Continued )

are older and have a high incidence of comorbid conditions.


received 5 points.
Nieuwenhuijzen

This, combined with the knowledge that almost 100% of


patients undergoing this procedure have a pelvic malig-
et al. [41]

et al. [42]

nancy and are frequently smokers, further specifies the


Parekh
Study

patients we see. Other factors reinforcing the uniquely


categorical nature of RC are the extensive operative times
*
594 EUROPEAN UROLOGY 64 (2013) 588–597

Table 3 – Observational cohort with nonconcurrent comparison

Study Definition No. of patients Comparison Results Conclusion Study


of ileus appraisal
*
Kouba Not defined 102: 51 received Effect of postoperative Gum chewing may
Times to flatus (2.4 vs 2.9 d; Newcastle-Ottawa
et al. [43] gum to chew gum chewing in p < 0.0001) and to BM enhance bowel Scale (7/9)
postoperatively (+); patients following recovery and resumption
(3.2 vs 3.9 d; p < 0.001)
51 received no ORC of function after radical
shorter in (+) group than
gum ( ) controls cystectomy and urinary
diversion
Pruthi Not defined 80; 40 received Postoperative Incidence of nausea/vomiting The use of metoclopramide *Newcastle-Ottawa
et al. [45] metoclopramide (+); metoclopramide lower (2.5% vs 12.5%; reduced rates of nausea Scale (6/9)
40 did not ( ) in patients p = 0.011) in the (+) group; and vomiting and possibly
following ORC POI incidence and length of gastrointestinal
stay were not significantly complications, although
different no definite effect on POI
or resumption of bowel
function was shown
*
Abraham Not defined 34; 14 patients had Compare perioperative RALRC group had shorter time RALRC associated with Newcastle-Ottawa
et al. [44] RALRC; 20 controls outcomes in RALRC to oral intake (2.3 vs 6.1 d; reduced time to Scale (8/9)
had LRC compared with controls p = 0.012) and regular diet resumption of normal
undergoing LRC (3.7 vs 7.0 d; p = 0.011) bowel function; hospital
than LRC group stay was not decreased
by RALRC

BM = bowel movement; LRC = laparoscopic radical cystectomy; ORC = open radical cystectomy; POI = postoperative paralytic ileus; RALRC = robot-assisted
laparoscopic radical cystectomy.
(+) received treatment; ( ) served as control.
*
Maximum score of 9 points (4 for selection, 2 for comparability, and 3 for outcome). Studies were considered high quality if they received 8 or 9 points,
medium quality if they received 6 or 7 points, and low quality if they received 5 points.

often required: 4.73–9.4 h (mean: 6.11 h) [5,26,46,47] and capture secondary measures of POI including time to flatus,
the presence of urine in the operative field, which has been time to first BM, time to toleration of first oral intake, and LOS.
theorized to delay the resumption of colonic motility [48]. These outcomes have less heterogeneous definitions and
We conducted a systematic review of the literature would be more amenable to the application of statistical
related to the incidence of POI and interventions to reduce techniques in the future such as in a meta-analysis.
POI as it relates to RC patients. Our review found three We also found that most of the information on
striking findings: (1) There is a paucity of studies addressing prophylaxis, prevention, and treatment of POI in RC
POI following RC, (2) POI in RC populations is ill defined, and populations comes from observational cohort studies, with
(3) there are very limited data about interventions to scarce data coming from experimental controlled studies.
prevent/reduce POI in patients undergoing RC. This area of Studies by Roth and colleagues [34] and by Nix and
research remains largely unexplored. colleagues [33] do shed light on the benefits of readaptation
Our review demonstrates an incidence of POI across all of the peritoneum and robotic techniques, respectively,
studies (9.86%) that is lower than data reported by several with regard to hastened bowel recovery. Additionally,
other authors (12–25.4%) [2,21,27,29,45,49], and we found Lightfoot and colleagues [35] conducted a RCT demonstrat-
that the reporting of POI is highly variable, observer ing that erythromycin did not hasten the return of bowel
dependent, and without a standard definition. Interestingly, function in RC patients. However, that study had a sample
the care pathways designed to optimize the return of bowel size of 22 patients and lacked substantial power to detect a
function also all reported relatively higher rates of POI. We small difference in treatment effect.
speculate that this is because of improved data collection A recently completed large RCT aimed to study the
with attention directed toward addressing POI directly. The effectiveness of alvimopan versus placebo among a RC
lower rates reflected by our review, which were remarkably population [50]. As noted earlier, alvimopan is a peripher-
variable and as low as 2% in one study, suggest possible ally acting m-opioid receptor antagonist, which because of
underreporting of POI by many published studies. its limited ability to cross the blood–brain barrier, is
Throughout our review we observed that the definition reported to avert the undesirable intestinal side effects of
of POI is not given or is inconsistent. When stated, the opioid agonists without affecting analgesia. The results of
definition varied considerably, in many cases grouped with this study will provide invaluable information regarding
more general GI complications and bearing a confusing this potential preventive option.
alternative nomenclature (eg, subileus). However, in most Robotic surgery has recently changed the face of
cases, it is not defined at all. Given that POI is generally a emerging literature. In a 2010 prospective cohort, Ng and
clinical diagnosis, it is critical to clearly report the definition colleagues [25] demonstrated that when compared with an
being used because it may vary significantly. Alternatively, open group, robotic RC had a significantly lower 30-d
an improved reporting strategy may be used routinely to complication rate. Our review identified several studies
EUROPEAN UROLOGY 64 (2013) 588–597 595

comparing a robotic approach with both a laparoscopic and somewhat subjective nature of the current validated tools.
open approach. All of these showed a trend toward a Although inadequate methodology and poor reporting
hastened return of bowel function in the robotic groups, but quality are usually coupled with bias and thus provide an
none were RCTs. A multi-institutional prospective random- indirect measure of it, there can certainly be instances of
ized trial to evaluate open versus robotic RC is underway, studies adhering to the highest of reporting standards yet
and the results will be highly informative [51]. they have fundamental flaws and biases. Finally, essentially
Identifying risk factors for POI is also important for all of these series come from single-institution studies in
counseling and targeting perioperative management. We academic centers with relatively high individual surgeon
reviewed the available literature addressing risk factors for volume, which might limit the interpretation and gener-
POI. Specifically, increasing age and BMI, ethnic minority, alizability of their findings.
increasing estimated blood loss/transfusion requirement, To better understand POI, so that we might curtail the
and the presence of a major complication were all impact it has on RC patients, future studies are needed
associated with POI in RC patients. This information is specifically targeting RC populations. It is critical to revisit
the cumulative knowledge from only five observational those interventions that this review has identified as
cohort studies. As hypothesized by Svatek and colleagues beneficial in curtailing POI by conducting prospective
[29], BMI and age may be associated with increasing POI controlled trials specifically for RC patients.
because of increased times to ambulation and more
extensive/difficult bowel manipulation. They also noted 4. Conclusions
increased overall rates of complications in these groups,
consistent with the findings of Chang and colleagues [2] The incidence and definition of POI after RC is highly
that the development of a major complication is an variable. An improved reporting strategy is needed to
independent risk factor for POI. Conversely, we did not identify true incidence and risk factors and, most important,
find any evidence that POI independently predicts occur- to guide future research for both potential preventive and
rence of other complications. The available literature therapeutic interventions.
regarding age and POI is conflicting, and addressing the
association of age with POI is challenging because it is
possible that elderly patients, especially those with cogni- Author contributions: Robert S. Svatek had full access to all the data in
tive impairment, do not reliably report markers of bowel the study and takes responsibility for the integrity of the data and the
function such as flatus and thus compound the variability of accuracy of the data analysis.
such measurements. Standardization in data collection and
Study concept and design: Svatek, Lawrence.
reporting of POI and secondary measures is imperative as Acquisition of data: Ramirez, McIntosh, Strehlow.
we move forward to allow for accurate comparisons among Analysis and interpretation of data: Svatek, Lawrence, Parekh, Ramirez,
populations and interventions. McIntosh, Strehlow.
At this time, it would be difficult to give solid Drafting of the manuscript: Svatek, Lawrence, Ramirez, McIntosh,
recommendations on POI prevention given the paucity of Strehlow.
data. It seems that gum chewing is a simple measure that Critical revision of the manuscript for important intellectual content: Svatek,
could potentially help prevent POI with minimal risk. Based Lawrence, Parekh, Ramirez, McIntosh, Strehlow.
on the available data, it also appears that omission of a NGT Statistical analysis: Svatek.
Obtaining funding: None.
may be beneficial in that it can lead to decreased LOS and
Administrative, technical, or material support: Svatek.
not result in an increased risk of POI. Bowel preparation also
Supervision: Svatek.
seems to be unnecessary and may appear to be detrimental
Other (specify): None.
in some of these cases. It also appears that, when possible,
readaptation of the peritoneum may hasten bowel recovery. Financial disclosures: Robert S. Svatek certifies that all conflicts of
However, it is important to acknowledge that, given the interest, including specific financial interests and relationships and
nonrandomized nature of these studies and the lack of affiliations relevant to the subject matter or materials discussed in the
standardized pathways used, potential confounders exist, manuscript (eg, employment/affiliation, grants or funding, consultan-
cies, honoraria, stock ownership or options, expert testimony, royalties,
particularly for the control cohorts that may have had a
or patents filed, received, or pending), are the following: None.
significant impact on the findings reported. Standardization
of perioperative management via care pathways has the Funding/Support and role of the sponsor: None.
potential to decrease adverse events, improve care, and
decrease health care expense. As far as robotic RC versus
ORC, it would be prudent to wait for the results of the multi-
institutional prospective RTC before making any recom- References
mendations. [1] Lawrentschuk N, Colombo R, Hakenberg OW, et al. Prevention and
This review certainly has limitations. Although we management of complications following radical cystectomy for
designed our search criteria to be very flexible and far bladder cancer. Eur Urol 2010;57:983–1001.
reaching, we acknowledge that we may have missed [2] Chang SS, Cookson MS, Baumgartner RG, Wells N, Smith Jr JA.
important factors that might influence POI. Our assessment Analysis of early complications after radical cystectomy: results
of the methodological quality of the studies is limited by the of a collaborative care pathway. J Urol 2002;167:2012–6.
596 EUROPEAN UROLOGY 64 (2013) 588–597

[3] Hollenbeck BK, Miller DC, Taub D, et al. Identifying risk factors for [23] Guillotreau J, Game X, Mouzin M, et al. Radical cystectomy for
potentially avoidable complications following radical cystectomy. bladder cancer: morbidity of laparoscopic versus open surgery.
J Urol 2005;174:1231–7, discussion 1237. J Urol 2009;181:554–9, discussion 559.
[4] Shabsigh A, Korets R, Vora KC, et al. Defining early morbidity of [24] Ramani VAC, Maddineni SB, Grey BR, Clarke NW. Differential
radical cystectomy for patients with bladder cancer using a stan- complication rates following radical cystectomy in the irradiated
dardized reporting methodology. Eur Urol 2009;55:164–76. and nonirradiated pelvis. Eur Urol 2010;57:1058–63.
[5] Svatek RS, Fisher MB, Matin SF, et al. Risk factor analysis in a [25] Ng CK, Kauffman EC, Lee MM, et al. A comparison of postoperative
contemporary cystectomy cohort using standardized reporting complications in open versus robotic cystectomy. Eur Urol 2010;
methodology and adverse event criteria. J Urol 2010;183: 57:274–82.
929–34. [26] Lowrance WT, Rumohr JA, Chang SS, Clark PE, Smith Jr JA, Cookson
[6] Resnick J, Greenwald DA, Brandt LJ. Delayed gastric emptying and MS. Contemporary open radical cystectomy: analysis of periopera-
postoperative ileus after nongastric abdominal surgery: part I. Am J tive outcomes. J Urol 2008;179:1313–8, discussion 1318.
Gastroenterol 1997;92:751–62. [27] Maffezzini M, Campodonico F, Canepa G, Gerbi G, Parodi D. Current
[7] Resnick J, Greenwald DA, Brandt LJ. Delayed gastric emptying and perioperative management of radical cystectomy with intestinal
postoperative ileus after nongastric abdominal surgery: part II. Am J urinary reconstruction for muscle-invasive bladder cancer and
Gastroenterol 1997;92:934–40. reduction of the incidence of postoperative ileus. Surg Oncol
[8] Schuster TG, Montie JE. Postoperative ileus after abdominal surgery. 2008;17:41–8.
Urology 2002;59:465–71. [28] Kang SG. Robot-assisted radical cystectomy and pelvic lymph node
[9] Iyer S, Saunders WB, Stemkowski S. Economic burden of postoper- dissection: a multi-institutional study from Korea. J Endourol
ative ileus associated with colectomy in the United States. J Manage 2010;24:1435–40.
Care Pharm 2009;15:485–94. [29] Svatek RS, Fisher MB, Williams MB, et al. Age and body mass index
[10] Goldstein JL, Matuszewski KA, Delaney CP, et al. Inpatient economic are independent risk factors for the development of postoperative
burden of postoperative ileus associated with abdominal surgery in paralytic ileus after radical cystectomy. Urology 2010;76:
the United States. P&T 2007;32:82–90. 1419–24.
[11] Noble EJ, Harris R, Hosie KB, Thomas S, Lewis SJ. Gum chewing [30] Khan MS. Analysis of early complications of robotic-assisted radical
reduces postoperative ileus? A systematic review and meta-analy- cystectomy using a standardized reporting system. Urology
sis. Int J Surg 2009;7:100–5. 2011;77:357–62.
[12] Traut U, Brugger L, Kunz R, et al. Systemic prokinetic pharmacologic [31] Lee CT, Dunn RL, Chen BT, Joshi DP, Sheffield J, Montie JE. Impact of
treatment for postoperative adynamic ileus following abdominal body mass index on radical cystectomy. J Urol 2004;172:1281–5.
surgery in adults. Cochrane Database Syst Rev 2008:CD004930. [32] Yamanaka K, Miyake H, Hara I, Inoue TA, Fujisawa M. Significance of
[13] Marret E, Remy C, Bonnet F. Meta-analysis of epidural analgesia radical cystectomy for bladder cancer in patients over 80 years old.
versus parenteral opioid analgesia after colorectal surgery. Br J Surg Int Urol Nephrol 2007;39:209–14.
2007;94:665–73. [33] Nix J, Smith A, Kurpad R, Nielsen ME, Wallen EM, Pruthi RS.
[14] Wolff BG, Michelassi F, Gerkin TM, et al. Alvimopan, a novel, Prospective randomized controlled trial of robotic versus open
peripherally acting mu opioid antagonist: results of a multicenter, radical cystectomy for bladder cancer: perioperative and patholog-
randomized, double-blind, placebo-controlled, phase III trial of ic results. Eur Urol 2010;57:196–201.
major abdominal surgery and postoperative ileus. Ann Surg [34] Roth B, Birkhäuser FD, Zehnder P, Burkhard FC, Thalmann GN,
2004;240:728–34, discussion 734–5. Studer UE. Readaptation of the peritoneum following extended
[15] Charoenkwan K, Phillipson G, Vutyavanich T. Early versus delayed pelvic lymphadenectomy and cystectomy has a significant benefi-
(traditional) oral fluids and food for reducing complications after cial impact on early postoperative recovery and complications:
major abdominal gynaecologic surgery. Cochrane Database Syst results of a prospective randomized trial. Eur Urol 2011;59:204–10.
Rev 2007:CD004508. [35] Lightfoot AJ, Eno M, Kreder KJ, O’Donnell MA, Rao SS, Williams RD.
[16] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items Treatment of postoperative ileus after bowel surgery with low-dose
for systematic reviews and meta-analyses: the PRISMA statement. intravenous erythromycin. Urology 2007;69:611–5.
Int J Surg 2010;8:336–41. [36] Porpiglia F, Renard J, Billia M, et al. Open versus laparoscopy-assisted
[17] Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports radical cystectomy: results of a prospective study. J Endourol 2007;
of randomized clinical trials: is blinding necessary? Control Clin 21:325–9.
Trials 1996;17:1–12. [37] Basillote JB, Abdelshehid C, Ahlering TE, Shanberg AM. Laparoscopic
[18] Altman DG, Schulz KF, Moher D, et al. The revised CONSORT assisted radical cystectomy with ileal neobladder: a comparison
statement for reporting randomized trials: explanation and elabo- with the open approach. J Urol 2004;172:489–93.
ration. Ann Intern Med 2001;134:663–94. [38] Wang GJ, Barocas DA, Raman JD, Scherr DS. Robotic vs open radical
[19] Wells G, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale cystectomy: prospective comparison of perioperative outcomes
(NOS) for assessing the quality of nonrandomized studies in meta- and pathological measures of early oncological efficacy. BJU Int
analyses. Paper presented at: 3rd Symposium on Systematic 2008;101:89–93.
Reviews: Beyond the Basics; July 3–5, 2000; Oxford, UK. [39] Donat SM, Slaton JW, Pisters LL, Swanson DA. Early nasogastric tube
[20] Brössner C. Does extended lymphadenectomy increase the mor- removal combined with metoclopramide after radical cystectomy
bidity of radical cystectomy? BJU Int 2004;93:64–6. and urinary diversion. J Urol 1999;162:1599–602.
[21] Park HK, Kwak C, Byun SS, Lee E, Lee SE. Early removal of nasogastric [40] Shafii M, Murphy DM, Donovan MG, Hickey DP. Is mechanical
tube after cystectomy with urinary diversion: does postoperative bowel preparation necessary in patients undergoing cystectomy
ileus risk increase? Urology 2005;65:905–8. and urinary diversion? BJU Int 2002;89:879–81.
[22] Inman BA, Harel F, Tiguert R, Lacombe L, Fradet Y. Routine naso- [41] Parekh DJ, Gilbert WB, Koch MO, Smith Jr JA. Continent urinary
gastric tubes are not required following cystectomy with urinary reconstruction versus ileal conduit: a contemporary single-
diversion: a comparative analysis of 430 patients. J Urol 2003; institution comparison of perioperative morbidity and mortality.
170:1888–91. Urology 2000;55:852–5.
EUROPEAN UROLOGY 64 (2013) 588–597 597

[42] Nieuwenhuijzen JA, de Vries RR, Bex A, et al. Urinary diversions [47] Wang SZ, Chen LW, Chen W, Chen JX, Lin HY. Hand-assisted versus
after cystectomy: the association of clinical factors, complications pure laparoscopic radical cystectomy: a clinical outcome compari-
and functional results of four different diversions. Eur Urol son. Int J Urol 2009;16:360–3.
2008;53:834–44, discussion 842–4. [48] Mattei A, Birkhaeuser FD, Baermann C, Warncke SH, Studer UE. To
[43] Kouba EJ, Wallen EM, Pruthi RS. Gum chewing stimulates bowel stent or not to stent perioperatively the ureteroileal anastomosis of
motility in patients undergoing radical cystectomy with urinary ileal orthotopic bladder substitutes and ileal conduits? Results of a
diversion. Urology 2007;70:1053–6. prospective randomized trial. J Urol 2008;179:582–6.
[44] Abraham JB, Young JL, Box GN, Lee HJ, Deane LA, Ornstein DK. [49] Jentzmik F, Schostak M, Stephan C, et al. Extraperitoneal radical
Comparative analysis of laparoscopic and robot-assisted radical cystectomy with extraperitonealization of the ileal neobladder: a
cystectomy with ileal conduit urinary diversion. J Endourol comparison to the transperitoneal technique. World J Urol 2010;
2007;21:1473–80. 28:457–63.
[45] Pruthi RS, Nielsen M, Smith A, Nix J, Schultz H, Wallen EM. Fast track [50] A study of alvimopan for the management of postoperative ileus in
program in patients undergoing radical cystectomy: results in 362 subjects undergoing radical cystectomy. ClinicalTrials.gov Web site.
consecutive patients. J Am Coll Surg 2010;210:93–9. www.clinicaltrials.gov/ct2/show/NCT00708201.
[46] Huang J, Lin T, Xu K, et al. Laparoscopic radical cystectomy with [51] Open vs robotic-assisted radical cystectomy: a randomized trial.
orthotopic ileal neobladder: a report of 85 cases. J Endourol 2008; ClinicalTrials.gov Web site. www.clinicaltrials.gov/ct2/show/
22:939–46. NCT01157676.