Accepted Manuscript The use of mechanical bowel preparation in laparoscopic gynecologic surgery: A decision analysis Kelly L. Kantartzis, MD, MSc, Jonathan P. Shepherd, MD, MSc PII:
S0002-9378(15)00480-9
DOI:
10.1016/j.ajog.2015.05.017
Reference:
YMOB 10400
To appear in:
American Journal of Obstetrics and Gynecology
Received Date: 20 December 2014 Revised Date:
26 April 2015
Accepted Date: 10 May 2015
Please cite this article as: Kantartzis KL, Shepherd JP, The use of mechanical bowel preparation in laparoscopic gynecologic surgery: A decision analysis, American Journal of Obstetrics and Gynecology (2015), doi: 10.1016/j.ajog.2015.05.017. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 1
The use of mechanical bowel preparation in laparoscopic gynecologic surgery: A
2
decision analysis
3
Kelly L. KANTARTZIS, MD, MSc
5
Jonathan P. SHEPHERD, MD, MSc
6
Division of Urogynecology, Department of Obstetrics, Gynecology, and Reproductive
7
Sciences, Magee Womens Hospital, University of Pittsburgh School of Medicine,
8
Pittsburgh, PA
9
M AN U
SC
RI PT
4
Corresponding author and contact for reprint requests:
11
Jonathan P. Shepherd, MD, MSc
12
300 Halket Street
13
Pittsburgh, PA 15213
14
[email protected]
15
Work Phone: 412-641-1440
16
Cell Phone: 412-926-5175
17
Fax: 412-641-1133
EP
AC C
18
TE D
10
19
Disclosure: The authors report no conflict of interest.
20
Financial Support: None
21
ACCEPTED MANUSCRIPT 2
Paper presentation: This research was presented in an oral presentation at the 41st
23
Annual Scientific Meeting for the Society of Gynecologic Surgeons (SGS), Orlando, FL,
24
March 22-25, 2015
25
Word Count: Abstract: 250 ; Main text: 1,738
26 27
SC
28 29
M AN U
30 31 32 33
38 39 40 41 42 43 44
EP
37
AC C
36
TE D
34 35
RI PT
22
ACCEPTED MANUSCRIPT 3
Condensation
46
Omitting mechanical bowel preparation was preferred, but the difference between this
47
and magnesium citrate was not clinically significant.
48 49
Short Version of Title
50
Bowel Preparation in Gynecologic Laparoscopy
SC
51 52
M AN U
53 54 55 56
61 62 63 64 65 66 67
EP
60
AC C
59
TE D
57 58
RI PT
45
ACCEPTED MANUSCRIPT 4
Abstract
69
Objective:
70
The use of mechanical bowel preparation prior to laparoscopy is common in
71
gynecology, but its use may affect rates of perioperative events and complications. Our
72
objective was to compare different mechanical bowel preparations using decision
73
analysis techniques to determine the optimal preparation prior to laparoscopic
74
gynecologic surgery.
SC
RI PT
68
75
Study Design:
77
A decision analysis was constructed modelling perioperative outcomes with the
78
following mechanical bowel preparations: magnesium citrate, sodium phosphate,
79
polyethylene glycol, enema, and no bowel preparation. Comparisons were made using
80
published utility values. Secondary analyses included the percentages that had ≥1
81
preoperative event and ≥1 intra- or postoperative complication.
TE D
82
M AN U
76
Results:
84
Overall, the highest utility values were for no bowel preparation (0.98) and magnesium
85
citrate (0.97), while the other values were: enema (0.95), sodium phosphate (0.94),
86
polyethylene glycol (0.91). The difference between no bowel preparation and
87
magnesium citrate was less than published minimally important differences for utilities,
88
so there is likely no real difference between these strategies. The probability of having
89
at least one preoperative event was lowest for no bowel preparation (1%) while the
AC C
EP
83
ACCEPTED MANUSCRIPT 5 90
probability of having at least one intra- or postoperative complication was lowest with
91
magnesium citrate (8%).
92
Conclusions:
94
The highest utilities were seen with no bowel preparation, but the absolute difference
95
between no bowel preparation and magnesium citrate was less than the minimally
96
important difference. With similar overall utilities, our model raises questions as to
97
whether mechanical bowel preparation is a necessary step prior to laparoscopic
98
gynecologic surgery. However, if a surgeon prefers a bowel preparation, magnesium
99
citrate is the preferred option.
100 101
Key Words: Gynecologic surgery, laparoscopy, mechanical bowel preparation
106 107 108 109 110 111 112
EP
105
AC C
104
TE D
102 103
M AN U
SC
RI PT
93
ACCEPTED MANUSCRIPT 6
Introduction
114
The use of mechanical bowel preparation prior to laparoscopy is common in gynecology
115
as well as other surgical specialties, yet studies across specialties have questioned its
116
merit.1-3 Historically, gynecologists have prescribed preoperative mechanical bowel
117
preparations in attempts to decrease the risk of infection, while also providing easier
118
bowel manipulation and better visualization.4 However, many of these proposed
119
benefits have never been proven, and bowel preparations may actually increase the risk
120
of surgical site infection.2, 4, 5 Despite such data, a survey of gynecologic oncologists
121
found that approximately half of the respondents still prescribe bowel preparations
122
despite 77% acknowledging that there is not data to support such use.4
M AN U
SC
RI PT
113
123
Two recent randomized trials have shown that for both vaginal prolapse repairs and
125
laparoscopic hysterectomies, mechanical bowel preparation with saline enemas
126
conferred no benefit for surgeon visualization and ultimately decreased patient
127
satisfaction.6, 7 This is similar to studies of both simple and more complex gynecologic
128
laparoscopy which also found no difference in the surgical field or operative difficulty if
129
either an oral sodium phosphate or saline enema was used preoperatively.1, 3
130
Unfortunately, these studies used various mechanical bowel preparations and various
131
routes of surgery thus making comparisons difficult. Our objective was to compare
132
multiple different mechanical bowel preparations, including no mechanical bowel
133
preparation in a decision analysis model, to determine the optimal bowel preparation
134
prior to laparoscopic gynecologic surgery.
135
AC C
EP
TE D
124
ACCEPTED MANUSCRIPT 7
Materials and Methods
137
After institutional review board approval was obtained, a decision analysis model was
138
created using TreeAge Pro (TreeAge Software, Inc, Williamstown, MA, USA). The
139
decision node included the following possible mechanical bowel preparations:
140
magnesium citrate, sodium phosphate, polyethylene glycol, enema, or no mechanical
141
bowel preparation. Each subsequent subtree was identical for each of the 5 bowel
142
preparation options. The differences in the subtrees were in the probability of each
143
perioperative event or complication occurring with the different bowel preparations.
SC
RI PT
136
M AN U
144
The first branch point in each subtree was a specific preoperative, intraoperative, or
146
postoperative complication/event dichotomized to present or absent. Subsequent
147
braches were for other possible complications. For example, with 2 complications there
148
were 4 terminal branches in the tree representing 4 possible outcomes. You could have
149
both complications, neither, only the first, or only the second. This simplified version of
150
the tree is displayed in Figure 1. Using this methodology, our tree was exponentially
151
expanded accounting for all nine perioperative events. The model was based on a 7-
152
day follow-up as most surgical complications related to bowel preparation are noted in
153
this time frame. This shorter time frame improved our ability to distinguish differences
154
between the preparations.
EP
AC C
155
TE D
145
156
The model included weighted average probabilities for perioperative events and
157
complications from 28 published trials. When laparoscopic gynecologic trials were not
158
available, colorectal, urologic, and general surgery literature was used. Eight trials taken
ACCEPTED MANUSCRIPT 8
from the colonoscopy literature evaluated preoperative patient outcomes specifically
160
related to bowel preparations due to the greater robustness of preoperative outcomes
161
which would occur regardless of whether a colonoscopy or gynecologic surgery follows
162
the bowel preparation. Our model included 5 preoperative events and complications
163
including nausea/vomiting/abdominal pain, diarrhea/fecal incontinence, anal discomfort,
164
hypotension, and arrhythmia/seizure. We also modelled 4 intra- or postoperative
165
complications including surgical site infection, abdominal infection, ileus/obstruction,
166
and bowel injury/colostomy. The base case probability was a composite of all available
167
data. Utility values were assigned to each perioperative event and were obtained from
168
published data. Utilities which represent a measure of quality of life ranged from 0
169
representing death to 1 representing perfect health.
M AN U
SC
RI PT
159
170
The primary objective was to determine which mechanical bowel preparation yielded the
172
highest utility value and thus was the preferred treatment option. Secondary analyses
173
include the percentage of patients that had at least one preoperative event or
174
complication and the percentage with at least one intra- or postoperative complication.
175
Multiple one-way sensitivity analyses were performed to test model robustness and to
176
determine if a threshold value existed where the results would then favor a different
177
bowel preparation. These analyses were carried out over the entire range of possible
178
probability values for events and complications as well as utilities, 0 to 1.
180
EP
AC C
179
TE D
171
Results
ACCEPTED MANUSCRIPT 9
Complication rates and utility values were obtained from the published literature (Table
182
1). Using simple roll-back methodology, the overall utility value for each mechanical
183
bowel preparation was calculated. This method calculates the average utility value a
184
patient will experience given the weighted average of each perioperative event and the
185
resultant decrease in quality of life utility if each of these events occurs. The highest
186
utility values were for no mechanical bowel preparation (0.98) and magnesium citrate
187
(0.97), while the other values were as follows: enema (0.95), sodium phosphate (0.94),
188
and polyethylene glycol (0.91). The difference between no mechanical bowel
189
preparation and magnesium citrate was less than the published minimally important
190
differences for health state utilities.31 Thus, there is likely no real difference between
191
these strategies.
M AN U
SC
RI PT
181
192
When the decision analysis tree was reduced to only account for preoperative events,
194
all preparations had similar utility values: no mechanical bowel preparation (1.00),
195
magnesium citrate (0.98), sodium phosphate (0.98), enema (0.98), and polyethylene
196
glycol (0.98). Conversely, magnesium citrate (0.99), no mechanical bowel preparation
197
(0.98), and enema (0.98) had higher utilities when the tree was reduced to account only
198
for intraoperative or postoperative events. Sodium phosphate (0.96) and polyethylene
199
glycol (0.92) were lower in this reduced model.
EP
AC C
200
TE D
193
201
The probability of having at least one preoperative event was lowest for no mechanical
202
bowel preparation (1%) versus any of the mechanical bowel preparations [magnesium
203
citrate (40%), polyethylene glycol (41%), sodium phosphate (55%), and enema (60%).
ACCEPTED MANUSCRIPT 10
The probability of having at least one intra- or postoperative complication was lowest
205
with magnesium citrate (8%) compared to no bowel preparation (16%) and other
206
mechanical bowel preparations [enema (12%), sodium phosphate (22%), and
207
polyethylene glycol (46%)].
RI PT
204
208
Multiple one-way sensitivity analyses were performed for the 12 probabilities at each of
210
the 5 branches and additionally for all 10 utilities. The purpose of a sensitivity analysis is
211
to determine if there is an alternate value for one of the input variables where the model
212
would favor a different strategy. When our variables were sampled throughout the entire
213
range of 0 to 1, the only thresholds identified changed the model from favoring no bowel
214
preparation to favoring magnesium citrate. No mechanical bowel preparation and
215
magnesium citrate are not meaningfully different due to their differential overall utility
216
being less than the minimally important difference for utilities. Thus, there were no
217
meaningful thresholds on one-way sensitivity analysis.
218
TE D
M AN U
SC
209
Comment
220
No mechanical bowel preparation had the highest utility value, making this option the
221
preferred strategy prior to laparoscopic gynecologic surgery. However, there is likely no
222
difference when this option is compared to magnesium citrate. This study confirms
223
previously published data across specialties and surgical routes, highlighting that the
224
elimination of mechanical bowel preparations does not negatively impact perioperative
225
outcomes.5,7,32 Given that patients report higher satisfaction when bowel preparation is
226
omitted6, that surgeons are unable to decipher between those who do and do not have
AC C
EP
219
ACCEPTED MANUSCRIPT 11 227
a bowel preparation3, 7, and that bowel preparations may delay return of postoperative
228
bowel function33, the routine use of mechanical bowel preparations should likely be
229
eliminated.
RI PT
230
However, there are still surgeons who maintain that certain patients must undergo a
232
preoperative bowel preparation. While no bowel preparation is preferred in our decision
233
analysis, there was no real difference between no bowel preparation and magnesium
234
citrate. Thus, if a surgeon desires a bowel preparation magnesium citrate is the
235
preferred option given it has the lowest probability of pre-, intra-, and postoperative
236
events compared to other mechanical bowel preparations in our model.
M AN U
SC
231
237
Despite the mounting evidence against routine use of bowel preparations, many
239
surgeons still prefer to give their patients preoperative bowel preparations.4 This lag
240
between evidence and physician practice is demonstrated across medical specialties,
241
but can be more pronounced in surgical disciplines.34 Studies have specifically
242
addressed changing behavior of general surgeons regarding bowel preparations and
243
found that at least in the short-term, a structured multi-modal educational strategy can
244
improve compliance with bowel preparation recommendations.35 As more literature is
245
published on this topic prior to gynecologic surgery, a similar education strategy will
246
likely be needed to promote physician understanding and adoption.
EP
AC C
247
TE D
238
248
A strength of our study is the use of multiple randomized controlled trials that
249
specifically evaluated bowel preparations prior to abdominal surgery and predominantly
ACCEPTED MANUSCRIPT 12
with a laparoscopic approach. Also, our multiple one-way sensitivity analyses
251
demonstrated that varying any single variable would not change our conclusions and
252
thus confirmed model robustness. A major limitation of this study is that the model
253
includes literature from a wide variety of surgical specialties and not just gynecology.
254
However, many of the variables included in the model, especially preoperative events,
255
do not rely on the proposed surgical procedure. For example, diarrhea and fecal
256
incontinence will occur with a given probability for a given mechanical bowel preparation
257
no matter what surgery is planned the following day. We ultimately decided to use
258
these alternative sources to improve our model where studies in laparoscopic
259
gynecology were lacking.
260
M AN U
SC
RI PT
250
Another limitation is that our model primarily included objective events and did not
262
include the potential subjective benefits of bowel preparations such as surgeon
263
perception of surgical difficulty or patient satisfaction. We would propose that surgical
264
difficulty would correlate with rates of bowel injury or obstruction and surgical infections,
265
variables which were accounted for in our model. Prior research has also found that
266
surgeons are poor predictors of which patients have had a bowel preparation.1 The use
267
of utility values is a method for incorporating patient satisfaction as quality of life is
268
measured and will decrease when events that impact patient satisfaction occur. Further
269
incorporating patient satisfaction would likely make our preferred strategy of no bowel
270
preparation a more favorable option as patients report higher satisfaction rates when no
271
bowel preparation is used.6 Finally, our model does not specifically discuss the insertion
272
of graft or mesh and the implications a bowel preparation has on potential negative
AC C
EP
TE D
261
ACCEPTED MANUSCRIPT 13
outcomes. Given the paucity of literature on the potential complications of mesh
274
placement at the time of a bowel injury, this was excluded and our model does not
275
adequately answer this question. Further research surrounding placement of graft is
276
needed.
RI PT
273
277
Our model adds to the current literature which questions the use of routine mechanical
279
bowel preparations prior to laparoscopic gynecologic surgery. Based on our findings,
280
surgeons can most likely omit preoperative bowel preparation. While surgeon behavior
281
is difficult to change, the use of a structured educational tool will likely ease the
282
transition.
M AN U
SC
278
283 284
288 289 290 291 292
EP
287
AC C
286
TE D
285
ACCEPTED MANUSCRIPT 14 293
References
294
1.
Muzzi L, Bellati F, Zullo MA, Manci N, Aangioli R, Panici PB. Mechanical bowel preparation before gynecologic laparoscopy: a randomized, single-blind,
296
controlled trial. Fertil Steril 2006;85:689-93.
297
2.
RI PT
295
Slim K, Vicaut E, Launay-Savary MV, Contant C, Chipponi J. Updated systematic review and meta-analysis of randomized clinical trials on the role of mechanical
299
bowel preparation before colorectal surgery. Annals of surgery 2009;249:203-9.
300
3.
SC
298
Yang LC, Aarden D, Lee TT, et al. Mechanical bowel preparation for gynecologic laparoscopy: a prospective randomized trial of oral sodium phosphate solution vs
302
single sodium phosphate enema. J Minim Invasive Gynecol 2011;18:149-56.
303
4.
M AN U
301
Wells T, Plante M, McAlpine JN. Preoperative bowel preparation in gynecologic oncology: a review of practice patterns and an impetus to change. International
305
journal of gynecological cancer : official journal of the International Gynecological
306
Cancer Society 2011;21:1135-42.
307
5.
TE D
304
Oliveira L, Wexner SD, Daneil N, et al. Mechanical bowel preparation for elective colorectal surgery. A prospective, randomized, surgeon-blinded trial comparing
309
sodium phosphate and polyethylene glycol-based oral lavage solutions. Diseases
310
of the colon and rectum 1997;40:585-91.
312 313
6.
AC C
311
EP
308
Ballard AC, Parker-Autry CY, Markland AD, Varner RE, Huisingh C, Richter HE.
Bowel preparation before vaginal prolapse surgery: a randomized controlled trial. Obstet Gynecol 2014;123:232-8.
ACCEPTED MANUSCRIPT 15 314
7.
Siedhoff MT, Clark LH, Hobbs KA, Findley AD, Moulder JK, Garrett JM.
315
Mechanical bowel preparation before laparoscopic hysterectomy: a randomized
316
controlled trial. Obstet Gynecol 2014;123:562-7. 8
Vradelis S, Kalaitakis E, Sharifi Y, et al. Addition of senna improves quality of
RI PT
317 318
colonoscopy preparation with magnesium citrate. World journal of
319
gastroenterology : WJG 2009;15:1759-63. 9.
Berkelhammer C, Ekambaram A, Silva RG. Low-volume oral colonoscopy bowel
SC
320
preparation: sodium phosphate and magnesium citrate. Gastrointestinal
322
endoscopy 2002;56:89-94.
323
10.
M AN U
321
Choi YS, Suh JP, Kim JK, et al. Magnesium citrate with a single dose of sodium
324
phosphate for colonoscopy bowel preparation. World journal of gastroenterology
325
: WJG 2011;17:242-8. 11.
Kao D, Lalor E, Sandha G, et al. A randomized controlled trial of four
TE D
326 327
precolonoscopy bowel cleansing regimens. Canadian journal of gastroenterology
328
= Journal canadien de gastroenterologie 2011;25:657-62. 12.
Jansen SV, Goedhard JG, Winkens B, VAN Deursen CT. Preparation before
EP
329
colonoscopy: a randomized controlled trial comparing different regimes.
331
European journal of gastroenterology & hepatology 2011;23:897-902.
332 333 334 335
13.
AC C
330
VAN Vugt VAN Pinxteren MW, VAN Kouwen MC, VAN Oijen
MG, VAN Achterberg T,
Nagengast FM. A prospective study of bowel preparation for colonoscopy with polyethylene glycol-electrolyte solution versus sodium phosphate in Lynch syndrome: a randomized trial. Familial cancer 2012;11:337-41.
ACCEPTED MANUSCRIPT 16 336
14.
McKenna T, Macgill A, Porat G, Friedenberg FK. Colonoscopy preparation: polyethylene glycol with Gatorade is as safe and efficacious as four liters of
338
polyethylene glycol with balanced electrolytes. Digestive diseases and sciences
339
2012;57:3098-105.
340
15.
RI PT
337
Hjelkrem M, Stengel J, Liu M, Jones DP, Harrison SA. MiraLAX is not as effective as GoLytely in bowel cleansing before screening colonoscopies. Clinical
342
gastroenterology and hepatology : the official clinical practice journal of the
343
American Gastroenterological Association 2011;9:326-32 e1. 16.
Forster JA, Thomas WM. Patient preferences and side effects experienced with
M AN U
344
SC
341
345
oral bowel preparations versus self-administered phosphate enema. Annals of
346
the Royal College of Surgeons of England 2003;85:185-6.
347
17.
Gold MR, Franks P, McCoy KI, Fryback DG. Toward consistency in cost-utility analyses: using national measures to create condition-specific values. Medical
349
care 1998;36:778-92.
350
18.
TE D
348
Tufts University Cost-Effectiveness Analysis Registry. Tufts Medical Center Institute for Clinical Research and Health Policy Studies. https://research.tufts-
352
nemc.org/cear4/. Accessed 1 December 2011. 19.
mechanical bowel preparation: a randomized prospective trial. Annals of surgery
354
2003;237:363-7.
355 356 357 358 359
Zmora O, Mahajna A, Bar-Zakai B, et al. Colon and rectal surgery without
AC C
353
EP
351
20.
Teng TO, Wallace A. One thousand health-related quality-of-life estimates.
Medical Care. 2000;38(6):583-637.
ACCEPTED MANUSCRIPT 17 360
21.
Won H, Maley P, Salim S, Rao A, Campbell NT, Abbott JA. Surgical and patient
361
outcomes using mechanical bowel preparation before laparoscopic gynecologic
362
surgery: a randomized controlled trial. Obstet Gynecol 2013;121:538-46. 22.
Miettinen RP, Laitinen ST, Makela JT, Paakkonen ME. Bowel preparation with
RI PT
363
oral polyethylene glycol electrolyte solution vs. no preparation in elective open
365
colorectal surgery: prospective, randomized study. Diseases of the colon and
366
rectum 2000;43:669-75; discussion 75-7.
367
23.
SC
364
Fa-Si-Oen P, Roumen R, Buitenweg J, et al. Mechanical bowel preparation or not? Outcome of a multicenter, randomized trial in elective open colon surgery.
369
Diseases of the colon and rectum 2005;48:1509-16.
370
24.
M AN U
368
Large MC, Kiriluk KJ, Decastro GJ, et al. The impact of mechanical bowel preparation on postoperative complications for patients undergoing cystectomy
372
and urinary diversion. The Journal of urology 2012;188:1801-5.
373
25.
TE D
371
Pena-Soria MJ, Mayol JM, Anula R, Arbeo-Escolar A, Fernandez-Represa JA. Single-blinded randomized trial of mechanical bowel preparation for colon
375
surgery with primary intraperitoneal anastomosis. Journal of gastrointestinal
376
surgery : official journal of the Society for Surgery of the Alimentary Tract
377
2008;12:2103-8; discussion 08-9.
379 380
26.
AC C
378
EP
374
Sugihara T, Yasunaga H, Horiguchi H, et al. Does mechanical bowel preparation
improve quality of laparoscopic nephrectomy? Propensity score-matched analysis in Japanese series. Urology 2013;81:74-9.
ACCEPTED MANUSCRIPT 18 381
27.
Ram E, Sherman Y, Weil R, Vishne T, Kravarusic D, Dreznik Z. Is mechanical
382
bowel preparation mandatory for elective colon surgery? A prospective
383
randomized study. Archives of surgery 2005;140:285-8. 28.
ChaleokittiI B. Comparative study between polyethylene glycol and sodium
RI PT
384 385
phosphate solution in elective colorectal surgery. Journal of the Medical
386
Association of Thailand = Chotmaihet thangphaet 2002;85:92-6. 29.
Itani KM, Wilson SE, Awad SS, Jensen EH, Finn TS, Abramson MA. Polyethylene
SC
387
glycol versus sodium phosphate mechanical bowel preparation in elective
389
colorectal surgery. American journal of surgery 2007;193:190-4.
390
30.
M AN U
388
Shepherd JP, Lowder JL, Jones KA, Smith KJ. Retropubic and transobturator
391
midurethral slings: a decision analysis to compare outcomes including efficacy
392
and complications. International urogynecology journal 2010;21:787-93. 31.
Walters SJ, Brazier JE. Comparison of the minimally important difference for two
TE D
393
health state utility measures: EQ-5D and SF-6D. Quality of life research : an
395
international journal of quality of life aspects of treatment, care and rehabilitation
396
2005;14:1523-32. 32.
before ileal urinary diversion: a systematic review and meta-analysis. Urologia
398
internationalis 2014;92:339-48.
399 400 401 402
Deng S, Dong Q, Wang J, Zhang P. The role of mechanical bowel preparation
AC C
397
EP
394
33.
Yamada T, Kan H, Matsumoto S, et al. Dysmotility by mechanical bowel
preparation using polyethylene glycol. The Journal of surgical research 2014;191:84-90.
ACCEPTED MANUSCRIPT 19 403
34.
the journey from evidence to effect. BMJ 2003;327:33-5.
404 405
Davis D, Evans M, Jadad A, et al. The case for knowledge translation: shortening
35.
Eskicioglu C, GagliardiI A, Fenech DS, Victor CJ, McLeod RS. Can a tailored knowledge translation strategy improve short term outcomes? A pilot study to
407
increase compliance with bowel preparation recommendations in general
408
surgery. Surgery 2011;150:68-74.
SC
409
RI PT
406
410
M AN U
411
AC C
EP
TE D
412
ACCEPTED MANUSCRIPT
20
RI PT
Table 1. Outcome probabilities and utility values
Outcome Probability (Range in Published Studies)
SC
No Outcome
Magnesium Sodium Mechanical
Enema
0.3055
Nausea/
0.1307
0.767 0.3067
(0.0313-
0.0000*
Diarrhea/ Fecal
EP
Preoperative
AC C
Anal Discomfort
0.3265#16 0.86)17,
0.6226)12-15 13
18
0.769
0.2857$
0.0052# 19
Incontinence
(0.68(0.0700-
0.4771) 9-
0.1688)8,9
Abdominal Pain
Event
(0.0214-
TE D
Vomiting/
0.0000*
Value
Phosphate Glycol(238g)
M AN U
Citrate Bowel Prep
Utility
Polyethylene
(0.610.2857#16
0.0749#3,8
0.3673#16
16,19
0.95)17, 18, 20
0.0214#9
0.0438#9
0.0469^
0.0499^
0.700#17
ACCEPTED MANUSCRIPT
21
Hypotension
0.0000*
0.0143#9
Arrhythmia/
0.0000 (0-
0.0438#9
0.0143$ 9
0.0000*
0.593#18
0.0000#9 Seizure
0.0063#9
21,22
0.0078)
RI PT
0.835 0.0145#22
0.0000*
(0.720.95)17
0.2012
SC
0.0832
0.0728 (0Surgical Site
(0.0667-
(0.0145-
Infection
0.12)3, 27,
M AN U
0.0236#26 19, 22-
0.1719)
0.3488)
0.0411#3
0.900#20
19, 22-
28
27
Intraoperative/ Abdominal
(0.0309-
0.1059
TE D
0.0622
Infection
0.32)22, 24, 26, 27
EP
Complications
0.696 0.2634
(0.0139-
0.0279&
(0.642(0.1812-
0.0685#3
0.1463)3,
0.75) 18, 0.3714)
22, 24
27
20
0.0636
0.0196
Ileus/
AC C
Postoperative
29
(0.0118-
Obstruction
0.650 (0.0139-
0.0285$
0.0374#19
0.0137#3
(0.6-
3,
0.0854)
0.0667)19,26,27
0.7)18 27
ACCEPTED MANUSCRIPT
22
0.00466 Bowel
0.0021 (00.0011$
(0.0396-
0.0000#3
0.550#30
RI PT
Injury/Colostomy 0.1467) 24, 26
0.0000#3
SC
0.0667) 24, 29
M AN U
Note: Outcome probabilities were derived from weighted averages of 28 published articles. Utilities were gathered from 3 articles and one publically available online database. Probabilities and utilities have a possible range of 0 to 1.
AC C
EP
TE D
#= Single value reported in the literature so a range is not reported *= Values for No mechanical bowel prep and enema were assumed to be 0 for certain preoperative outcomes given lack of mechanism to cause this event with no or minimal intervention. $= This value was not available in the literature so baseline value was assumed to be median of other bowel preparation values. This was accounted for in sensitivity analysis by varying over entire range of 0-1. ^= This value was not available in the literature so baseline value was calculated using the ratio of nausea/ vomiting/ abdominal pain to anal discomfort from the magnesium citrate and sodium phosphate preparations. The known value for nausea/ vomiting/ abdominal pain was then divided by this multiplier of 6.5412. This assumption was accounted for in sensitivity analysis by varying over entire range of 0-1. &= This value was not available in the literature so baseline value was calculated using the ratio of surgical site infection to abdominal infection from other four preparations. The known value for surgical site infection was then divided by this multiplier of 0.8472. This assumption was accounted for in sensitivity analysis by varying over entire range of 0-1.
ACCEPTED MANUSCRIPT 23
Figure 1. Simplified decision analysis tree
The tree modeling all possible complications is too complex to graphically illustrate
AC C
EP
TE D
M AN U
SC
RI PT
here.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0002-9378(15)00480-9
DOI:
10.1016/j.ajog.2015.05.017
Reference:
YMOB 10400
To appear in:
American Journal of Obstetrics and Gynecology
Received Date: 20 December 2014 Revised Date:
26 April 2015
Accepted Date: 10 May 2015
Please cite this article as: Kantartzis KL, Shepherd JP, The use of mechanical bowel preparation in laparoscopic gynecologic surgery: A decision analysis, American Journal of Obstetrics and Gynecology (2015), doi: 10.1016/j.ajog.2015.05.017. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 1
The use of mechanical bowel preparation in laparoscopic gynecologic surgery: A
2
decision analysis
3
Kelly L. KANTARTZIS, MD, MSc
5
Jonathan P. SHEPHERD, MD, MSc
6
Division of Urogynecology, Department of Obstetrics, Gynecology, and Reproductive
7
Sciences, Magee Womens Hospital, University of Pittsburgh School of Medicine,
8
Pittsburgh, PA
9
M AN U
SC
RI PT
4
Corresponding author and contact for reprint requests:
11
Jonathan P. Shepherd, MD, MSc
12
300 Halket Street
13
Pittsburgh, PA 15213
14
[email protected]
15
Work Phone: 412-641-1440
16
Cell Phone: 412-926-5175
17
Fax: 412-641-1133
EP
AC C
18
TE D
10
19
Disclosure: The authors report no conflict of interest.
20
Financial Support: None
21
ACCEPTED MANUSCRIPT 2
Paper presentation: This research was presented in an oral presentation at the 41st
23
Annual Scientific Meeting for the Society of Gynecologic Surgeons (SGS), Orlando, FL,
24
March 22-25, 2015
25
Word Count: Abstract: 250 ; Main text: 1,738
26 27
SC
28 29
M AN U
30 31 32 33
38 39 40 41 42 43 44
EP
37
AC C
36
TE D
34 35
RI PT
22
ACCEPTED MANUSCRIPT 3
Condensation
46
Omitting mechanical bowel preparation was preferred, but the difference between this
47
and magnesium citrate was not clinically significant.
48 49
Short Version of Title
50
Bowel Preparation in Gynecologic Laparoscopy
SC
51 52
M AN U
53 54 55 56
61 62 63 64 65 66 67
EP
60
AC C
59
TE D
57 58
RI PT
45
ACCEPTED MANUSCRIPT 4
Abstract
69
Objective:
70
The use of mechanical bowel preparation prior to laparoscopy is common in
71
gynecology, but its use may affect rates of perioperative events and complications. Our
72
objective was to compare different mechanical bowel preparations using decision
73
analysis techniques to determine the optimal preparation prior to laparoscopic
74
gynecologic surgery.
SC
RI PT
68
75
Study Design:
77
A decision analysis was constructed modelling perioperative outcomes with the
78
following mechanical bowel preparations: magnesium citrate, sodium phosphate,
79
polyethylene glycol, enema, and no bowel preparation. Comparisons were made using
80
published utility values. Secondary analyses included the percentages that had ≥1
81
preoperative event and ≥1 intra- or postoperative complication.
TE D
82
M AN U
76
Results:
84
Overall, the highest utility values were for no bowel preparation (0.98) and magnesium
85
citrate (0.97), while the other values were: enema (0.95), sodium phosphate (0.94),
86
polyethylene glycol (0.91). The difference between no bowel preparation and
87
magnesium citrate was less than published minimally important differences for utilities,
88
so there is likely no real difference between these strategies. The probability of having
89
at least one preoperative event was lowest for no bowel preparation (1%) while the
AC C
EP
83
ACCEPTED MANUSCRIPT 5 90
probability of having at least one intra- or postoperative complication was lowest with
91
magnesium citrate (8%).
92
Conclusions:
94
The highest utilities were seen with no bowel preparation, but the absolute difference
95
between no bowel preparation and magnesium citrate was less than the minimally
96
important difference. With similar overall utilities, our model raises questions as to
97
whether mechanical bowel preparation is a necessary step prior to laparoscopic
98
gynecologic surgery. However, if a surgeon prefers a bowel preparation, magnesium
99
citrate is the preferred option.
100 101
Key Words: Gynecologic surgery, laparoscopy, mechanical bowel preparation
106 107 108 109 110 111 112
EP
105
AC C
104
TE D
102 103
M AN U
SC
RI PT
93
ACCEPTED MANUSCRIPT 6
Introduction
114
The use of mechanical bowel preparation prior to laparoscopy is common in gynecology
115
as well as other surgical specialties, yet studies across specialties have questioned its
116
merit.1-3 Historically, gynecologists have prescribed preoperative mechanical bowel
117
preparations in attempts to decrease the risk of infection, while also providing easier
118
bowel manipulation and better visualization.4 However, many of these proposed
119
benefits have never been proven, and bowel preparations may actually increase the risk
120
of surgical site infection.2, 4, 5 Despite such data, a survey of gynecologic oncologists
121
found that approximately half of the respondents still prescribe bowel preparations
122
despite 77% acknowledging that there is not data to support such use.4
M AN U
SC
RI PT
113
123
Two recent randomized trials have shown that for both vaginal prolapse repairs and
125
laparoscopic hysterectomies, mechanical bowel preparation with saline enemas
126
conferred no benefit for surgeon visualization and ultimately decreased patient
127
satisfaction.6, 7 This is similar to studies of both simple and more complex gynecologic
128
laparoscopy which also found no difference in the surgical field or operative difficulty if
129
either an oral sodium phosphate or saline enema was used preoperatively.1, 3
130
Unfortunately, these studies used various mechanical bowel preparations and various
131
routes of surgery thus making comparisons difficult. Our objective was to compare
132
multiple different mechanical bowel preparations, including no mechanical bowel
133
preparation in a decision analysis model, to determine the optimal bowel preparation
134
prior to laparoscopic gynecologic surgery.
135
AC C
EP
TE D
124
ACCEPTED MANUSCRIPT 7
Materials and Methods
137
After institutional review board approval was obtained, a decision analysis model was
138
created using TreeAge Pro (TreeAge Software, Inc, Williamstown, MA, USA). The
139
decision node included the following possible mechanical bowel preparations:
140
magnesium citrate, sodium phosphate, polyethylene glycol, enema, or no mechanical
141
bowel preparation. Each subsequent subtree was identical for each of the 5 bowel
142
preparation options. The differences in the subtrees were in the probability of each
143
perioperative event or complication occurring with the different bowel preparations.
SC
RI PT
136
M AN U
144
The first branch point in each subtree was a specific preoperative, intraoperative, or
146
postoperative complication/event dichotomized to present or absent. Subsequent
147
braches were for other possible complications. For example, with 2 complications there
148
were 4 terminal branches in the tree representing 4 possible outcomes. You could have
149
both complications, neither, only the first, or only the second. This simplified version of
150
the tree is displayed in Figure 1. Using this methodology, our tree was exponentially
151
expanded accounting for all nine perioperative events. The model was based on a 7-
152
day follow-up as most surgical complications related to bowel preparation are noted in
153
this time frame. This shorter time frame improved our ability to distinguish differences
154
between the preparations.
EP
AC C
155
TE D
145
156
The model included weighted average probabilities for perioperative events and
157
complications from 28 published trials. When laparoscopic gynecologic trials were not
158
available, colorectal, urologic, and general surgery literature was used. Eight trials taken
ACCEPTED MANUSCRIPT 8
from the colonoscopy literature evaluated preoperative patient outcomes specifically
160
related to bowel preparations due to the greater robustness of preoperative outcomes
161
which would occur regardless of whether a colonoscopy or gynecologic surgery follows
162
the bowel preparation. Our model included 5 preoperative events and complications
163
including nausea/vomiting/abdominal pain, diarrhea/fecal incontinence, anal discomfort,
164
hypotension, and arrhythmia/seizure. We also modelled 4 intra- or postoperative
165
complications including surgical site infection, abdominal infection, ileus/obstruction,
166
and bowel injury/colostomy. The base case probability was a composite of all available
167
data. Utility values were assigned to each perioperative event and were obtained from
168
published data. Utilities which represent a measure of quality of life ranged from 0
169
representing death to 1 representing perfect health.
M AN U
SC
RI PT
159
170
The primary objective was to determine which mechanical bowel preparation yielded the
172
highest utility value and thus was the preferred treatment option. Secondary analyses
173
include the percentage of patients that had at least one preoperative event or
174
complication and the percentage with at least one intra- or postoperative complication.
175
Multiple one-way sensitivity analyses were performed to test model robustness and to
176
determine if a threshold value existed where the results would then favor a different
177
bowel preparation. These analyses were carried out over the entire range of possible
178
probability values for events and complications as well as utilities, 0 to 1.
180
EP
AC C
179
TE D
171
Results
ACCEPTED MANUSCRIPT 9
Complication rates and utility values were obtained from the published literature (Table
182
1). Using simple roll-back methodology, the overall utility value for each mechanical
183
bowel preparation was calculated. This method calculates the average utility value a
184
patient will experience given the weighted average of each perioperative event and the
185
resultant decrease in quality of life utility if each of these events occurs. The highest
186
utility values were for no mechanical bowel preparation (0.98) and magnesium citrate
187
(0.97), while the other values were as follows: enema (0.95), sodium phosphate (0.94),
188
and polyethylene glycol (0.91). The difference between no mechanical bowel
189
preparation and magnesium citrate was less than the published minimally important
190
differences for health state utilities.31 Thus, there is likely no real difference between
191
these strategies.
M AN U
SC
RI PT
181
192
When the decision analysis tree was reduced to only account for preoperative events,
194
all preparations had similar utility values: no mechanical bowel preparation (1.00),
195
magnesium citrate (0.98), sodium phosphate (0.98), enema (0.98), and polyethylene
196
glycol (0.98). Conversely, magnesium citrate (0.99), no mechanical bowel preparation
197
(0.98), and enema (0.98) had higher utilities when the tree was reduced to account only
198
for intraoperative or postoperative events. Sodium phosphate (0.96) and polyethylene
199
glycol (0.92) were lower in this reduced model.
EP
AC C
200
TE D
193
201
The probability of having at least one preoperative event was lowest for no mechanical
202
bowel preparation (1%) versus any of the mechanical bowel preparations [magnesium
203
citrate (40%), polyethylene glycol (41%), sodium phosphate (55%), and enema (60%).
ACCEPTED MANUSCRIPT 10
The probability of having at least one intra- or postoperative complication was lowest
205
with magnesium citrate (8%) compared to no bowel preparation (16%) and other
206
mechanical bowel preparations [enema (12%), sodium phosphate (22%), and
207
polyethylene glycol (46%)].
RI PT
204
208
Multiple one-way sensitivity analyses were performed for the 12 probabilities at each of
210
the 5 branches and additionally for all 10 utilities. The purpose of a sensitivity analysis is
211
to determine if there is an alternate value for one of the input variables where the model
212
would favor a different strategy. When our variables were sampled throughout the entire
213
range of 0 to 1, the only thresholds identified changed the model from favoring no bowel
214
preparation to favoring magnesium citrate. No mechanical bowel preparation and
215
magnesium citrate are not meaningfully different due to their differential overall utility
216
being less than the minimally important difference for utilities. Thus, there were no
217
meaningful thresholds on one-way sensitivity analysis.
218
TE D
M AN U
SC
209
Comment
220
No mechanical bowel preparation had the highest utility value, making this option the
221
preferred strategy prior to laparoscopic gynecologic surgery. However, there is likely no
222
difference when this option is compared to magnesium citrate. This study confirms
223
previously published data across specialties and surgical routes, highlighting that the
224
elimination of mechanical bowel preparations does not negatively impact perioperative
225
outcomes.5,7,32 Given that patients report higher satisfaction when bowel preparation is
226
omitted6, that surgeons are unable to decipher between those who do and do not have
AC C
EP
219
ACCEPTED MANUSCRIPT 11 227
a bowel preparation3, 7, and that bowel preparations may delay return of postoperative
228
bowel function33, the routine use of mechanical bowel preparations should likely be
229
eliminated.
RI PT
230
However, there are still surgeons who maintain that certain patients must undergo a
232
preoperative bowel preparation. While no bowel preparation is preferred in our decision
233
analysis, there was no real difference between no bowel preparation and magnesium
234
citrate. Thus, if a surgeon desires a bowel preparation magnesium citrate is the
235
preferred option given it has the lowest probability of pre-, intra-, and postoperative
236
events compared to other mechanical bowel preparations in our model.
M AN U
SC
231
237
Despite the mounting evidence against routine use of bowel preparations, many
239
surgeons still prefer to give their patients preoperative bowel preparations.4 This lag
240
between evidence and physician practice is demonstrated across medical specialties,
241
but can be more pronounced in surgical disciplines.34 Studies have specifically
242
addressed changing behavior of general surgeons regarding bowel preparations and
243
found that at least in the short-term, a structured multi-modal educational strategy can
244
improve compliance with bowel preparation recommendations.35 As more literature is
245
published on this topic prior to gynecologic surgery, a similar education strategy will
246
likely be needed to promote physician understanding and adoption.
EP
AC C
247
TE D
238
248
A strength of our study is the use of multiple randomized controlled trials that
249
specifically evaluated bowel preparations prior to abdominal surgery and predominantly
ACCEPTED MANUSCRIPT 12
with a laparoscopic approach. Also, our multiple one-way sensitivity analyses
251
demonstrated that varying any single variable would not change our conclusions and
252
thus confirmed model robustness. A major limitation of this study is that the model
253
includes literature from a wide variety of surgical specialties and not just gynecology.
254
However, many of the variables included in the model, especially preoperative events,
255
do not rely on the proposed surgical procedure. For example, diarrhea and fecal
256
incontinence will occur with a given probability for a given mechanical bowel preparation
257
no matter what surgery is planned the following day. We ultimately decided to use
258
these alternative sources to improve our model where studies in laparoscopic
259
gynecology were lacking.
260
M AN U
SC
RI PT
250
Another limitation is that our model primarily included objective events and did not
262
include the potential subjective benefits of bowel preparations such as surgeon
263
perception of surgical difficulty or patient satisfaction. We would propose that surgical
264
difficulty would correlate with rates of bowel injury or obstruction and surgical infections,
265
variables which were accounted for in our model. Prior research has also found that
266
surgeons are poor predictors of which patients have had a bowel preparation.1 The use
267
of utility values is a method for incorporating patient satisfaction as quality of life is
268
measured and will decrease when events that impact patient satisfaction occur. Further
269
incorporating patient satisfaction would likely make our preferred strategy of no bowel
270
preparation a more favorable option as patients report higher satisfaction rates when no
271
bowel preparation is used.6 Finally, our model does not specifically discuss the insertion
272
of graft or mesh and the implications a bowel preparation has on potential negative
AC C
EP
TE D
261
ACCEPTED MANUSCRIPT 13
outcomes. Given the paucity of literature on the potential complications of mesh
274
placement at the time of a bowel injury, this was excluded and our model does not
275
adequately answer this question. Further research surrounding placement of graft is
276
needed.
RI PT
273
277
Our model adds to the current literature which questions the use of routine mechanical
279
bowel preparations prior to laparoscopic gynecologic surgery. Based on our findings,
280
surgeons can most likely omit preoperative bowel preparation. While surgeon behavior
281
is difficult to change, the use of a structured educational tool will likely ease the
282
transition.
M AN U
SC
278
283 284
288 289 290 291 292
EP
287
AC C
286
TE D
285
ACCEPTED MANUSCRIPT 14 293
References
294
1.
Muzzi L, Bellati F, Zullo MA, Manci N, Aangioli R, Panici PB. Mechanical bowel preparation before gynecologic laparoscopy: a randomized, single-blind,
296
controlled trial. Fertil Steril 2006;85:689-93.
297
2.
RI PT
295
Slim K, Vicaut E, Launay-Savary MV, Contant C, Chipponi J. Updated systematic review and meta-analysis of randomized clinical trials on the role of mechanical
299
bowel preparation before colorectal surgery. Annals of surgery 2009;249:203-9.
300
3.
SC
298
Yang LC, Aarden D, Lee TT, et al. Mechanical bowel preparation for gynecologic laparoscopy: a prospective randomized trial of oral sodium phosphate solution vs
302
single sodium phosphate enema. J Minim Invasive Gynecol 2011;18:149-56.
303
4.
M AN U
301
Wells T, Plante M, McAlpine JN. Preoperative bowel preparation in gynecologic oncology: a review of practice patterns and an impetus to change. International
305
journal of gynecological cancer : official journal of the International Gynecological
306
Cancer Society 2011;21:1135-42.
307
5.
TE D
304
Oliveira L, Wexner SD, Daneil N, et al. Mechanical bowel preparation for elective colorectal surgery. A prospective, randomized, surgeon-blinded trial comparing
309
sodium phosphate and polyethylene glycol-based oral lavage solutions. Diseases
310
of the colon and rectum 1997;40:585-91.
312 313
6.
AC C
311
EP
308
Ballard AC, Parker-Autry CY, Markland AD, Varner RE, Huisingh C, Richter HE.
Bowel preparation before vaginal prolapse surgery: a randomized controlled trial. Obstet Gynecol 2014;123:232-8.
ACCEPTED MANUSCRIPT 15 314
7.
Siedhoff MT, Clark LH, Hobbs KA, Findley AD, Moulder JK, Garrett JM.
315
Mechanical bowel preparation before laparoscopic hysterectomy: a randomized
316
controlled trial. Obstet Gynecol 2014;123:562-7. 8
Vradelis S, Kalaitakis E, Sharifi Y, et al. Addition of senna improves quality of
RI PT
317 318
colonoscopy preparation with magnesium citrate. World journal of
319
gastroenterology : WJG 2009;15:1759-63. 9.
Berkelhammer C, Ekambaram A, Silva RG. Low-volume oral colonoscopy bowel
SC
320
preparation: sodium phosphate and magnesium citrate. Gastrointestinal
322
endoscopy 2002;56:89-94.
323
10.
M AN U
321
Choi YS, Suh JP, Kim JK, et al. Magnesium citrate with a single dose of sodium
324
phosphate for colonoscopy bowel preparation. World journal of gastroenterology
325
: WJG 2011;17:242-8. 11.
Kao D, Lalor E, Sandha G, et al. A randomized controlled trial of four
TE D
326 327
precolonoscopy bowel cleansing regimens. Canadian journal of gastroenterology
328
= Journal canadien de gastroenterologie 2011;25:657-62. 12.
Jansen SV, Goedhard JG, Winkens B, VAN Deursen CT. Preparation before
EP
329
colonoscopy: a randomized controlled trial comparing different regimes.
331
European journal of gastroenterology & hepatology 2011;23:897-902.
332 333 334 335
13.
AC C
330
VAN Vugt VAN Pinxteren MW, VAN Kouwen MC, VAN Oijen
MG, VAN Achterberg T,
Nagengast FM. A prospective study of bowel preparation for colonoscopy with polyethylene glycol-electrolyte solution versus sodium phosphate in Lynch syndrome: a randomized trial. Familial cancer 2012;11:337-41.
ACCEPTED MANUSCRIPT 16 336
14.
McKenna T, Macgill A, Porat G, Friedenberg FK. Colonoscopy preparation: polyethylene glycol with Gatorade is as safe and efficacious as four liters of
338
polyethylene glycol with balanced electrolytes. Digestive diseases and sciences
339
2012;57:3098-105.
340
15.
RI PT
337
Hjelkrem M, Stengel J, Liu M, Jones DP, Harrison SA. MiraLAX is not as effective as GoLytely in bowel cleansing before screening colonoscopies. Clinical
342
gastroenterology and hepatology : the official clinical practice journal of the
343
American Gastroenterological Association 2011;9:326-32 e1. 16.
Forster JA, Thomas WM. Patient preferences and side effects experienced with
M AN U
344
SC
341
345
oral bowel preparations versus self-administered phosphate enema. Annals of
346
the Royal College of Surgeons of England 2003;85:185-6.
347
17.
Gold MR, Franks P, McCoy KI, Fryback DG. Toward consistency in cost-utility analyses: using national measures to create condition-specific values. Medical
349
care 1998;36:778-92.
350
18.
TE D
348
Tufts University Cost-Effectiveness Analysis Registry. Tufts Medical Center Institute for Clinical Research and Health Policy Studies. https://research.tufts-
352
nemc.org/cear4/. Accessed 1 December 2011. 19.
mechanical bowel preparation: a randomized prospective trial. Annals of surgery
354
2003;237:363-7.
355 356 357 358 359
Zmora O, Mahajna A, Bar-Zakai B, et al. Colon and rectal surgery without
AC C
353
EP
351
20.
Teng TO, Wallace A. One thousand health-related quality-of-life estimates.
Medical Care. 2000;38(6):583-637.
ACCEPTED MANUSCRIPT 17 360
21.
Won H, Maley P, Salim S, Rao A, Campbell NT, Abbott JA. Surgical and patient
361
outcomes using mechanical bowel preparation before laparoscopic gynecologic
362
surgery: a randomized controlled trial. Obstet Gynecol 2013;121:538-46. 22.
Miettinen RP, Laitinen ST, Makela JT, Paakkonen ME. Bowel preparation with
RI PT
363
oral polyethylene glycol electrolyte solution vs. no preparation in elective open
365
colorectal surgery: prospective, randomized study. Diseases of the colon and
366
rectum 2000;43:669-75; discussion 75-7.
367
23.
SC
364
Fa-Si-Oen P, Roumen R, Buitenweg J, et al. Mechanical bowel preparation or not? Outcome of a multicenter, randomized trial in elective open colon surgery.
369
Diseases of the colon and rectum 2005;48:1509-16.
370
24.
M AN U
368
Large MC, Kiriluk KJ, Decastro GJ, et al. The impact of mechanical bowel preparation on postoperative complications for patients undergoing cystectomy
372
and urinary diversion. The Journal of urology 2012;188:1801-5.
373
25.
TE D
371
Pena-Soria MJ, Mayol JM, Anula R, Arbeo-Escolar A, Fernandez-Represa JA. Single-blinded randomized trial of mechanical bowel preparation for colon
375
surgery with primary intraperitoneal anastomosis. Journal of gastrointestinal
376
surgery : official journal of the Society for Surgery of the Alimentary Tract
377
2008;12:2103-8; discussion 08-9.
379 380
26.
AC C
378
EP
374
Sugihara T, Yasunaga H, Horiguchi H, et al. Does mechanical bowel preparation
improve quality of laparoscopic nephrectomy? Propensity score-matched analysis in Japanese series. Urology 2013;81:74-9.
ACCEPTED MANUSCRIPT 18 381
27.
Ram E, Sherman Y, Weil R, Vishne T, Kravarusic D, Dreznik Z. Is mechanical
382
bowel preparation mandatory for elective colon surgery? A prospective
383
randomized study. Archives of surgery 2005;140:285-8. 28.
ChaleokittiI B. Comparative study between polyethylene glycol and sodium
RI PT
384 385
phosphate solution in elective colorectal surgery. Journal of the Medical
386
Association of Thailand = Chotmaihet thangphaet 2002;85:92-6. 29.
Itani KM, Wilson SE, Awad SS, Jensen EH, Finn TS, Abramson MA. Polyethylene
SC
387
glycol versus sodium phosphate mechanical bowel preparation in elective
389
colorectal surgery. American journal of surgery 2007;193:190-4.
390
30.
M AN U
388
Shepherd JP, Lowder JL, Jones KA, Smith KJ. Retropubic and transobturator
391
midurethral slings: a decision analysis to compare outcomes including efficacy
392
and complications. International urogynecology journal 2010;21:787-93. 31.
Walters SJ, Brazier JE. Comparison of the minimally important difference for two
TE D
393
health state utility measures: EQ-5D and SF-6D. Quality of life research : an
395
international journal of quality of life aspects of treatment, care and rehabilitation
396
2005;14:1523-32. 32.
before ileal urinary diversion: a systematic review and meta-analysis. Urologia
398
internationalis 2014;92:339-48.
399 400 401 402
Deng S, Dong Q, Wang J, Zhang P. The role of mechanical bowel preparation
AC C
397
EP
394
33.
Yamada T, Kan H, Matsumoto S, et al. Dysmotility by mechanical bowel
preparation using polyethylene glycol. The Journal of surgical research 2014;191:84-90.
ACCEPTED MANUSCRIPT 19 403
34.
the journey from evidence to effect. BMJ 2003;327:33-5.
404 405
Davis D, Evans M, Jadad A, et al. The case for knowledge translation: shortening
35.
Eskicioglu C, GagliardiI A, Fenech DS, Victor CJ, McLeod RS. Can a tailored knowledge translation strategy improve short term outcomes? A pilot study to
407
increase compliance with bowel preparation recommendations in general
408
surgery. Surgery 2011;150:68-74.
SC
409
RI PT
406
410
M AN U
411
AC C
EP
TE D
412
ACCEPTED MANUSCRIPT
20
RI PT
Table 1. Outcome probabilities and utility values
Outcome Probability (Range in Published Studies)
SC
No Outcome
Magnesium Sodium Mechanical
Enema
0.3055
Nausea/
0.1307
0.767 0.3067
(0.0313-
0.0000*
Diarrhea/ Fecal
EP
Preoperative
AC C
Anal Discomfort
0.3265#16 0.86)17,
0.6226)12-15 13
18
0.769
0.2857$
0.0052# 19
Incontinence
(0.68(0.0700-
0.4771) 9-
0.1688)8,9
Abdominal Pain
Event
(0.0214-
TE D
Vomiting/
0.0000*
Value
Phosphate Glycol(238g)
M AN U
Citrate Bowel Prep
Utility
Polyethylene
(0.610.2857#16
0.0749#3,8
0.3673#16
16,19
0.95)17, 18, 20
0.0214#9
0.0438#9
0.0469^
0.0499^
0.700#17
ACCEPTED MANUSCRIPT
21
Hypotension
0.0000*
0.0143#9
Arrhythmia/
0.0000 (0-
0.0438#9
0.0143$ 9
0.0000*
0.593#18
0.0000#9 Seizure
0.0063#9
21,22
0.0078)
RI PT
0.835 0.0145#22
0.0000*
(0.720.95)17
0.2012
SC
0.0832
0.0728 (0Surgical Site
(0.0667-
(0.0145-
Infection
0.12)3, 27,
M AN U
0.0236#26 19, 22-
0.1719)
0.3488)
0.0411#3
0.900#20
19, 22-
28
27
Intraoperative/ Abdominal
(0.0309-
0.1059
TE D
0.0622
Infection
0.32)22, 24, 26, 27
EP
Complications
0.696 0.2634
(0.0139-
0.0279&
(0.642(0.1812-
0.0685#3
0.1463)3,
0.75) 18, 0.3714)
22, 24
27
20
0.0636
0.0196
Ileus/
AC C
Postoperative
29
(0.0118-
Obstruction
0.650 (0.0139-
0.0285$
0.0374#19
0.0137#3
(0.6-
3,
0.0854)
0.0667)19,26,27
0.7)18 27
ACCEPTED MANUSCRIPT
22
0.00466 Bowel
0.0021 (00.0011$
(0.0396-
0.0000#3
0.550#30
RI PT
Injury/Colostomy 0.1467) 24, 26
0.0000#3
SC
0.0667) 24, 29
M AN U
Note: Outcome probabilities were derived from weighted averages of 28 published articles. Utilities were gathered from 3 articles and one publically available online database. Probabilities and utilities have a possible range of 0 to 1.
AC C
EP
TE D
#= Single value reported in the literature so a range is not reported *= Values for No mechanical bowel prep and enema were assumed to be 0 for certain preoperative outcomes given lack of mechanism to cause this event with no or minimal intervention. $= This value was not available in the literature so baseline value was assumed to be median of other bowel preparation values. This was accounted for in sensitivity analysis by varying over entire range of 0-1. ^= This value was not available in the literature so baseline value was calculated using the ratio of nausea/ vomiting/ abdominal pain to anal discomfort from the magnesium citrate and sodium phosphate preparations. The known value for nausea/ vomiting/ abdominal pain was then divided by this multiplier of 6.5412. This assumption was accounted for in sensitivity analysis by varying over entire range of 0-1. &= This value was not available in the literature so baseline value was calculated using the ratio of surgical site infection to abdominal infection from other four preparations. The known value for surgical site infection was then divided by this multiplier of 0.8472. This assumption was accounted for in sensitivity analysis by varying over entire range of 0-1.
ACCEPTED MANUSCRIPT 23
Figure 1. Simplified decision analysis tree
The tree modeling all possible complications is too complex to graphically illustrate
AC C
EP
TE D
M AN U
SC
RI PT
here.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
