Surg Clin N Am 85 (2005) 807–817

Long Limb Roux en Y Gastric

Bypass Revisited
Robert E. Brolin, MDa,b,*
a
University of Pittsburgh Medical Center, 200 Lothrop Street, Pittsburgh,
PA 15213-2582, USA
b
University Medical Center at Princeton, 253 Witherspoon Street, Princeton, NJ 08540, USA

In the early era of Roux en Y gastric bypass (RYGB), Roux limb length
typically was in the range of 50 cm to 75 cm and rarely exceeded 100 cm
[1–3]. In that era, Roux limb length was focused on elimination of bile
reflux. In our early experience with RYGB, when using a 50-cm to 75-cm
Roux limb, many of the heaviest patients failed to achieve satisfactory
weight loss postoperatively. This observation led us to design a prospective
randomized clinical study to learn whether a modest increase in Roux limb
length would improve weight loss results without producing a higher
incidence of metabolic and other complications. In our prospective
randomized study, a 150-cm Roux limb was compared with a conventional
75-cm Roux limb in 45 patients. The 150-cm measurement was chosen
arbitrarily and dubbed ‘‘long limb.’’ This modification is illustrated in
Fig. 1. Gastric pouch volume was the same in both groups. Because weight
loss results generally were satisfactory in less obese patients when using a 50-
to 75-cm limb RYGB, we restricted our protocol to super obese patients
who weighed at least 200 pounds more than their ideal body weight
according to standard life insurance tables [4]. We used the number of
pounds overweight in our selection of patients because body mass index
(BMI) was not a commonly used weight measurement in 1984 when the first
patient was entered into our prospective study. Extending Roux limb length
beyond 100 cm was not evaluated carefully before the publication of the
results of our prospective study in 1992 [5].

* NJBariatrics, 4250 U.S. Route 1 North, Suite 1, Monmouth Junction, NJ 08852.

E-mail address: rbrolin@njbariatricspc.com

0039-6109/05/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.suc.2005.03.003 surgical.theclinics.com
808 BROLIN

Fig. 1. Roux-en-Y gastric bypass in which the TA 90-B stapler (U.S. Surgical Corp., Norwalk,
Connecticut) is fired across the cardia of the stomach to create a 20 G 5 cm3 upper pouch. The
jejunum is divided approximately 30 cm distal to the ligament of Treitz with the distal end
anastomosed to the upper stomach using a circular stapler to create a 1.1-cm diameter
anastomosis. The proximal end of the jejunum is anastomosed 150 cm below the gastro-
jejunostomy. (From Brolin RE, Kenler HA, Gorman JG, et al. Long-limb gastric bypass in the
super-obese: A prospective randomized study. Ann Surg 1992;215:388; with permission.)

What is long limb gastric bypass?

The patients who had a 150-cm Roux limb achieved significantly greater
weight loss at 2 and 3 years postoperatively versus the group that had a
75-cm Roux limb (Fig. 2). Moreover, there was no difference in the
incidence of metabolic or other complications between the two groups.
These results prompted many bariatric surgeons to extend the length of the
Roux limb in their patients; however, many surgeons did not follow our
technique ‘‘to the letter’’ but rather used other measurements for the Roux
and biliopancreatic limb.
Only a few surgeons have reported their results with longer Roux limbs.
In 1991, Bruder et al [6] reported a series of 55 patients who had RYGB
performed with a 45-cm or 90-cm Roux limb. The patients were matched by
age and gender, but the comparison of limb lengths was not randomized.
Although limb length in the longer limb group was double that of the
patients who had short limbs, 90 cm is ‘‘short’’ by current standards. These
investigators reported that mean excess weight loss was 6% greater in the
longer Roux limb group after 6 months with no difference in diarrhea or
nutritional complications between the two groups. A follow-up study was
published by Freeman et al [7] in 1997. Roux limb length was extended to
 LONG LIMB ROUX EN Y GASTRIC BYPASS 809

Fig. 2. Weight loss in pounds through 4 years postoperatively. *Significant difference between
the two groups (P ! 0.02 by unpaired Student t test). (From Brolin RE, Kenler HA, Gorman
JG, et al. Long-limb gastric bypass in the super-obese: A prospective randomized study. Ann
Surg 1992;215:389; with permission.)

225 cm in a few patients, although the numbers of patients who had limb
measurements at 225 cm or any other specific length was not stated in that
report. Weight loss at 2 years postoperatively was substantially greater in
the patients who had longer Roux limbs. There was no difference in
metabolic sequelae between patients who had short and long limbs;
however, several patients who had long limbs had ‘‘troublesome’’ diarrhea.
In 1997, Sugerman et al [8] reported a series of 22 superobese patients who
failed to lose weight following conventional RYGB. They had revisional
operations that incorporated a 140-cm Roux limb, a long unmeasured
biliopancreatic limb, and a 150-cm common channel. This series followed
a group of 5 patients who underwent RYGB with unsatisfactory weight loss
who underwent a revisional RYGB that incorporated the same Roux limb
measurement and a 50-cm common channel. All 5 patients developed severe
protein calorie malnutrition with two late deaths secondary to hepatic failure.
Sugerman’s group concluded that biliopancreatic diversion that incorporated
a small upper gastric pouch and a 50-cm common channel was unduly severe
and could not be recommended. Conversely, the same procedure with a 150-
cm common channel resulted in a mean loss of excess weight of 69% at 5 years
postoperatively and manageable nutritional sequelae.
In 2000, MacLean et al [9] reported a significant difference in successful
weight loss outcome following isolated RYGB between 96 superobese
patients and 178 patients with a BMI of less than 50 kg/m2. Although the
superobese patients lost more weight than their less obese counterparts,
their final mean BMI was 35 G 7 kg/m2. Moreover, 41 of the 96
superobese patients (43%) failed to lose 50% of their excess weight. All
810 BROLIN

274 patients were randomized in a short (40 cm) versus long (100 cm)
Roux limb comparison [10]. MacLean et al also varied the length of the
biliopancreatic limb (10 cm in the short limb group versus 100 cm in the
long limb group). The superobese patients who had longer Roux and
biliopancreatic limbs had significantly greater weight loss. Conversely, in
the less obese patients (BMI ! 50 kg/m2) there was no difference in
weight loss between the two groups. There also was no difference in the
incidence of metabolic sequelae.
In 2001, Choban and Flancbaum [11] reported the results of a pro-
spective, randomized comparison of three Roux limb lengths in 133
patients, including a 75-cm versus 150-cm limb length in 69 patients with
a BMI of up to 50 kg/m2. Sixty-four superobese patients were randomized
to receive Roux limbs of 150 cm or 250 cm. Gastric pouch volume and
length of the biliopancreatic limb were the same in all patients. In the less
obese patients, the difference in limb length had no impact on postoperative
weight loss. Conversely, the superobese patients who had 250-cm Roux
limbs had significantly greater weight loss compared with the patients who
had 150-cm limbs at 18 months postoperatively. This difference seemed to
persist beyond 18 months but lost statistical significance as the number of
patients that was available for follow-up decreased over time. The relative
incidence of nutritional deficiencies was not addressed in this study.
In 2003, Feng et al [12] compared Roux limb lengths in 58 patients who
underwent laparoscopic RYGB. The short Roux limb group included 45
patients who had lengths that ranged from 45 cm to 100 cm, whereas the
limbs were 150 cm in remaining 13 patients. In this study, the difference in
weight loss between the short and long limb groups was not significant;
however, the number of patients that had long limbs was small and the
mean BMI for the entire series was 44 kg/m2. Nonetheless, a ‘‘trend toward
an increased portion of patients with O50% excess weight loss (P ¼ 0.07)
was observed in the extended Roux limb group.’’ This trend might have
achieved statistical significance if more patients who had long limbs were
included in this study.
In 2002, we reported the results of a 10-year evaluation of a ‘‘distal’’
RYGB in which the Roux enteroenterostomy was performed at 75 cm above
the ileocecal junction [13]. This distal RYGB incorporated a 15-cm to 25-cm
biliopancreatic limb and an upper gastric pouch with a capacity of up to 30
cm3. Forty-eight superobese patients with a mean BMI of 68 kg/m2, who
had the distal RYGB were compared retrospectively with superobese
patients who had ‘‘short’’ (50–75 cm) and ‘‘long’’ (150 cm) Roux limbs. Figs.
3 and 4 show weight loss expressed in pounds and BMI units through 5
years postoperatively. There were significant differences in weight loss
among the three groups that began at 6 months postoperatively and
persisted throughout the study. Greater weight loss was associated con-
sistently with progressively longer Roux limb lengths. The duration of
weight loss before stabilization also correlated with limb length. The short
 LONG LIMB ROUX EN Y GASTRIC BYPASS 811

Fig. 3. Weight loss in pounds through 5 years postoperatively. There were significant
differences between each of the three groups at at least 1 year postoperatively. *Significant
difference between the short limb group and distal RY gastric bypass (D-RY) and 150-cm
patients at 6 months postoperatievly (P ! 0.05 by ANOVA with Student-Newman-Keuls test).
(From Brolin RE, Lamarca LB, Kenler HA, et al. Malabsorptive gastric bypass in patients with
super obesity. J Gastrointest Surg 2002;6:198; with permission.)

limb group stabilized between 12 and 18 months, whereas the patients who
had 150-cm limbs or distal RYGB stabilized at 24 and 36 months,
respectively. The percentage of weight that was regained from the mean time
of stabilization was the same in patients who had short limbs and 150-cm
limbs. Less weight was regained following distal malabsorptive RYGB; this

Fig. 4. Change in BMI through 5 years postoperatively. There were significant differences
between each of the groups at 12, 24, and 36 months postoperatively. *Significant difference
between the Dry patients and the shorter limb groups noted at 6, 18, 48, and 60 months
(P ! 0.05; by ANOVA with Student-Newman-Keuls test). (From Brolin RE, Lamarca LB,
Kenler HA, et al. Malabsorptive gastric bypass in patients with super obesity. J Gastrointest
Surg 2002;6:199; with permission.)
812 BROLIN

suggests that the malabsorptive component may contribute to long-term

weight maintenance. Variability in eating behavior among individual
patients is likely the most important factor in weight loss maintenance
following bariatric operations. The incidence of metabolic problems was
significantly greater following distal RYGB (Table 1).
In 1998, the Mayo Clinic group reported their experience with a very very
long Roux limb in 26 superobese patients; the mean BMI was 67 G 3 kg/m2
[14]. Roux limb length was unmeasured; however, the common channel
between the enteroenterostomy and the ileocecal junction was constant at
100 cm. At 2 years, the mean percentage of excess weight loss was 57%,
which was less than the 64% mean excess weight loss that was observed in
our patients who underwent a distal RYGB. The incidence of postoperative
nutritional problems was not mentioned in the Mayo report.
In the early era of bariatric surgery there was only one current procedural
terminology (CPT) code for RYGB. In 1994, a second CPT code for
RYGB, #43847, was approved. This coding change distinguished short- and
long-limb RYGB. CPT #43846 is described as ‘‘gastric restrictive procedure,
with gastric bypass for morbid obesity; with short limb (less than 100 cm)
Roux-en-Y gastroenterostomy.’’ CPT #43847 is described as ‘‘gastric
restrictive procedure, with gastric bypass for morbid obesity with small
intestine reconstruction to limit absorption’’ [15]. Although a limb
measurement is not listed for #43847, it is de facto that any limb
measurement that exceeds 100 cm must be assigned this code because there
are only two codes listed for RYGB!

Determination of experimental and investigative procedures

Several major insurance carriers deny coverage for any RYGB that is
coded under #43847 on the basis that all procedures that are submitted
under this code are ‘‘experimental or investigative.’’ One major carrier’s
definition of ‘‘experimental/investigative’’ is an ‘‘intervention’’ that is ‘‘not
proven to be as safe or effective in achieving an outcome equal to or

Table 1
Postoperative metabolic deficiencies
Operation Iron B-12 Anemia Vit A Vit D Calcium Albumin
a a
Short (N ¼ 80) 42 (52%) 30 (37%) 33 (41%) – – –a –a
150 cm (N ¼ 102) 46 (45%) 34 (33%) 36 (35%) –a –a –a –a
b b
D-RY (N ¼ 39) 19 (49%) 3 (8%) 36 (74%) 4 (10%) 20 (51%) 4 (10%) 5 (13%)
a
Not measured.
b
Significant difference between D-RY patients and the other two groups. (Less than 0.003
by chi square test).
From Brolin RE, Lamarca LB, Kenler HA, et al. Malabsorptive gastric bypass in patients
with super obesity. J Gastrointest Surg 2002;6:201; with permission.
 LONG LIMB ROUX EN Y GASTRIC BYPASS 813

exceeding the outcome of alternative therapies,’’ or ‘‘does not improve

health outcomes,’’ or ‘‘does not permit conclusions concerning the effect of
the interventions on health outcomes,’’ or ‘‘is not proven to be applicable
outside the research setting.’’ The author recently submitted a request for
approval of RYGB, CPT #43847 to be performed in a middle-aged woman
(BMI ¼ 63 kg/m2) who had refractory hypertension, type II diabetes, sleep
apnea, and degenerative arthritis. The carrier’s denial included their criteria
for ‘‘experimental/investigative’’ procedures as detailed above. The author
appealed the denial and cited our 1992 prospective randomized study and
MacLean et al’s [10] 2001 report as evidence to refute the criteria in their
denial letter. Not surprisingly, a second denial was received based upon their
previously cited criteria without comment on the data in our 1992 paper or
MacLean et al’s report. A subsequent conversation with one of the carrier’s
medical directors was nonproductive. The medical director cited a recent
‘‘in-house’’ review of available bariatric surgical procedures that totally
ignored our 1992 study and our more recent publication with longer follow-
up [13]. Rather than continue with an apparently losing battle, this patient
consented to have a less than 100-cm Roux limb gastric bypass which
proceeded uneventfully.
It seems likely that the deepest root of the authorization problem for
long-limb RYGB is the current descriptions of the two existing CPT codes.
CPT #43846 and #43847 were approved for use more than 10 years ago. All
of the prospective studies that confirmed the safety and efficacy of the long-
limb RYGB that was described in our 1992 report were published during the
past 5 years [10–12]. Unfortunately, CPT #43847 has lumped the 150-cm
RYGB with the very very long limb/distal RYGB, which were introduced as
malabsorptive procedures. All of the available data on the 150-cm RYGB
show it to be as safe as an up to 100-cm RYGB in terms of complications,
and more effective in producing weight loss. Conversely, the distal RYGB
with a very (very) long Roux limb and an up to 150-cm common channel
seems to produce better weight loss than the 150-cm RYGB, but at a defined
metabolic price in terms of a greater variety and increased incidence of
nutritional sequelae [8,13]. Although some degree of biliopancreatic
diversion is inherent in RYGB, the degree of diversion that is likely to
result in fat soluble vitamin deficiency and steatorrhea remains undeter-
mined. There are sufficient data to justify the position that RYGB with
a 150-cm Roux limb does not result in more metabolic sequelae than
conventional short-limb RYGB [10,13].
Another factor that must be considered in the authorization process for
bariatric operations is the motivation of individual insurance carriers.
Fischer [16] recently recounted his experience as a member of the medical
advisory panel (MAP) for the Technology Evaluation Center (TEC) which
represents a ‘‘joint effort between the Kaiser Permanente Foundation and
the Blue Cross and Blue Shield Association.’’ Fischer commented that the
TEC panel had difficulty believing that bariatric surgery improved ‘‘net
814 BROLIN

health outcome.’’ Moreover, he recounted that patients ‘‘often’’ regain their

weight ‘‘sometimes needing a second procedure if they are not followed
rigorously in a program with long term nutritional counseling.’’ He later
commented that no American study showed that gastric bypass resulted in
improvement in ‘‘net health outcome.’’ This conclusion flies in the face of
hundreds of papers that showed dramatic improvement in all of the
comorbidities that are associated with severe obesity following successful
weight loss surgery. Fischer did not define ‘‘net health outcome’’ in his
report. An important note in Fischer’s commentary was that ‘‘the staff’’ of
the insurance carriers collected all of the papers for review by the MAP and
already had ‘‘whittled down’’ the collection to ‘‘prospective studies’’ by the
time of presentation to the MAP. Our 1992 prospective randomized study
apparently was not reviewed by the MAP. As a cynic regarding the moti-
vations of health insurance carriers, the author seriously questions how or
why the staff could not find at least one American study that showed
improvement in ‘‘net health outcome’’ following bariatric surgery.
The TEC report prompted Horizon Blue Cross Blue Shield of New Jersey
to deny CPT code #43847, after covering this code since its inception. In their
justification for denial, long-limb gastric bypass is described as a ‘‘malab-
sorptive’’ procedure which has ‘‘the potential complications similar to those of
the biliopancreatic bypass.’’ Coding problems notwithstanding, it seems that
denial is the primary agenda here.
In 2005 several new CPT codes for bariatric operations were approved
for general use including separate codes for laparoscopic and open
procedures. Among the new codes is a change in the descriptor for CPT
#43846 to acknowledge an up to 150-cm Roux limb as ‘‘short.’’ Justification
for this change is the absence of data which show that a 150-cm limb
produces more malabsorptive manifestations than shorter Roux limb
measurements.

Very (very) long Roux limbs and biliopancreatic diversion

Cynicism aside, none of the malabsorptive modifications of RYGB have
been subjected to a randomized, prospective comparison with conventional
short-limb RYGB, CPT #43846; however, the outcome of such a study
seems predictable based upon the available data. Weight loss likely would be
significantly greater with the more malabsorptive modifications of RYGB.
There also is likely to be a significantly greater incidence of several
metabolic deficiencies with more malabsorption (see Table 1). Improve-
ments in obesity-related comorbidities would likely be greater with more
weight loss, although short-term differences in comorbidity improvement
between the two procedures may not be dramatic. Long-term weight
maintenance might be better with more malabsorption, which could result in
more sustained improvement of comorbidities. A comparison of revision
rates also would be relevant because a substantial number of superobese
 LONG LIMB ROUX EN Y GASTRIC BYPASS 815

patients that have conventional RYGB require revision for inadequate

weight loss. Conversely, ongoing nutritional problems might be the leading
cause for revision after malabsorptive RYGB. Only 1.6% of the author’s
more than 200 patients who underwent a distal RYGB have required
revision for refractory nutritional deficiencies, although 3.8% of these
patients temporarily required total parenteral nutrition in that regard.
Moreover, the author has seen a substantial number of superb weight loss
results in extremely heavy (BMI O 70 kg/m2) patients after distal RYGB.
Patients in this weight class frequently experience poor weight loss after
short-limb RYGB [9,13].
In summary, the risk/benefit analysis of distal/malabsorptive RYGB is
complex. A well-controlled prospective comparative study of short limb
versus distal/malabsorptive RYGB would require years to complete. It
would be extremely helpful for the insurance carriers to enjoin bariatric
surgeons in conduct of such a study. It seems that the National Institute of
Health has no interest in supporting such a study because they dropped the
surgical arm of the Study of Health Outcomes of Weight-loss Trial (and
changed its name) in 2000.

Revisional bariatric operations

The long-term ‘‘success rate’’ of conventional short-limb RYGB ranges
between 50% and 80%, which implies ‘‘failure’’ for the remaining patients
[3,9,13,17]. Weight loss ‘‘failure’’ is significantly more prevalent in superobese
patients (BMI R 50 kg/m2). Fortunately, surgical options are available for
patients who fail to achieve satisfactory loss after short-limb RYGB. Adding
restriction alone to a failed RYGB rarely results in substantial weight loss
[18–20]. Conversely, adding malabsorption by lengthening the Roux limb has
provided good long-term weight loss for many of these challenging patients. In
Sugerman et al’s [8] report, the 22 patients who failed conventional RYGB lost
an additional 30% of their excess weight with improvement or resolution of
the comorbidities that reoccurred with cessation of weight loss following
conventional RYGB. Likewise, Fobi et al [21] reported substantial additional
weight loss after conversion of patients who failed conventional banded
RYGB to a more malabsorptive procedure, again at the expense of more
metabolic complications. The author has performed 29 distal RYGBs with up
to a 75-cm common channel in patients who had unsatisfactory weight loss
after conventional RYGB, including 10 who had a 150-cm Roux limb. Mean
weight loss was 32.5 kg at 1 year postconversion. Twenty of the 29 patients
(72.4%) lost at least 50% of their excess weight. Metabolic sequelae and
diarrhea accompanied these revisions. One patient required takedown of her
RYGB because of poor weight loss in conjunction with severe hypocalcemia
and hypoproteinemia. Noncompliant, unreliable patients with poor weight
loss after conventional RYGB should not be offered revisional surgery;
816 BROLIN

however, preoperative identification of these noncompliant patients is

problematic.

Summary
RYGB that is performed with at least a 150-cm Roux limb results in
significantly greater weight loss than shorter (!100-cm) Roux limb
procedures in superobese patients (BMI R50 kg/m2). Conversely, longer
Roux limb procedures do not provide greater weight loss in less obese (BMI
! 50 kg/m2) patients. Modest elongation of the Roux limbdin the range of
150 cm to 200 cmddoes not result in more frequent nutritional sequelae
compared with shorter Roux limb procedures. Conversely, RYGBs, in
which the Roux or the biliopancreatic limb is very long with anastomosis to
the mid or distal ileum (very, very long), usually results in more metabolic
problems than RYGBs in which the Roux limb measures up to 150 cm and
the biliopancreatic limb is short. The current (2005) CPT codes do not
stratify Roux limb length adequately on the basis of weight loss outcome or
late nutritional sequelae.

References
[1] Lechner GW, Callender K. Subtotal gastric exclusion and gastric partitioning: a randomized
prospective comparison of one hundred patients. Surgery 1981;90:637–44.
[2] Pories WJ, Flickinger EG, Meelheim D, et al. The effectiveness of gastric bypass over gastric
partition in morbid obesity. Consequence of distal gastric and duodenal exclusion. Ann Surg
1982;196:389–99.
[3] Yale CE. Gastric surgery for morbid obesity: Complications and long term weight control.
Arch Surg 1989;124:941–7.
[4] Measurement of overweight. Stat Bull NY Metropol Life Insur Co 1984;54:20–3.
[5] Brolin RE, Kenler HA, Gorman JG, et al. Long-limb gastric bypass in the superobese:
a prospective randomized study. Ann Surg 1992;215:387–95.
[6] Bruder SJ, Freeman JB, Brazeau-Gravelle P. Lengthening the Roux-Y limb increases weight
loss after gastric bypass: a preliminary report. Obes Surg 1991;7:414–9.
[7] Freeman JB, Kotlarewsky M, Phoenix C. Weight loss after extended gastric bypass. Obes
Surg 1997;7:337–44.
[8] Sugerman JH, Kellum JM, DeMaria EJ. Conversion of proximal to distal gastric bypass for
failed gastric bypass for superobesity. J Gastrointest Surg 1997;1:517–25.
[9] MacLean LD, Rhode BM, Nohr CW. Late outcome of isolated gastric bypass. Ann Surg
2000;231:524–8.
[10] MacLean LD, Rhode BM, Nohr CW. Long or short-limb gastric bypass? J Gastrointest
Surg 2001;5:525–30.
[11] Choban PS, Flancbaum LJ. The effect of Roux limb lengths on outcome after Roux-en-Y
gastric bypass: a prospective randomized clinical trial. Obes Surg 2002;12:540–5.
[12] Feng JJ, Gagner M, Pomp A, et al. Effect of standard vs. extended Roux limb length on
weight loss outcomes after laparoscopic Roux-en-Y gastric bypass. Surg Endosc 2003;17:
1055–60.
 LONG LIMB ROUX EN Y GASTRIC BYPASS 817

[13] Brolin RE, Lamarca LB, Kenler HA, et al. Malabsorptive gastric bypass in patients with
superobesity. J Gastrointest Surg 2002;6:195–205.
[14] Murr MM, Balsiger BM, Kennedy FP, et al. Malabsorptive procedures for severe obesity:
comparison of pancreaticobiliary bypass and very, very long Roux-en-Y gastric bypass.
J Gastrointest Surg 1998;3:607–12.
[15] Current Procedure Terminology CPT 2004. Chicago: AMA Press; 2004.
[16] Fischer JE. Serving on the MAP of the Blue Cross and Blue Shield Association’s TEC. Bull
Am Coll Surg 2004;89:22–5.
[17] Halverson JD, Zuckerman GR, Koehler RE, et al. Gastric bypass for morbid obesity:
a medical-surgical assessment. Ann Surg 1981;194:152–60.
[18] Sugerman HJ, Wolper JL. Failed gastroplasty for morbid obesity. Am J Surg 1984;148:
331–6.
[19] Behrns KE, Smith CD, Kelly KA, et al. Reoperative bariatric surgery: lessons learned to
improve patient selection and results. Ann Surg 1993;218:646–53.
[20] Naslund I. The size of the gastric outlet and the outcome of surgery for obesity. Acta Chir
Scand 1986;152:205–10.
[21] Fobi M, Lee H, Igwe D Jr, et al. Revision of failed gastric bypass to distal Roux-en-Y gastric
bypass; a review of 65 cases. Obes Surg 2001;11:190–5.