Comparison of Clinical and Radiographic

Outcomes of Vertical Simple Stitch

Versus Modified Mason-Allen Stitch
in Arthroscopic Bankart Repairs
A Prospective Randomized Controlled Study
Jin-Young Park,* MD, PhD, Seok Won Chung,yz MD, PhD, Jong Soo Lee,* MD,
Kyung-Soo Oh,y MD, PhD, and Jae-Hyung Lee,* MD
Investigation performed at Neon Orthopaedic Clinic, Seoul, Republic of Korea,
and Konkuk University School of Medicine, Seoul, Republic of Korea

Background: In spite of the probable advantages of Bankart repair with modified Mason-Allen technique, there has been no
study to evaluate the clinical outcomes of the modified Mason-Allen technique for Bankart repair.
Purpose: To prospectively compare the anatomic and clinical outcomes between the vertical simple stitch and the modified
Mason-Allen stitch with respect to the labral height, retear rate, redislocation/apprehension, and various functional outcome
scores.
Study Design: Randomized controlled trial; Level of evidence, 2.
Methods: Patients who underwent arthroscopic Bankart repair with double-loaded suture anchors were randomly allocated to 1
of 2 groups: the vertical simple stitch method (SS; n = 45) or the Bankart repair using modified Mason-Allen technique (BRUMA;
n = 41). All patients underwent computed tomography arthrography at 6 months postoperatively and various functional outcome
measurements at least 2 years postoperatively. The labral height and width at the 3-, 4-, and 5-o’clock positions were measured
preoperatively and at 6 months after surgery on axial computed tomography arthrography; the redislocation/apprehension rate,
the retear rate, and various functional outcome scores were evaluated at each follow-up visit.
Results: Postoperative labral height and width were significantly increased at all locations (all P \ .001) in both groups, but they
were not statistically different between groups (all P . .05). Two patients in the SS group (4.4%) and 2 in the BRUMA group (4.9%)
experienced redislocation after surgery, and 4 patients in the SS group (8.9%) and 2 in the BRUMA group (4.9%) group showed
apprehension after surgery. Additionally, 5 patients in the SS group (11.1%) and 2 in the BRUMA group (4.9%) showed retear at 6
months (P = .239). There were no differences in any functional outcome scores (all P . .05).
Conclusion: There was no difference in the radiologic outcomes at 6 months and the clinical outcomes assessed at least 2 years
after surgery between the groups.
Keywords: modified Mason-Allen stitch; simple stitch; Bankart repair; labral height; outcomes

Anterior shoulder instability is a disabling pathology, espe-
cially in the young active population. In most cases, a Bank-
art lesion occurs concomitantly. A Bankart lesion is a tear of
the capsulolabral tissue from the anteroinferior glenoid that
results in decreased labral height. It is considered an essen-
tial lesion for anterior shoulder instability with the laxity of
the capsuloligamentous complex.2 Many procedures have
The American Journal of Sports Medicine
1–10
DOI: 10.1177/0363546518816679
Ó 2018 The Author(s)
been introduced to address this Bankart lesion, and
arthroscopic Bankart repair with suture anchors has
recently become the preferred surgical method for shoul-
der stabilization in most cases of anterior shoulder
instability.4,18,29
The outcomes of arthroscopic Bankart repair with
suture anchors are generally acceptable. However, retear
or redislocation after arthroscopic Bankart repair remains
problematic, reaching an incidence of 34%.6,29 The failure
of arthroscopic Bankart repair with suture anchors often
occurs when the quality of the capsulolabral tissue is
poor as in cases of chronic recurrent instability, when the
bone loss of the anterior glenoid is substantial, and when

1
2 Park et al The American Journal of Sports Medicine

the labral hump is not adequately restored with sufficient
tension of the inferior glenohumeral ligament.1,13,24,30
Various repair techniques have been described to
enhance the outcomes of the Bankart repair with suture
anchors, such as simple stitch, horizontal mattress suture,
knotless suture, double-row repair, and modified Mason-
Allen technique.3,5,21,23 Among these, the Bankart repair
using modified Mason-Allen technique (BRUMA), as first
described by Castagna et al,5 is a combination of horizontal
mattress suture and simple stitch. The BRUMA can
improve a tissue’s holding property and create a rip stop
to reduce the risk of cutting through capsulolabral tissue,
by tying the horizontal mattress suture before the vertical
stitch with a double-loaded suture anchor.5 Thus, the mod-
ified Mason-Allen suture may be a good option to strongly
repair the Bankart lesion and restore capsulolabral ten-
sion, especially in poorly attenuated capsulolabral tissue,
where tensioning by simple stitch may not be enough to
avoid failure.5 Furthermore, as opposed to a vertical simple
stitch technique, the horizontal mattress stitch of the mod-
ified Mason-Allen suture was shown to better restore lab-
ral height,3 which may help to restore the capsulolabral
buttress and concavity/compression functions of the
labrum, thus resulting in increased glenohumeral stability
and better functional outcomes.16,19
In spite of these probable advantages of the BRUMA,
there has been no study to evaluate the outcomes of this
technique. Therefore, the purpose of this study was to pro-
spectively compare the anatomic and clinical outcomes
between the vertical simple stitch and the modified
Mason-Allen stitch with respect to labral height, retear
rate, redislocation/apprehension, and various functional
outcome scores. We hypothesized that the Mason-Allen
stitch would increase labral height, reduce the rate of retear
or redislocation/apprehension, and improve functional out-
comes when compared with the vertical simple stitch.
METHODS
Sample Size Calculation and Patient Allocation
This was a prospective randomized controlled study. We
conducted this study in accordance with the principles of
the Declaration of Helsinki. The reporting of data from
this trial complies with the CONSORT statement (Consol-
idated Standards of Reporting Trials). Sample sizes are
calculated to detect a significant difference in Rowe scores
(mean difference, 6 points; SD, 8 points; effect size, 0.75)
and in labral height (mean difference, 1.1 mm; SD,
1.4 mm; effect size, 0.79), according to a previous study.13
A sample size of 47 patients in each group was required
for a power of 90% at a type I error level of .05, with an
expected dropout rate of 20%. Initially, 184 consecutive
patients were prospectively enrolled who underwent
arthroscopic Bankart repair for anterior shoulder instabil-
ity at Neon Orthopaedic Clinic and Konkuk University
Hospital between June 2013 and December 2015. Patients
with anterior shoulder instability were considered surgical
candidates for arthroscopic Bankart repair if they had per-
sistent restriction in performing daily or sports activities
and had not responded to at least 6 months of adequate
nonoperative treatment.14,15 Patients with a single episode
of shoulder dislocation could be surgical candidates if their
functional restrictions were severe in spite of an adequate
nonoperative treatment and if they were high-risk patients
involved in contact or collision sports near the end of the
season. Patients were excluded with evidence of rotator
cuff lesion (n = 24), multidirectional instability (n = 6),
engaging Hill-Sachs lesion (n = 15), bony Bankart lesion
(n = 25), combined reverse Bankart lesion (n = 8), and revi-
sion Bankart repair surgery (n = 12). If a humeral head
was engaged at the glenoid during application of maximal
external rotation, forward flexion, and horizontal exten-
sion force at the time of surgery, it was regarded as an
engaging Hill-Sachs lesion requiring remplissage proce-
dure and was excluded from this study. All erosion and
fragment types were not defined as bony Bankart lesions
and were thus included in this study. No patient refused
to participate in the study, and no patient showed degener-
ative arthritis of the glenohumeral joint. The remaining 94
patients were randomly allocated into the vertical simple
stitch method (SS) group or the BRUMA group (47 patients
in each group). Patients were randomized with a computer-
generated block randomization sequence (http://www.
randomizer.org) by an independent researcher, and the
group assignment was disclosed to the physician at the
time of surgery. Among these 94 patients, 2 in the SS
group and 6 in the BRUMA group were lost before 2-year
postoperative follow-up and were excluded. Finally, 45
patients in the SS group and 41 in the BRUMA group
were available for this final analysis (Figure 1).
For the anatomic outcome evaluation, computed tomog-
raphy (CT) arthrogram was performed preoperatively and
at postoperative 6 months. In addition, various functional
outcomes were evaluated for every patient, including the
Rowe score, American Shoulder and Elbow Surgeons
(ASES) score, Constant score, and visual analog scale
(VAS) for pain at 6 months, 1 year, and final follow-up
(at least 2 years after surgery). The mean 6 SD patient
age at the time of surgery was 26.6 6 9.6 years (range,
16-56 years) in the SS group and 23.5 6 6.4 years (range,

z
Address correspondence to Seok Won Chung, MD, PhD, Department of Orthopaedic Surgery, Konkuk University School of Medicine, 120-1
Neungdong-ro (Hwayang-dong), Gwangjin-gu, Seoul 143-729, Republic of Korea (email: smilecsw@gmail.com).
*Center for Shoulder, Elbow and Sports Medicine, Neon Orthopaedic Clinic, Seoul, Republic of Korea.
y
Department of Orthopaedic Surgery, School of Medicine, Konkuk University, Seoul, Republic of Korea.
One or more of the authors has declared the following potential conflict of interest or source of funding: This research was supported by the Basic Sci-
ence Research Program through the National Research Foundation of Korea, funded by the Ministry of Science and ICT (NRF-2017R1A2B4003343).
AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD
and disclaims any liability or responsibility relating thereto.
AJSM Vol. XX, No. X, XXXX
Figure 1. Flowchart.
14-45 years) in the BRUMA group, with 41 men and 4
women in the SS group and 34 men and 7 women in the
BRUMA group. The mean follow-up period was 27.9 months
(range, 24-46 months) and 28.2 months (range, 24-40
months) in the SS and BRUMA groups, respectively. Data
regarding factors that can affect the outcome of arthroscopic
Bankart repair and capsular shift were prospectively col-
lected for all patients and evaluated. Level of sports activity
was recorded as high, medium, or low.8 A high level of sports
activity was defined as participating in dynamic or contact
sports (eg, boxing, wrestling, ice hockey, skiing, basketball,
baseball, football, and badminton); a medium level, as partici-
pating in static sports (eg, golf, yoga, cycling, and running);
and a low level, as rarely participating in sports. Thirteen
patients in the SS group and 9 patients in the BRUMA group
experienced a single episode of shoulder dislocation without
recurrent dislocation. There were 12 athletes in the SS group
(8 amateur and 4 elite/professional) and 17 athletes in the
BRUMA group (10 amateur and 7 elite/professional). Sports
included baseball (n = 2, SS group; n = 8, BRUMA group),
handball (n = 1, SS group; n = 3, BRUMA group), soccer
(n = 2, BRUMA group), golf (n = 2, BRUMA group), martial
arts (eg, Muay Thai, wrestling, judo; n = 2, SS group; 1,
BRUMA group), short track speed skating (n = 1, SS group),
bicycle racing (n = 1, SS group), rowing (n = 1, SS group),
snowboarding (n = 1, SS group), diving (n = 1, SS group),
weight lifting (n = 1, BRUMA group), ice hockey (n = 1,
BRUMA group), and rugby (n = 1, BRUMA group).
Hill-Sachs lesion size was measured as the percentage
of the articular cartilage involved and classified as small
(\20% of involvement), medium (20%-30%), and large
Modified M-A Stitch in Bankart Repair 3
(.30%).7 Glenoid bone loss was measured according to
the method of d’Elia et al.9 In brief, by drawing a circle
based on the inferior part of the glenoid on a sagittal image
cut and outlining the contour of the glenoid bone defect,
the percentage of glenoid bone loss was calculated (patho-
logic glenoid area divided by the area of circle). Twenty-five
patients showed definite glenoid bone loss .10% and were
excluded from this study. However, glenoid bone loss
\10% (eg, erosion or fragment type) was included in this
study. Bankart lesion tear size was measured as the differ-
ence of the most superior and inferior clock positions.
Demographic and clinical data in both groups are summa-
rized in Table 1.
Surgical Procedure
All surgical procedures were performed by a single surgeon
(J-Y.P.) with the patient in the lateral decubitus position
and the arm in a traction device composed of 3 directions
(3-point shoulder distraction system; Arthrex) at 7 kg of
traction. We evaluated anterior translation under anesthe-
sia, and all patients showed grade .2 anterior translation.
The posterior, anterosuperior, and anteroinferior portals
were established according to the procedure of Wolf.31
The posterior portal was established approximately 2 cm
inferior and 1 cm medial to the acromial angle, and the
anterosuperior portal was created in the rotator interval
just inferior to the anterior edge of the acromion. In addi-
tion, the anteroinferior portal was made just over the supe-
rior border of the subscapularis tendon with an outside-in
4 Park et al The American Journal of Sports Medicine

TABLE 1
Patient Demographics and Clinical Dataa

Group, Mean 6 SD (Range) or n

Variable SS (n = 45) BRUMA (n = 41) P Value

Age, y 26.6 6 9.6 (16-56) 23.5 6 6.4 (14-45) .064

Sex .256
Male 41 34
Female 4 7
Interval from first episode to surgery, mo 36.9 6 37.6 36.9 6 37.6 .712
Follow-up, mo 27.9 6 6.7 (24-46) 28.2 6 6.5 (24-40) .850
Side of involvement .309
Dominant 29 22
Nondominant 16 19
Dislocations before surgery 4.8 6 6.9 4.3 6 6.3 .681
Initial dislocation event .631
Sports associated 29 30
Other traumatic 16 11
Sports activity .276
Low 2 5
Medium 8 4
High 35 32
Athlete .147
Yes 12 17
No 33 24
ALPSA lesion .039
Yes 6 14
No 39 27
Hill-Sachs defect .624
Yes 38 34
No 7 7
Hill-Sachs defect sizeb .195
Small 31 25
Medium 4 8
Large 3 1
SLAP lesion .800
Yes 11 11
No 34 30
Tear size, o’clockc 3.9 6 1.1 4.0 6 1.1 .731
Anchors used 3.7 6 0.5 3.9 6 0.5 .117
Preoperative
FF 174.55 6 11.01 169.87 6 12.52 .069
ER, at side 60.11 6 20.35 58.29 6 17.16 .657
ER, 90° of abduction 72.44 6 12.46 69.39 6 12.15 .254
IR, at back 8.22 6 2.29 8.31 6 2.45 .854
VAS for pain 3.4 6 0.6 3.0 6 1.0 .142
ASES score 62.6 6 7.9 64.0 6 8.4 .424
Constant score 59.4 6 5.9 60.4 6 7.4 .465

a
ALPSA, anterior labral periosteal sleeve avulsion; ASES, American Shoulder and Elbow Surgeons; BRUMA, Bankart repair using mod-
ified Mason-Allen technique; ER, external rotation; FF, forward flexion; IR, internal rotation; SLAP, superior labral anterior and posterior;
SS group, vertical simple stitch method; VAS, visual analog scale.
b
Hill-Sachs lesion was classified according to the percentage of the articular cartilage involved: small, \20% of involvement; medium,
20%-30%; large, .30%.7
c
Tear size was measured as the difference of the most superior and most inferior clock positions.

technique and a spinal needle to determine the most appro-
priate position of the cannula for anchor placement. From
the perspective of the anteroinferior portal, the anterior
capsulolabral tissue was released and mobilized from the
anterior glenoid surface with a soft tissue liberator (Linva-
tec). Full mobilization was carefully checked by pulling the
capsulolabral tissue up to the glenoid level. Soft tissue on
the medial wall of the glenoid was removed, and decortica-
tion of the anterior glenoid rim was performed with
a motorized shaver, bur, and rasp to increase healing
potency. The first anchor hole was created on the anterior
glenoid rim at approximately the 5-o’clock position for the
right shoulder (7 o’clock for the left shoulder). For simplic-
ity, throughout the rest of the article, the analyses and
AJSM Vol. XX, No. X, XXXX
Figure 2. Vertical simple stitch method for the Bankart
repair. (A) Pilot hole drilling on the face of the glenoid 1 to
2 mm from the glenoid rim. (B) Insertion of a double-loaded
suture anchor into the hole. (C) Passage of 2 suture strands
from 2 different suture threads. (D) Bankart repair with a ver-
tical simple stitch method.
descriptions of the clock positions are given as if all involved
shoulders were right shoulders. After pilot hole drilling on
the face of the glenoid 1 to 2 mm from the glenoid rim, a dou-
ble-loaded suture anchor (Gryphon T BR Suture Anchor;
DePuy Synthes Inc) was inserted into the hole through
the anteroinferior portal, with a couple of sutures aligned
in the direction of the anterior inferior ligament.
For the SS, with use of a spectrum hook (Conmed-
Linvatec Inc) loaded with No. 0 PDS (Maxon o 3.5 metric;
Covidien), a capsulolabral suture was created approxi-
mately 1 cm lateral to and inferior from the anchor hole to
obtain appropriate superior and medial capsular shift to
the glenoid. The PDS was retrieved through the anterosupe-
rior portal. One end of the suture was engaged into the knot
of the PDS and then pulled back out of the anteroinferior
portal (shuttle relay method). Finally, an arthroscopic slid-
ing knot (SMC knot) was made at the capsular side after
the capsule was sufficiently shifted. For the other suture
thread loaded in the same suture anchor, this capsulolabral
suture procedure of an SS was repeated at approximately
5 mm next to the first suture (Figure 2).
For the modified Mason-Allen method, one end of the
first suture thread was passed through the similar portion
of the capsulolabral complex as the SS; the other end of the
same suture thread was then passed through approxi-
mately 5 mm next to the first passage (second passage),
with the same shuttle relay method as the SS. Subse-
quently, one end of the second suture thread of the same
anchor was passed just medial to the previous 2 suture
strands of the first suture thread (third passage). After
Modified M-A Stitch in Bankart Repair 5
Figure 3. Bankart repair using modified Mason-Allen method.
(A) Pilot hole drilling on the face of the glenoid 1 to 2 mm from
the glenoid rim. (B) Insertion of a double-loaded suture anchor
into the hole. (C) Passage of 3 suture strands (2 from the first
suture thread for the horizontal mattress suture and 1 from the
second suture thread for the additional vertical simple stitch).
(D) Bankart repair using modified Mason-Allen method. (E)
Illustration of the Bankart repair procedure using modified
Mason-Allen method.
passages were made for 3 suture strands, the horizontal
mattress capsulolabral suture was made with the 2 suture
strands of the first suture thread (first and second pas-
sages) at the capsular side with good restoration of an
anterior hump. Afterward, the vertical simple suture
with the second suture thread was added just medial and
perpendicular to the horizontal loop of the first suture
thread to emphasize the strength grip of the attenuated
capsulolabral tissue, to reduce the risk of pull-through of
suture, and to increase the contact force between capsulo-
labral tissue and glenoid bone (Figure 3).
An additional 2 to 4 anchors were implanted on the gle-
noid edge at approximately the 2- to 5-o’clock position with
the same procedures of the SS or the modified Mason-Allen
method, and successful Bankart repair and capsular shift,
with tightened and tensioned capsule, were created for all
patients of each group. Careful palpation with a probe was
performed to confirm the stability of the repaired capsulo-
labral complex at the glenoid rim for both groups. There
were 11 patients in the SS group with superior labral
6 Park et al
anterior and posterior (SLAP) lesions (type 2, n = 9; type 5,
n = 2) and 11 in the BRUMA group (type 2, n = 6; type 3,
n = 2; type 4, n = 2; type 5, n = 1). These SLAP lesions
were repaired because all the lesions were unstable. Type
1 SLAP lesions were neglected. The remplissage procedure
was not performed for any patient.
Rehabilitation
The same rehabilitation protocol was applied to all patients.
Shoulders were immobilized for 6 weeks in a 30° abduction
brace. Shrugging of both shoulders, active elbow flexion/
extension, active forearm supination/pronation, and active
hand and wrist motion were encouraged immediately after
surgery. Gentle passive forward flexion exercise with a pul-
ley was initiated after weaning off the sling and lasted for 3
weeks. After the pulley period, patients started external
rotation exercises with a stick for 3 or 4 weeks. Muscle
strengthening was started approximately at 12 weeks
postoperatively. Sports activities were gradually permitted
beginning at 4 to 5 months after the operation.
Evaluation
The standardized CT arthrogram was performed for all
patients preoperatively and at 6 months postoperatively.
The time frame of the 6 months after Bankart repair is
generally accepted in other studies for anatomic evaluation
of the reconstructed labrum.28,32 Evaluation of postopera-
tive CT arthrogram was performed on a picture archiving
and communication system monitor (Centricity Enterprise
Archive; GE) with mouse-point cursor and automated
computer calculation for distance. A radiologist (S.G.M.)
injected 12 to 20 mL of diluted (65%) iodinated contrast
material (Bonorex 300 Isohexol; Central Medical Service)
into the glenohumeral joint with ultrasound guidance via
a posterior approach. The scan was performed with a 16-
channel multidetector CT system (Brivo CT 385; GE) with
the patient’s shoulder in neutral rotation and 0° of abduc-
tion. For data analysis, oblique coronal, oblique sagittal,
and axial reconstructions were produced at a 3-dimensional
workstation, with a 2-mm section thickness and no recon-
struction interval for axial images and with a 2-mm recon-
struction interval for the oblique coronal and sagittal
images. Oblique coronal images were reconstructed parallel
to the supraspinatus muscle and the oblique sagittal images
parallel to the joint surface of the glenoid, with an identical
section thickness and reconstruction interval.
For the anatomic outcome, the retear rate after Bankart
repair and the labral height and width of the repaired
labrum were assessed with the postoperative CT arthro-
gram. A musculoskeletal radiologist who had .10 years
of experience and was unaware of the present study per-
formed and interpreted the CT arthrogram and further
evaluated the presence of retear of the repaired labrum.
The labral height and width were measured following the
method of Kim et al.13 Briefly, the 3-, 4-, and 5-o’clock posi-
tions were decided from the oblique coronal cut with the
longest superior-inferior length after the cut numbers
The American Journal of Sports Medicine
Figure 4. The measurement of the labral height and width on
an axial computed tomography arthrogram image.
were extrapolated via the 12-o’clock (most superior) and
6-o’clock (most inferior) image cuts. Then, each labral
height and width at the corresponding 3-, 4-, and 5-o’clock
positions were measured on axial CT arthrogram images.
The labral height was defined as the vertical distance (in
millimeters) from the lowest portion of the glenoid to the
maximum height tip of the labrum and the width as the
distance (in millimeters) from the anterior glenoid rim to
the most anterior portion of the labrum (Figure 4).
To assess interobserver reliability, 2 orthopaedic surgeons
with 6 and 8 years of experience (J.S.L. and colleague,
respectively) separately and randomly measured the heights
and widths for all patients. Each rater was unaware of the
other’s ratings. For intraobserver reliability, 1 rater (J.S.L.)
performed a second measurement with the same images 3
weeks after the first measurement without knowing the first
rating. The means of both raters’ ratings (each labral height
and width) at the 3-, 4-, and 5-o’clock positions were com-
pared between groups. For the functional outcome, shoulder
range of motion, VAS for pain, ASES score, Constant score,
and Rowe score were evaluated for all patients by the senior
author and fellowship trainees (J-Y.P., J.S.L.) and data were
analyzed from the preoperative and final follow-up assess-
ments at least 2 years after surgery. For the evaluation of
shoulder range of motion, passive motion in 4 directions
was measured by a senior investigator (J-Y.P.): (1) forward
flexion was measured in degrees between the arm and the
thorax with the elbow held straight, (2) external rotation at
the side was measured in degrees between the thorax and
the forearm with the arm held in an adducted position and
with the elbow in 90° of flexion, (3) external rotation at 90°
AJSM Vol. XX, No. X, XXXX
TABLE 2
Inter- and Intraobserver Reliabilities of Labral Height and Width Measurements
at Each Location of Computed Tomography Arthrograma
Preoperative
Variable Height
Intraobserver, rater 1
3 o’clock 0.922
4 o’clock 0.916
5 o’clock 0.909
Interobserver
3 o’clock 0.876
4 o’clock 0.880
5 o’clock 0.848
a
Values are presented as intraclass correlation coefficients.
abduction was measured in degrees of the rotation angle of
the forearm from the transverse plane of the upper arm
with the shoulder abducted to 90° and the elbow flexed to
90°, and (4) internal rotation across the back was measured
by the vertebral level that the patient could reach with the
tip of his or her thumb. For the analysis, the vertebral level
was numbered serially: 12 for the 12th thoracic vertebra, 13
for the 1st lumbar vertebra, 17 for the 5th lumbar vertebra,
and 18 for any level below the sacral region.22 The VAS for
pain was scored from 0 to 10, with 10 indicating the highest
level of pain. The ASES score consisted of a summation score
with a 100-point system (50 points for daily function and 50
points for pain). The history of redislocation was recorded,
and the presence of an anterior apprehension sign was
checked by a senior investigator (J-Y. P.) at an outpatient
clinic at every follow-up visit for each patient.
Statistical Analysis
Inter- and intraobserver reliabilities were calculated with the
intraclass correlation coefficient between 2 measurements of
different raters and different times. Either a paired t test or
a Wilcoxon signed-rank test was performed to determine any
significant differences in the labral heights and widths at each
o’clock position and between pre- and postoperative functional
scores. In addition, a Student t test or Mann-Whitney U test
was used to compare any differences in labral heights and
widths between groups at each o’clock position and in every
continuous variable, including every functional outcome
parameter. The chi-square test or Fisher exact test was
used to determine the differences between groups for categor-
ical variables. Analyses were performed with SPSS (v 18.0;
IBM), and P \ .05 was considered statistically significant.
RESULTS
Redislocation/Apprehension
Two patients in the simple repair group (4.4%) and 2 in the
BRUMA group (4.9%) experienced redislocation after
Modified M-A Stitch in Bankart Repair 7
Postoperative, 6 mo
Width Height Width
0.897 0.928 0.930
0.894 0.906 0.927
0.925 0.936 0.915
0.823 0.896 0.887
0.879 0.849 0.851
0.906 0.913 0.904
surgery. In addition, 4 patients in the simple repair group
(8.9%) and 2 in the BRUMA group (4.9%) showed a positive
apprehension sign after surgery (P = .399). All patients
who exhibited redislocation or a positive apprehension
sign had retears on postoperative imaging.
Anatomic Outcome
Retear. Five cases in the simple repair group (11.1%)
and 2 in the BRUMA group (4.9%) showed retear at 6
months after repair (P = .239). Among the 5 patients in
the simple repair group, 2 were athletes (1 snowboarder
and 1 diver), and 3 experienced trauma after repair (the
snowboarder hit a wall while snowboarding, the diver got
hit by an outside force while diving, the third patient expe-
rienced an external force during physical therapy). The
patient with trauma during physical therapy showed
apprehension but did not exhibit redislocation; the 2 ath-
letes showed redislocation and apprehension. The trau-
matic dislocation events of both athletes in \6 months
were in part attributed to the lack of compliance (ie,
returning too quickly to sport without sufficient rehabilita-
tion). The 2 patients who showed retear in the BRUMA
group were athletes (1 judo player and 1 baseball player),
and both experienced trauma with redislocation (positive
apprehension test) after repair during a match (the judo
player) and off-season training (the baseball player). All
5 patients who experienced postoperative trauma under-
went revision surgery (the physical therapy patient, at 6
months after initial surgery; the snowboarder, at 16
months; the diver, at 13.5 months; the judo player, at 7.5
months; and the baseball player, at 12 months), and no fur-
ther redislocation had occurred at final follow-up. The
other 3 patients without a history of trauma or redisloca-
tion did not complain of instability or pain, and additional
surgery was not required.
Labral Height and Width. The inter- and intraobserver
reliabilities of labral height and width measurements at
each location on CT arthrogram were excellent (Table 2).
At 6 months postoperatively, the height and width of
the reconstructed labrum were significantly increased at
8 Park et al The American Journal of Sports Medicine

TABLE 3
Comparison of the Labral Height and Width Measurements Between Groups at Each Time Pointa

Height, mm Width, mm

Glenoid Location Simple Repair BRUMA P Value Simple Repair BRUMA P Value

Preoperative
3 o’clock 1.44 6 1.02 1.77 6 1.14 .143 1.13 6 0.81 1.46 6 0.99 .116
4 o’clock 1.25 6 0.93 1.55 6 0.93 .235 1.22 6 0.99 1.17 6 0.69 .838
5 o’clock 1.59 6 1.39 1.75 6 1.01 .636 1.09 6 0.71 1.42 6 1.17 .134
Postoperative 6 mo
3 o’clock 5.06 6 1.48 5.27 6 1.62 .534 4.81 6 1.66 4.78 6 1.51 .932
4 o’clock 5.43 6 1.55 5.76 6 1.69 .358 5.30 6 1.66 5.49 6 1.78 .622
5 o’clock 5.67 6 1.61 6.15 6 2.09 .243 5.16 6 1.40 5.56 6 1.49 .206

a
Values are presented as mean 6 SD. Simple repair, n = 45; BRUMA, n = 41. BRUMA, Bankart repair using modified Mason-Allen
technique.

TABLE 4 technique.12,25 For the Bankart repair, Castagna et al5

Postoperative Functional Outcomes introduced this modified Mason-Allen technique to
Between Groups at Final Follow-upa enhance the tissue-holding power and to prevent the cut-
ting through of the attenuated capsulolabral tissue. By
Variable SS Group BRUMA Group P Value using this modified Mason-Allen technique in the context
Postoperative of Bankart repair, the grip strength and initial fixation
FF 178.77 6 6.32 178.29 6 6.08 .718 strength can be improved, especially in shoulder with
ER, at side 62.67 6 16.77 62.31 6 12.94 .915 attenuated capsulolabral tissue, and the failure rate of
ER, 90° abduction 77.67 6 5.99 75.48 6 7.81 .148 the Bankart repair may be decreased.26 To date, there is
IR, at back 8.42 6 1.68 7.92 6 1.70 .180 no report evaluating the clinical outcomes of the modified
VAS for pain 0.26 6 0.49 0.12 6 0.34 .159 Mason-Allen stitch for Bankart repair. However, several
ASES score 93.9 6 4.1 94.2 6 3.7 .759 studies have assessed the outcomes of the horizontal mat-
Constant score 91.3 6 3.8 91.5 6 4.0 .752 tress suture for the Bankart repair, which may support the
Rowe score 95.6 6 3.5 95.6 6 2.9 .949
effectiveness of the modified Mason-Allen technique, as the
a
ASES, American Shoulder and Elbow Surgeons; BRUMA,
horizontal mattress suture is one of the main components
Bankart repair using modified Mason-Allen technique; ER, exter- for this technique.10,17 Dines and ElAttrache10 reported sec-
nal rotation; FF, forward flexion; IR, internal rotation; SS, vertical ond-look arthroscopic surgery among patients 7 months after
simple stitch method; VAS, visual analog scale. horizontal mattress repair, showing a well-healed labrum
with a restored normal appearance, as compared with a sim-
ple stitch that failed to re-create the normal labral anatomy.
all locations of the glenoid in both groups, especially at the In addition, Lee et al17 reported that the arthroscopic Bank-
5-o’clock position, as compared with the preoperative art repair with a horizontal mattress suture is a safe and reli-
labrum (all P \ .001). There were no statistical differences able treatment for shoulder instability, with low recurrence
in labral height and width at all clock positions between rates. However, we failed to detect significant differences in
groups (all P . .05) (Table 3). the retear rate (SS group: 11.1%, 5 of 45; BRUMA group,
4.9%, 2 of 41; P = .239) and the redislocation/apprehension
rate (SS group: 8.9%, 4 of 45; BRUMA group: 4.9%, 2 of 41;
Functional Outcomes P = .399), probably because of the small incidences.
Given that the previously reported recurrence rate after
Postoperatively, all functional outcome scores were signif- Bankart repair with the horizontal mattress suture is
icantly improved in both groups (all P \ .001). However, 7% to 8% (7.89% in the study of Lee et al,17 7.0% in the
there was no significant difference between groups in any study of Mishra and Fanton20), the 4.9% redislocation/
functional outcomes assessed at final follow-up (.2 years apprehension rate at 2 years after the BRUMA seems to
after surgery, all P . .05) (Table 4). be quite promising. We cannot say that the modified
Mason-Allen technique for the Bankart repair is better
than the horizontal mattress suture, as we did not compare
DISCUSSION those 2 techniques directly and because of the limited num-
ber of patients who showed recurrence. Unfortunately, we
The modified Mason-Allen technique was first described in also failed to prove the anatomic superiority of the BRUMA
rotator cuff repair surgery by Schneeberger et al,27 and in terms of labral height and width. However, given that
various other authors subsequently showed excellent bio- the modified Mason-Allen technique is a suture technique
mechanical results and good clinical outcomes using this that combines an additional vertical simple stitch on top of
AJSM Vol. XX, No. X, XXXX
the horizontal mattress suture, it may be natural to think
that this technique would have mechanical superiority to
the horizontal mattress suture as well as a vertical simple
stitch and would further improve clinical outcomes. The
importance of restoring labral height in a Bankart repair
has been well demonstrated,13 and its restoration seems to
be mainly derived from the horizontal mattress suture of
the modified Mason-Allen technique. Hagstrom and
Marzo11 reported that the horizontal mattress suture
achieved better labral anatomy restoration and increased
stability to glenohumeral repair. Similarly, Lazarus et al16
demonstrated that this suture technique better restored lab-
ral height and anatomy when compared with a simple stitch
technique in the repair of acute Bankart lesions; they also
reported that restoration of labral height re-created glenola-
bral concavity and stabilized the glenohumeral joint. Proper
recovery of labral height with proper tensioning of the cap-
suloligamentous tissue would be an essential process for
successful arthroscopic Bankart repair.
For the functional outcomes, there was no difference in
the range of motion, including external rotation and vari-
ous functional outcome scores. There could be a concern
that the modified Mason-Allen suture for Bankart lesion
may sacrifice external rotation, as this technique involves
placation of more inferior glenohumeral ligaments by hor-
izontal mattress sutures. However, the mean range of
motion, including external rotation (determined clinically),
was not statistically different between the SS group and
the BRUMA group. In addition, the various functional out-
come scores were not different between groups.
This was the first study to prospectively compare the
radiologic and clinical outcomes of the vertical simple stitch
and the modified Mason-Allen stitch for Bankart repair,
especially according to time of repair. Nevertheless, several
limitations should be considered. First, the evaluation time
of 6 months may be relatively early for the final diagnosis of
labral healing and for the evaluation of the height and
width of the repaired labrum, even though the time frame
of 6 months after Bankart repair was generally accepted
in other studies for anatomic evaluation of the recon-
structed labrum.28,32 There may be further change of the
labral healing status and the height and width of the
repaired labrum. Further study with longer image follow-
up may be needed. Second, there could be an error in the
measurement of labral height and width, even though the
reliability of the measurements was excellent in this study.
CONCLUSION
In conclusion, there was no difference in the radiologic out-
comes at 6 months and clinical outcomes assessed at least 2
years after surgery between the vertical simple stitch
group and the modified Mason-Allen stitch group.
REFERENCES
1. Bahk MS, Karzel RP, Snyder SJ. Arthroscopic posterior stabilization
and anterior capsular plication for recurrent posterior glenohumeral
instability. Arthroscopy. 2010;26:1172-1180.
Modified M-A Stitch in Bankart Repair 9
2. Baker CL, Uribe JW, Whitman C. Arthroscopic evaluation of acute ini-
tial anterior shoulder dislocations. Am J Sports Med. 1990;18:25-28.
3. Boddula MR, Adamson GJ, Gupta A, McGarry MH, Lee TQ. Restora-
tion of labral anatomy and biomechanics after superior labral anterior-
posterior repair: comparison of mattress versus simple technique. Am
J Sports Med. 2012;40:875-881.
4. Bottoni CR, Franks BR, Moore JH, DeBerardino TM, Taylor DC,
Arciero RA. Operative stabilization of posterior shoulder instability.
Am J Sports Med. 2005;33:996-1002.
5. Castagna A, Conti M, Mouhsine E, Delle Rose G, Massazza G, Gar-
ofalo R. A new technique to improve tissue grip and contact force in
arthroscopic capsulolabral repair: the MIBA stitch. Knee Surg Sports
Traumatol Arthrosc. 2008;16:415-419.
6. Castagna A, Delle Rose G, Borroni M, et al. Arthroscopic stabilization
of the shoulder in adolescent athletes participating in overhead or
contact sports. Arthroscopy. 2012;28:309-315.
7. Chen AL, Hunt SA, Hawkins RJ, Zuckerman JD. Management of
bone loss associated with recurrent anterior glenohumeral instability.
Am J Sports Med. 2005;33:912-925.
8. Chung SW, Kim SH, Tae SK, Yoon JP, Choi JA, Oh JH. Is the supra-
spinatus muscle atrophy truly irreversible after surgical repair of rota-
tor cuff tears? Clin Orthop Surg. 2013;5:55-65.
9. d’Elia G, Di Giacomo A, D’Alessandro P, Cirillo LC. Traumatic anterior
glenohumeral instability: quantification of glenoid bone loss by spiral
CT. Radiol Med. 2008;113:496-503.
10. Dines JS, ElAttrache NS. Horizontal mattress with a knotless anchor
to better recreate the normal superior labrum anatomy. Arthroscopy.
2008;24:1422-1425.
11. Hagstrom LS, Marzo JM. Simple versus horizontal suture anchor
repair of Bankart lesions: which better restores labral anatomy?
Arthroscopy. 2013;29:325-329.
12. Hawi N, Dratzidis A, Kraemer M, et al. Biomechanical evaluation of
the simple cinch stitch for arthroscopic rotator cuff repair. Clin Bio-
mech (Bristol, Avon). 2016;36:21-25.
13. Kim JY, Chung SW, Kwak JY. Morphological characteristics of the
repaired labrum according to glenoid location and its clinical relevance
after arthroscopic Bankart repair: postoperative evaluation with com-
puted tomography arthrography. Am J Sports Med. 2014;42:1304-1314.
14. Kim YS, Ok, JH. Arthroscopic reconstruction of bony defect in shoul-
der instability. Clin Shoulder Elbow. 2011;14:117-124.
15. Ko SH JK, Shin SM, Park HC. Using the arthroscopic remplissage of
anterior shoulder instability with Hill-Sachs lesion. Clin Shoulder
Elbow. 2011;14:53-58.
16. Lazarus MD, Sidles JA, Harryman DT 2nd, Matsen FA 3rd. Effect of
a chondral-labral defect on glenoid concavity and glenohumeral sta-
bility: a cadaveric model. J Bone Joint Surg Am. 1996;78:94-102.
17. Lee KH, Soeharno H, Chew CP, Lie D. Arthroscopic Bankart repair
augmented by plication of the inferior glenohumeral ligament via hor-
izontal mattress suturing for traumatic shoulder instability. Singapore
Med J. 2013;54:555-559.
18. Marquardt B, Witt KA, Liem D, Steinbeck J, Potzl W. Arthroscopic
Bankart repair in traumatic anterior shoulder instability using a suture
anchor technique. Arthroscopy. 2006;22:931-936.
19. Metcalf MH, Pon JD, Harryman DT 2nd, Loutzenheiser T, Sidles JA.
Capsulolabral augmentation increases glenohumeral stability in the
cadaver shoulder. J Shoulder Elbow Surg. 2001;10:532-538.
20. Mishra DK, Fanton GS. Two-year outcome of arthroscopic Bankart
repair and electrothermal-assisted capsulorrhaphy for recurrent trau-
matic anterior shoulder instability. Arthroscopy. 2001;17:844-849.
21. Nho SJ, Frank RM, Van Thiel GS, et al. A biomechanical analysis of
anterior Bankart repair using suture anchors. Am J Sports Med.
2010;38:1405-1412.
22. Oh JH, Kim SH, Lee HK, Jo KH, Bin SW, Gong HS. Moderate preop-
erative shoulder stiffness does not alter the clinical outcome of rotator
cuff repair with arthroscopic release and manipulation. Arthroscopy.
2008;24:983-991.
23. Ostermann RC, Hofbauer M, Platzer P, Moen TC. The ‘‘labral
bridge’’: a novel technique for arthroscopic anatomic knotless Bank-
art repair. Arthrosc Tech. 2015;4:e91-e95.
10 Park et al
24. Park JY, Lee SJ, Lhee SH, Oh JH. Change in labrum height after
arthroscopic Bankart repair: correlation with preoperative tissue qual-
ity and clinical outcome. J Shoulder Elbow Surg. 2012;21:1712-1720.
25. Porto Fde M, Alves MW, de Andrade AL. Evaluation of patients
undergoing rotator cuff suture with the modified Mason-Allen tech-
nique. Acta Ortop Bras. 2013;21:167-169.
26. Scheibel MT, Habermeyer P. A modified Mason-Allen technique for
rotator cuff repair using suture anchors. Arthroscopy. 2003;19:330-333.
27. Schneeberger AG, Beck M, Gerber C. In-vitro stability of rotator cuff
repair techniques [in German]. Orthopade. 1995;24:268-274.
28. Stein T, Mehling AP, Reck C, et al. MRI assessment of the structural
labrum integrity after Bankart repair using knotless bio-anchors.
Knee Surg Sports Traumatol Arthrosc. 2011;19:1771-1779.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.
The American Journal of Sports Medicine
29. van der Linde JA, van Kampen DA, Terwee CB, Dijksman LM, Klein-
jan G, Willems WJ. Long-term results after arthroscopic shoulder sta-
bilization using suture anchors: an 8- to 10-year follow-up. Am J
Sports Med. 2011;39:2396-2403.
30. Wiesel BB, Gartsman GM, Press CM, et al. What went wrong and
what was done about it: pitfalls in the treatment of common shoulder
surgery. J Bone Joint Surg Am. 2013;95:2061-2070.
31. Wolf EM. Anterior portals in shoulder arthroscopy. Arthroscopy.
1989;5:201-208.
32. Yoo JC, Lee YS, Tae SK, Park JH, Park JW, Ha HC. Magnetic res-
onance imaging appearance of a repaired capsulolabral complex
after arthroscopic Bankart repair. Am J Sports Med. 2008;36:
2310-2316.