Effect of Two Rehabilitation Protocols on Range of Motion and

Healing Rates After Arthroscopic Rotator Cuff Repair:

Aggressive Versus Limited Early Passive Exercises
Bong Gun Lee, M.D., Nam Su Cho, M.D., and Yong Girl Rhee, M.D.

Purpose: To compare range of motion and healing rates between 2 different rehabilitation protocols
after arthroscopic single-row repair for full-thickness rotator cuff tear. Methods: Sixty-four shoul-
ders available for postoperative magnetic resonance imaging (MRI) evaluation after arthroscopic
rotator cuff repair were enrolled in this study. Aggressive early passive rehabilitation (manual therapy
[2 times per day] and unlimited self–passive stretching exercise) was performed in 30 shoulders
(group A) and limited early passive rehabilitation (limited continuous passive motion exercise and
limited self–passive exercise) in 34 shoulders (group B). A postoperative MRI scan was performed
at a mean of 7.6 months (range, 6 to 12 months) after surgery. Results: Regarding range of motion,
group A improved more rapidly in forward flexion, external rotation at the side, internal and external
rotation at 90° of abduction, and abduction than group B until 3 months postoperatively with
significant differences. However, there were no statistically significant differences between the 2
groups at 1-year follow-up (P ⫽ .827 for forward flexion, P ⫽ .132 for external rotation at the side,
P ⫽ .661 for external rotation at 90° of abduction, and P ⫽ .252 for abduction), except in internal
rotation at 90° of abduction (P ⫽ .021). In assessing the repair integrity with postoperative MRI
scans, 7 of 30 cases (23.3%) in group A and 3 of 34 cases (8.8%) in group B had retears, but the
difference was not statistically significant (P ⫽ .106). Conclusions: Pain, range of motion, muscle
strength, and function all significantly improved after arthroscopic rotator cuff repair, regardless of
early postoperative rehabilitation protocols. However, aggressive early motion may increase the possi-
bility of anatomic failure at the repaired cuff. A gentle rehabilitation protocol with limits in range of
motion and exercise times after arthroscopic rotator cuff repair would be better for tendon healing without
taking any substantial risks. Level of Evidence: Level II, randomized controlled trial.

A rthroscopic rotator cuff repair has widely been

used in replacement of an open repair. Various
operative techniques are now available, reporting
largely good clinical results. Initially, single-row re-
pair techniques were performed,1,2 and then double-
row repair techniques were developed to increase the
contact area for tendon healing.3 More recently, a
suture bridge technique that is known to provide bio-
From the Shoulder & Elbow Clinic, Department of Orthopaedic
Surgery, College of Medicine, Kyung Hee University, Seoul, South Korea.
The authors report no conflict of interest.
Received September 26, 2010; accepted July 19, 2011.
Address correspondence to Yong Girl Rhee, M.D., Department
of Orthopaedic Surgery, College of Medicine, Kyung Hee Univer-
sity, 1 Hoegi-dong, Dongdaemun-gu, Seoul 130-702, South Korea.
E-mail: shoulderrhee@hanmail.net
© 2012 by the Arthroscopy Association of North America
0749-8063/10569/$36.00
doi:10.1016/j.arthro.2011.07.012
mechanically strong reconstruction of the footprint
has drawn attention.4 Because arthroscopic repairs
allow rigid fixation comparable to open repairs, early
passive exercises are generally recommended.3,5-7 De-
spite advancement and refinement in arthroscopic
techniques, however, high rates of structural failure
after arthroscopic rotator cuff repair have been re-
ported in the literature.1,8 Many surgical and nonsur-
gical factors affect whether a tendon will successfully
heal to the tuberosity after repair. Major factors that
the surgeon cannot control include patient age, size
and chronicity of the tear, and muscle fatty degener-
ation and atrophy. The surgeon can control factors
related to surgical technique and postoperative reha-
bilitation regimens. Among various factors, the early
rehabilitation usually recommended for preventing
postoperative stiffness may also act as 1 of the factors
that affects clinical and structural outcome.9 To date,

34 Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 28, No 1 (January), 2012: pp 34-42
 EFFECT OF TWO REHABILITATION PROTOCOLS
most studies have mainly focused on operative tech-
niques for anatomic reconstruction of the footprint and
rigid fixation. However, few studies have compared
clinical outcomes of different postoperative rehabili-
tation protocols, especially not to mention anatomic
results after each rehabilitation protocol. The purpose
of our study was to compare the clinical and structural
outcome between 2 different rehabilitation protocols
after arthroscopic single-row repair for full-thickness
rotator cuff tear. The patients were divided into 2
different groups according to the method of rehabili-
tation protocols until 6 weeks postoperatively because
the first 6 weeks after surgery is an important period
for ingrowth of the tendon to the bone surface. We
hypothesized that the range of motion (ROM) in the
aggressive early rehabilitation protocol after ar-
throscopic rotator cuff repair would improve more
rapidly with better final results than in the limited
early rehabilitation protocol. However, the aggressive
early rehabilitation protocol would inevitably show a
higher rate of structural failure than the limited early
rehabilitation protocol.
METHODS
Patient Selection
From November 2005 to December 2007, 432 con-
secutive shoulders received arthroscopic rotator cuff
FIGURE 1. Patient selection and randomization process. (AC, acromioclavicular.)
35
repair for the treatment of full-thickness rotator cuff
tears. For precise comparative study under homoge-
neous conditions, only medium-sized or large-sized
tears that could be repaired without undue tension
with a single-row repair technique based on ar-
throscopic findings were enrolled in this study. The
cuff tears were diagnosed primarily by physical ex-
amination and magnetic resonance imaging (MRI).
Tear size was measured intraoperatively with a cali-
brated probe. According to the classification of DeOrio
and Cofield,10 tear size was categorized as medium (1
to 3 cm) or large (3 to 5 cm). Partial rotator cuff tears,
small-sized tears, or massive tears were excluded.
Those tears repaired with other repair techniques in-
cluding suture bridge repair and modified Mason-
Allen single-row repair were also excluded from the
study. Those who had SLAP lesions, acromioclavic-
ular arthritis that required distal clavicle resection,
advanced glenohumeral arthritis, or rotator cuff tears
with a Workers’ Compensation claim or those who
needed tenotomy or tenodesis of the long head of the
biceps were excluded from the study. Patients under-
going revision procedures were also excluded (Fig 1).
Before undergoing the operation after being diag-
nosed with rotator cuff tears, patients consented to be
included in the study after receiving comprehensive
information regarding the study protocol. The surgeon
was blinded to the randomization in rehabilitation
36 B. G. LEE ET AL.

before surgery. After surgery, candidates who did not
have any exclusion criteria were randomized alterna-
tively into 1 of 2 rehabilitation protocols (type A or
B). A total of 85 patients were initially enrolled in this
study. Six patients were lost to follow-up within 1
year after the operation. Fifteen patients refused post-
operative MRI evaluation. Therefore 64 patients avail-
able for postoperative MRI evaluation at least 6
months after arthroscopic rotator cuff repair were fi-
nally enrolled in this study. According to the early
postoperative rehabilitation protocol, 30 shoulders
were enrolled in group A and 34 shoulders in group B.
The arthroscopic findings of group A showed medium-
sized tears in 21 shoulders (70%) and large-sized tears
in 9 (30%). Group B had medium-sized tears in 20
shoulders (59%) and large-sized tears in 14 (41%).
The mean age at the time of the operation was 54.5
years (range, 39 to 66 years) in group A and 55.2 years
(range, 40 to 65 years) in group B. The mean follow-up
period was 23.8 months (range, 13 to 67 months) and 26.4
months (range, 14 to 64 months), respectively. The mean
number of anchors used for repair was 2.12 (range, 1 to 4)
and 2.31 (range, 1 to 4), respectively. There were no sig-
nificant differences in the demographic data (Table 1).
Preoperative MRI was available in 25 cases in
group A and 29 cases in group B. Fatty degeneration
was evaluated for each muscle with the 5-stage grad-
ing system developed by Goutallier et al.11 Muscle
atrophy was evaluated on the oblique sagittal plane
image medial to the level of the coracoid process with
the 4-stage grading system (normal, mild, moderate,
or severe) developed by Warner et al.12 There were no
significant differences in fatty degeneration and mus-
cle atrophy between the 2 groups.
Preoperative and Postoperative Evaluations
A single senior author (Y.G.R.) who was an ortho-
paedic surgeon examined the patients both before and
after surgery. All patients underwent a physical exam-
ination 1 day before the operation. Postoperative evalu-
ations were performed regularly at 3 months, 6 months,
9 months, and 1 year postoperatively on an outpatient
basis, and the results from the last follow-up were ana-
lyzed. Preoperative and postoperative subjective pain
scores were measured with the visual analog scale
(VAS). Quantitative muscle strength measurement of the
rotator cuff was assessed with use of a portable, handheld
Nottingham Mecmesin Myometer (Mecmesin, Notting-
ham, England). Elevation strength was tested with the
patient in the seated position with the arm flexed to 90°
in the scapular plane. External and internal rotation was
tested with the shoulder in a neutral position and the
elbow in 90° of flexion. For shoulder ROM, forward
flexion, abduction, external rotation at the side and at 90°
of abduction, and internal rotation at 90° of abduction
were assessed before and after the operation. In this
study ROM is the sum of glenohumeral motion and
scapulothoracic motion. We did not record 2 joint mo-
tions separately.
Operative Techniques
All operations were performed by the senior author
with the patient in the beach-chair position with the
back of the bed flexed about 70°. Both shoulders were
examined with the patient under general anesthesia for
ROM. A posterior portal was established for the initial
assessment of the joint. An anterior portal through the
rotator interval and a lateral portal were established as

TABLE 1. Patient Demographics and Healing Rate

Group A (n ⫽ 30) Group B (n ⫽ 34) P Value

Male/female 21/9 20/14 .437

Right/left 22/8 23/11 .785
Mean age (range) (yr) 54.5 (39-66) 55.2 (40-65) .731
Dominant/nondominant 20/10 22/12 .872
Medium/large tear 21/9 20/14 .437
Mean no. of anchors (range) 2.12 (1-4) 2.31 (1-4) .419
Mean symptom duration (range) (mo) 16.5 (1-42) 14.3 (3-38) .681
Length of follow-up (range) (mo) 23.8 (13-67) 26.4 (14-64) .257
Fatty degeneration* 0.83 (0-2) 0.71 (0-3) .337
Global fatty degeneration index† 0.62 (0-1.3) 0.53 (0-1.7) .300
MRI follow-up time (range) (mo) 7.9 (6-12) 7.5 (6-12) .314
Healing/retear 23/7 31/3 .106

* Fatty degeneration was evaluated for the supraspinatus with the 5-stage grading system developed by
Goutallier et al.11
† Mean value of grades for supraspinatus, infraspinatus, and subscapularis.
 EFFECT OF TWO REHABILITATION PROTOCOLS
the working portals. After subacromial decompression
was completed, the posterolateral portal was used as a
viewing portal for the “Grand Canyon” view,13 and
the posterior and anterosuperior portals were used as
the “waiting room” portal. After adequate visualiza-
tion, preparation, and release of the tendon, the upper
surface of the greater tuberosity was abraded widely
with a shaver, with removal of all soft tissue and
cortical bone, to create a bleeding cancellous bone
bed. The greater tuberosity was gently debrided and
smoothed of irregularities and the superficial bone was
decorticated, but the medullary canal was not exposed.
A formal bone trough was not made. The suture
anchor, a 5-mm Super Revo (Linvatec, Largo, FL)
prethreaded with 2 No. 2 braided polyester sutures
(green and white), was inserted through the accessory
superolateral portal. The rotator cuff repair was per-
formed by first placing a suture shuttle through the
tendon by use of a suture hook (Linvatec) or Banana
SutureLasso (Arthrex, Naples, FL). A suture hook or
Banana SutureLasso was inserted through the working
portal or “3 sister portals”13 including the modified
Neviaser portal and was used to pass the suture some
distance medial to the tendon edge, close to the mus-
culotendinous junction. The shuttle was used to bring
1 end of the suture through the tendon. An ar-
throscopic knot was tied, reducing the tendon to the
bone. The number of suture anchors used depended on
tear size and configuration. For medium-sized tears, 1
or 2 anchors were used, and for large-sized tears, 3 or
4 anchors were used. After routine portal closure, a
shoulder-immobilizing sling with abduction pillow
was applied to the patients with the shoulder in 30° of
abduction and neutral rotation.
Postoperative Rehabilitation
The aggressive early passive rehabilitation (group
A) targeted recovering ROM to the preoperative level
at an early stage. Immediately after the operation,
passive shoulder stretching exercises and manual ther-
apy by a physical therapist started without a limitation
on ROM. In other words, from 1 day postoperatively,
passive ROM exercises including forward flexion and
external rotation were performed twice a day with the
support of a physical therapist. In addition to the
exercises led by the therapist, the patients themselves
were allowed to perform pendulum exercises, passive
forward flexion up to a tolerable range, and passive
external rotation to 30° in the supine position. They
performed self–passive ROM exercises in a tolerable
range at first and were instructed to perform the ex-
37
ercises 3 times a day, 10 rounds each time. The
patients wore a shoulder abduction brace during rest
until 6 weeks postoperatively and were strongly ad-
vised to perform home rehabilitation self-exercises
after discharge from the hospital. Active exercises
were not allowed until 6 weeks postoperatively or
until full passive ROM had been regained. Active-
assisted exercises were started at 6 weeks postopera-
tively, and muscle strengthening exercises were intro-
duced thereafter gradually. A return to recreational
activity with heavy demands on the shoulder or to
manual labor was delayed for 6 months (Fig 2).
The limited early passive rehabilitation protocol
(group B) targeted the minimum passive exercises to
prevent joint stiffness until the repaired rotator cuff
was healed. By use of a continuous passive motion
machine, only shoulder stretching exercises limited to
90° were performed 2 times a day until 3 weeks
postoperatively and these exercises gradually in-
creased to the possible range. Preoperatively, the pa-
tients were instructed on the use of the machine so that
they could perform only passive forward flexion up to
90° using continuous passive motion on day 1 post-
operatively without the support of a physical therapist.
Other exercises including external rotation were not
allowed. Just like group A patients, group B patients
wore a shoulder abduction brace until 6 weeks post-
operatively. From 3 weeks postoperatively, the group
B patients were instructed to increase the range of
passive forward flexion to the tolerable level and
gradually start other exercises such as external rota-
tion. As with group A, active exercises were not
allowed until 6 weeks postoperatively or full passive
ROM had been regained. Once active-assisted exer-
cises were allowed, group B followed the same reha-
bilitation protocol as group A (Fig 2). All patients
were educated repeatedly on performing the exact
exercise during the hospitalization period, and their
performance was checked at discharge and at 3 weeks
and 6 weeks postoperatively on an outpatient basis.
Assessment of Tendon Healing
To assess tendon healing, anatomic evaluation of
the cuff repair was performed with MRI as the inves-
tigation of choice, because it provides the benefits of
multiplanar imaging of the postoperative shoulder. A
postoperative MRI scan was performed at a mean of
7.6 months (range, 6 to 12 months) after surgery. All
studies were obtained with a 1.5-T unit (Signa; GE
Medical Systems, Milwaukee, WI) by use of the routine
38 B. G. LEE ET AL.

FIGURE 2. The 2 rehabilitation protocols used in this study. Asterisk, one time equals 10 rounds of each exercise. (CPM, continuous passive motion.)

pulse sequences. The images were reviewed by one
experienced senior radiologist who was informed that
the patients had undergone surgery for rotator cuff repair
and blinded to the size and location of the tear that had
been repaired. Continuity and rerupture of the tendon
were assessed on magnetic resonance images according
to established MRI criteria.3,14 When a fluid-equivalent
signal or nonvisualization of the supraspinatus, infraspi-
natus, or subscapularis tendon was found on at least 1
T2-weighted or proton density–weighted image, the di-
agnosis of a full-thickness retear (i.e., anatomic failure of
healing) was made (Fig 3).
Statistical Analysis
The Wilcoxon signed rank test was performed to
assess the difference in preoperative and postoper-

FIGURE 3. Postoperative coronal oblique MRI scans after single-row rotator cuff repair show (A) well-healed cuff and (B) retear at insertion site.
 EFFECT OF TWO REHABILITATION PROTOCOLS 39

TABLE 2. ROM
Preoperative 3 mo 6 mo 1 yr

Mean (°) P Mean (°) P Mean (°) P Mean (°) P

Forward flexion .672 .021* .368 .729

Group A 149.0 ⫾ 16.6 149.7 ⫾ 12.7 157.3 ⫾ 11.4 155.3 ⫾ 13.0
Group B 151.9 ⫾ 12.6 133.8 ⫾ 27.4 151.9 ⫾ 18.2 153.0 ⫾ 12.2
External rotation at side .223 .010* .007* .078
Group A 49.8 ⫾ 14.5 44.2 ⫾ 14.6 50.3 ⫾ 11.2 53.0 ⫾ 11.6
Group B 52.8 ⫾ 13.6 34.1 ⫾ 19.2 41.6 ⫾ 14.9 48.1 ⫾ 13.9
Internal rotation at 90° of abduction .323 .001* .003* .057
Group A 57.8 ⫾ 16.4 59.0 ⫾ 17.9 63.8 ⫾ 14.3 65.7 ⫾ 13.3
Group B 51.3 ⫾ 26.3 38.5 ⫾ 24.1 47.3 ⫾ 22.7 54.9 ⫾ 21.5
External rotation at 90° of abduction .056 .009* .266 .778
Group A 74.2 ⫾ 9.10 70.5 ⫾ 14.0 74.0 ⫾ 11.0 76.3 ⫾ 12.1
Group B 78.1 ⫾ 12.5 54.0 ⫾ 24.5 67.8 ⫾ 18.1 77.7 ⫾ 11.6
Abduction .168 .048* .315 .884
Group A 158.8 ⫾ 20.4 161.5 ⫾ 22.0 165.3 ⫾ 13.9 167.8 ⫾ 12.8
Group B 162.1 ⫾ 24.3 143.6 ⫾ 35.7 154.4 ⫾ 30.1 161.8 ⫾ 27.3

* The difference between the 2 groups was statistically significant (P ⬍ .05).

ative results of each group. The Mann-Whitney test
was used to compare continuous variables between
the groups. The Fisher exact test was performed to
evaluate the statistical significance of the difference
in quantitative data including retear rate between
the 2 groups. The SPSS software package (version
16.0; SPSS, Chicago, IL) was used for all statistical
analyses, with the ␣ level set at .05.
RESULTS
Pain
In group A the VAS score for pain at rest decreased
from a preoperative mean of 1.77 (range, 0 to 5) to
0.23 (range, 0 to 3) at 1 year postoperatively (P ⬍
.001). The VAS score for pain during motion decreased
from a preoperative mean of 6.36 (range, 4 to 9) to 1.47
(range, 0 to 5) at 1 year postoperatively (P ⬍ .001). In
group B the VAS score for pain at rest decreased from
a preoperative mean of 1.56 (range, 0 to 4) to 0.15
(range, 0 to 3) at 1 year postoperatively (P ⬍ .001).
The VAS score for pain during motion decreased from
a preoperative mean of 5.79 (range, 4 to 8) to 1.53
(range, 0 to 5) at 1 year postoperatively (P ⬍ .001).
Both groups showed significant improvement from
preoperative levels, but there was no statistically sig-
nificant difference between the 2 groups at 3 months
postoperatively, at 6 months postoperatively, and at
1-year follow-up (P ⫽ .702, P ⫽ .209, and P ⫽ .382,
respectively, for VAS at rest and P ⫽ .696, P ⫽ .154,
and P ⫽ .808, respectively, for VAS during motion).
Range of Motion
The preoperative and postoperative ROM is listed
in Table 2. In forward flexion both groups showed
decreased ROM at 3 months postoperatively (149.7°
in group A and 133.8° in group B), compared with
preoperative motion (149.0° in group A and 151.9° in
group B) but gradually recovered. When compared
with group A, group B showed statistically signifi-
cantly decreased ROM at 3 months postoperatively
(P ⫽ .021). After 6 months postoperatively, there was
no statistically significant difference between the 2
groups (P ⫽ .368).
In abduction and external rotation at 90° of abduc-
tion, group B also showed statistically significantly
decreased ROM at 3 months postoperatively com-
pared with group A (P ⫽ .010 and P ⫽ .048, respec-
tively). After 6 months postoperatively, there was no
statistically significant difference between the 2
groups (P ⫽ .315 and P ⫽ .266, respectively). In
external rotation at the side and internal rotation at 90°
of abduction, group B presented slower recovery than
did group A, with significant differences until 6
months postoperatively (P ⫽ .007 and P ⫽ .003,
respectively). However, there was no statistically sig-
nificant difference between the 2 groups at 1 year
postoperatively (P ⫽ .078 and P ⫽ .057, respectively)
(Table 2).
40 B. G. LEE ET AL.
Muscle Strength
Both groups recorded significant postoperative im-
provement in forward flexion, external rotation, and
internal rotation muscle strength (P ⬍ .001). Group B
measured 0.83 kg lower than group A in forward
flexion muscle strength at 3 months postoperatively,
but the difference was not statistically significant (P ⫽
.062). The mean muscle strength of group A during
elevation, external rotation, and internal rotation was
7.76, 7.94, and 8.90 kg, respectively, at 1 year post-
operatively. The corresponding strength was 7.33,
7.62, and 8.44 kg, respectively, in group B. However,
there was no statistical difference between the 2
groups (P ⫽ .227, P ⫽ .542, and P ⫽ .450, respec-
tively).
Clinical Assessment
At the last follow-up, the University of California,
Los Angeles score improved to 32.3 points in group A
and 31.8 points in group B from the preoperative
mean of 15.7 points and 16.0 points, respectively. The
difference between the groups was significant, with
29.4 points in group A and 26.5 points in group B, at
3 months postoperatively (P ⫽ .009). After 6 months
postoperatively, there was no statistically significant
difference between the 2 groups (P ⫽ .158). At the
last follow-up, group A had 16 excellent cases
(47.1%), 13 good (38.2%), 4 fair (11.8%), and 1 poor
(2.9%). Group B had 15 excellent cases (50%), 12
good (40%), and 3 fair (10%). Both groups reported
statistically significant improvement in clinical assess-
ment, but there was no statistical difference between
the 2 groups (P ⫽ .341).
Structural Outcome
In assessing the repair integrity of both groups with
postoperative MRI scans, we found complete healing
in 23 of 30 shoulders (76.7%) in group A and retear
was observed in 7 (23.3%). In group B complete
healing was found in 31 of 34 shoulders (91.2%) and
retear was observed in 3 (8.8%). The retear rate in
group A was higher than that in group B (23.3% v
8.8%). However, the difference was not statistically
significant (P ⫽ .106) (Table 1).
DISCUSSION
Arthroscopic repairs have recently been accepted as
a standard treatment for rotator cuff tears. However, a
high failure rate of tendon healing to the tuberosity
occurs even in procedures performed carefully by
experienced surgeons. The generally accepted retear
rate is 25% to 40%.9,15,16 Many factors are involved in
the successful healing or structural failure of the re-
paired tendon. Some of these factors that affect heal-
ing of the rotator cuff are beyond the surgeon’s con-
trol. The uncontrollable factors associated with rotator
cuff healing include muscle fatty degeneration and
atrophy, larger tear size, poor tendon quality, and
repetitive trauma from impingement. The surgeon can
control factors related to surgical technique (proper
surgical technique, tear pattern recognition, adequate
subacromial decompression, cuff mobilization, prep-
aration of the tuberosity, suture and knot tying tech-
nique, anchor placement, and surgeon experience) and
postoperative rehabilitation regimens.9,17-19 Careful
consideration of preoperative patient factors, followed
by repair using a meticulous surgical technique and
appropriate postoperative rehabilitation, leads to good
clinical and anatomic results.9 Commonly, a retear of
the repaired cuff has an influence on the clinical result,
even though a small retear without much retraction
does not lead to a loss of abduction strength.16 In this
regard, technical efforts to lower the retear rate have
been attempted.20-23 Current arthroscopic techniques
for rotator cuff tears have emphasized the superiority
of a double-row repair in the footprint strength and the
resultant decrease of anatomic failure. Recent studies
mainly focused on surgical factors, in particular oper-
ative techniques. However, postoperative rehabilita-
tion can be as crucial as surgical technique for suc-
cessful healing of the rotator cuff. Postoperative
rehabilitation, though not performed under the direct
supervision of the surgeon, is controlled by the sur-
geon, including postoperative immobilization. There-
fore we considered that comparative analysis of the
effects of postoperative rehabilitation on clinical re-
sults and tendon healing may become a basis for
establishing an appropriate rehabilitation protocol and
contribute to decreased structural failure. After the
operation, the joint motion affects rotator cuff heal-
ing,24,25 and the motion created in the process of
rehabilitation also puts stress on the repaired cuff
tendon, which may lead to structural failure. Sano et
al.26 showed that the single-row model represented a
high stress concentration on the bursal surface of the
tendon in a simulated situation. Cummins et al.27
showed that the predominant mode of failure in pa-
tients undergoing revision surgery was suture pulling
through tendon. Bigliani et al.18 reported that 5 of 31
failed rotator cuff repairs were correlated with post-
operative physical therapy. In a biomechanical study
 EFFECT OF TWO REHABILITATION PROTOCOLS
by Park et al.,28 external rotation motion after cuff
repair affected gap formation in the anterior supraspi-
natus tendon region. Gerber et al.29 showed that no
repair was able to withstand the high loads imposed by
postoperative weight bearing in an experimental sheep
model, even if optimum repair technique was used. In
our study the group undergoing aggressive early pas-
sive rehabilitation showed better recovery of ROM
until 3 months postoperatively than the group with
limited early passive rehabilitation, but after 6 months
postoperatively, no difference was found between the
2 groups except for internal rotation at 90° of abduc-
tion. However, the retear rate of the aggressive early
passive rehabilitation group was more than twice the
rate of the limited early passive rehabilitation group,
although there was no statistically significant differ-
ence in the final retear rate. Therefore we believe that
postoperative early motion during rehabilitation could
affect structural integrity of the repaired cuff tendon
adversely and limited early passive rehabilitation
would be better than aggressive early passive rehabil-
itation after arthroscopic rotator cuff repair.
Various postoperative rehabilitation protocols are
being discussed in the literature. Some groups apply
immobilization for the first 3 weeks postoperatively,
but others allow early passive motion.3,5-7 The range
and direction of motion allowed are also various.
Some authors restricted overhead stretching exercise
until 6 weeks postoperatively to avoid damaging the
repair.6 According to Frank et al.,30 external rotation
was limited to less than 20° with the arm adducted in
the early postoperative period. Instead of applying the
same postoperative rehabilitation protocol to all pa-
tients, Frank et al. adjusted the protocol in consider-
ation of preoperative patient factors and operative
findings. Rehabilitation therapy can be tailored to the
size of the tear. Accousti and Flatow9 recommended
that smaller, stable tears can be mobilized earlier than
large, retracted tears that should be protected with an
abduction pillow for 6 weeks after repair. Millett et
al.17 insisted that momentum and ROM be gradually
increased depending on the quality and structural integ-
rity of the rotator cuff. In our study only the medium-
sized or large-sized tears that could be repaired without
undue tension with a single-row repair technique
based on arthroscopic findings were enrolled for pre-
cise comparative study under homogeneous condi-
tions. Therefore the early rehabilitation protocol that
allowed rehabilitation immediately postoperatively
was adopted. The only difference was that the patients
were divided into 2 different groups according to the
method of passive exercises until 6 weeks postopera-
41
tively. Because the first 6 weeks after surgery is an
important period for ingrowth of the tendon to the
bone surface, we intended to determine the impact of
the exercise during this period. It can be easily sup-
posed that the aggressive early passive rehabilitation
protocol put more stress on the repaired rotator cuff
than the limited early passive rehabilitation protocol.
The advantages of our study are that (1) all operations
were performed by 1 surgeon with only a single-row
repair technique to exclude any technical differences
between the 2 groups, (2) it was a homogeneous study
that excluded other factors and selection bias that could
have impacted results by targeting only medium-sized or
large-sized tears repaired without undue tension, (3)
postoperative serial evaluations were performed regu-
larly on an outpatient basis, and (4) anatomic assess-
ment through postoperative MRI evaluation was per-
formed for all patients.
Our study has several limitations. First, this study’s
sample size was too small to determine any statistical
significance by parametric analysis. As a result, sta-
tistical analysis depended on only nonparametric
methods. Second, some possible confounding factors
that could have resulted in retears were not considered
in this study. These included compliance with postop-
erative rehabilitation and immobilization, diabetes,
and smoking. Third, the decision regarding retears
was made by MRI without contrast. However, repair
integrity has been evaluated by MRI without contrast
in several studies5,8,15,30 including that of Sugaya et
al.3 In addition, the interobserver reliability and in-
traobserver reliability were not evaluated on MRI
interpretation. However, we believed that the radiol-
ogist participating in this study had ripe experience in
the field of musculoskeletal radiology. Finally, the
glenohumeral joint motion was not isolated from the
scapulothoracic motion. We did not record 2 joint
motions separately. Actually, it is very difficult to
measure 2 joint motions separately on an outpatient
basis.
CONCLUSIONS
Pain, ROM, muscle strength, and function all sig-
nificantly improved after arthroscopic rotator cuff re-
pair, regardless of early postoperative rehabilitation
protocols. However, aggressive early passive motion
may increase the possibility of anatomic failure at the
repaired cuff. A gentle rehabilitation with limits in
ROM and exercise times after arthroscopic rotator
cuff repair would be better for tendon healing without
taking any substantial risks.
42 B. G. LEE ET AL.
REFERENCES
1. Galatz LM, Ball CM, Teefey SA, Middleton WD, Yamaguchi
K. The outcome and repair integrity of completely arthroscopi-
cally repaired large and massive rotator cuff tears. J Bone Joint
Surg Am 2004;86:219-224.
2. Wolf EM, Pennington WT, Agrawal V. Arthroscopic rotator
cuff repair: 4- to 10-year results. Arthroscopy 2004;20:5-12.
3. Sugaya H, Maeda K, Matsuki K, Moriishi J. Functional and
structural outcome after arthroscopic full-thickness rotator cuff
repair: Single-row versus dual-row fixation. Arthroscopy 2005;
21:1307-1316.
4. Park MC, ElAttrache NS, Tibone JE, Ahmad CS, Jun BJ, Lee
TQ. Part I: Footprint contact characteristics for a transosseous-
equivalent rotator cuff repair technique compared with a double-
row repair technique. J Shoulder Elbow Surg 2007;16:461-468.
5. Burks RT, Crim J, Brown N, Fink B, Greis PE. A prospective
randomized clinical trial comparing arthroscopic single- and dou-
ble-row rotator cuff repair: Magnetic resonance imaging and early
clinical evaluation. Am J Sports Med 2009;37:674-682.
6. Franceschi F, Ruzzini L, Longo UG, et al. Equivalent clinical
results of arthroscopic single-row and double-row suture an-
chor repair for rotator cuff tears: A randomized controlled trial.
Am J Sports Med 2007;35:1254-1260.
7. Charousset C, Grimberg J, Duranthon LD, Bellaiche L, Petro-
ver D. Can a double-row anchorage technique improve tendon
healing in arthroscopic rotator cuff repair? A prospective,
nonrandomized, comparative study of double-row and single-
row anchorage techniques with computed tomographic ar-
thrography tendon healing assessment. Am J Sports Med 2007;
35:1247-1253.
8. Bishop J, Klepps S, Lo IK, Bird J, Gladstone JN, Flatow EL.
Cuff integrity after arthroscopic versus open rotator cuff re-
pair: A prospective study. J Shoulder Elbow Surg 2006;15:
290-299.
9. Accousti KJ, Flatow EL. Technical pearls on how to maximize
healing of the rotator cuff. Instr Course Lect 2007;56:3-12.
10. DeOrio JK, Cofield RH. Results of a second attempt at surgical
repair of a failed initial rotator-cuff repair. J Bone Joint Surg
Am 1984;66:563-567.
11. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC.
Fatty muscle degeneration in cuff ruptures. Pre- and postop-
erative evaluation by CT scan. Clin Orthop Relat Res 1994:
78-83.
12. Warner JJ, Higgins L, Parsons IM IV, Dowdy P. Diagnosis
and treatment of anterosuperior rotator cuff tears. J Shoulder
Elbow Surg 2001;10:37-46.
13. Rhee YG, Vishvanathan T, Kishan B, Theera B, Lim CT. The
‘three sister portals’ for arthroscopic repair of massive rotator
cuff tears. Tech Shoulder Elbow Surg 2007;8:53-57.
14. Gusmer PB, Potter HG, Donovan WD, O’Brien SJ. MR im-
aging of the shoulder after rotator cuff repair. AJR Am J
Roentgenol 1997;168:559-563.
15. Cole BJ, McCarty LP III, Kang RW, Alford W, Lewis PB,
Hayden JK. Arthroscopic rotator cuff repair: Prospective
functional outcome and repair integrity at minimum 2-year
follow-up. J Shoulder Elbow Surg 2007;16:579-585.
16. Liem D, Bartl C, Lichtenberg S, Magosch P, Habermeyer P.
Clinical outcome and tendon integrity of arthroscopic versus
mini-open supraspinatus tendon repair: A magnetic resonance
imaging-controlled matched-pair analysis. Arthroscopy 2007;
23:514-521.
17. Millett PJ, Wilcox RB III, O’Holleran JD, Warner JJ. Reha-
bilitation of the rotator cuff: An evaluation-based approach.
J Am Acad Orthop Surg 2006;14:599-609.
18. Bigliani LU, Cordasco FA, McIlveen SJ, Musso ES. Operative
treatment of failed repairs of the rotator cuff. J Bone Joint Surg
Am 1992;74:1505-1515.
19. Boileau P, Brassart N, Watkinson DJ, Carles M, Hatzidakis
AM, Krishnan SG. Arthroscopic repair of full-thickness tears
of the supraspinatus: Does the tendon really heal? J Bone Joint
Surg Am 2005;87:1229-1240.
20. Gerber C, Schneeberger AG, Beck M, Schlegel U. Mechanical
strength of repairs of the rotator cuff. J Bone Joint Surg Br
1994;76:371-380.
21. Tao SS, Kaltenbach J. Arthroscopic placement of a modified
Mason-Allen stitch. Arthroscopy 2006;22:1248.e1-1248.e3.
Available online at www.arthroscopyjournal.org.
22. White CD, Bunker TD, Hooper RM. The strength of suture
configurations in arthroscopic rotator cuff repair. Arthroscopy
2006;22:837-841.
23. Gerber C, Fuchs B, Hodler J. The results of repair of massive
tears of the rotator cuff. J Bone Joint Surg Am 2000;82:505-515.
24. Thomopoulos S, Williams GR, Soslowsky LJ. Tendon to bone
healing: Differences in biomechanical, structural, and compo-
sitional properties due to a range of activity levels. J Biomech
Eng 2003;125:106-113.
25. Galatz LM, Charlton N, Das R, Kim HM, Havlioglu N,
Thomopoulos S. Complete removal of load is detrimental to
rotator cuff healing. J Shoulder Elbow Surg 2009;18:669-675.
26. Sano H, Yamashita T, Wakabayashi I, Itoi E. Stress distribu-
tion in the supraspinatus tendon after tendon repair: Suture
anchors versus transosseous suture fixation. Am J Sports Med
2007;35:542-546.
27. Cummins CA, Murrell GA. Mode of failure for rotator cuff
repair with suture anchors identified at revision surgery. J
Shoulder Elbow Surg 2003;12:128-133.
28. Park MC, Jun BJ, Park CJ, Ahmad CS, ElAttrache NS, Lee
TQ. The biomechanical effects of dynamic external rotation on
rotator cuff repair compared to testing with the humerus fixed.
Am J Sports Med 2007;35:1931-1939.
29. Gerber C, Schneeberger AG, Perren SM, Nyffeler RW. Ex-
perimental rotator cuff repair. A preliminary study. J Bone
Joint Surg Am 1999;81:1281-1290.
30. Frank JB, ElAttrache NS, Dines JS, Blackburn A, Crues J,
Tibone JE. Repair site integrity after arthroscopic transosse-
ous-equivalent suture-bridge rotator cuff repair. Am J Sports
Med 2008;36:1496-1503.