See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/14333776

Lag signs in the diagnosis of rotator cuff

rupture
Article in Journal of Shoulder and Elbow Surgery February 1995
DOI: 10.1016/S1058-2746(95)80101-4 Source: PubMed

CITATIONS

READS

114

521

4 authors, including:
Ralph Hertel

Christian Gerber

Lindenhofspital Bern

University of Zurich

129 PUBLICATIONS 2,900 CITATIONS

377 PUBLICATIONS 16,354 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

SEE PROFILE

Available from: Ralph Hertel

Retrieved on: 31 October 2016

Lag signs in the diagnosis of rotator

cuff rupture
R. Hertel, MD, F. T. Ballmer, MD, S. M. Lambert, FRCS, and Ch. Gerber, MD,
Berne, Switzerland

We assessed the relative value of lag signs for the evaluation of rotator cuff rupture in
a prospective study of 100 consecutive painful shoulders with impingement syndrome,
stages 1 to 3. Lag signs were compared with the Jobe and lift-off signs. Three tests
were designed to assess the main components of the rotator cuff: the external rotation
lag sign (ERLS) for the supraspinatus and the infraspinatus tendons, the drop sign for
the infraspinatus, and the internal rotation lag sign flRLS) for the subscapularis tendon.
For assessment of the supraspinatus and infraspinatus the ERLS was less sensitive but
more specific than the Jobe sign. The drop sign was the least sensitive but was as
specific as the ERLS. Partial ruptures of the supraspinatus remained concealed to the
ERLS. For assessment of the subscapularis the IRLS was as specific but more sensitive
than the lift-off sign. Partial ruptures of the subscapularis tendon could be missed by the
lift-off sign but were detected by the IRLS. The magnitude of the lag correlated with the
size of the rupture for both the ERLSand the IRLS. Clinical testing for lag signs was
efficient, reproducible, and reliable. In patients with little or no restriction of motion it
enhanced the accuracy of clinical diagnosis in rotator cuff lesions. (.I SHOULDERELBOW
SURG 1996;5:307-13.)
The clinical diagnosis of rotator cuff rupture is
based on the history and physical examination.
Although large ruptures can be diagnosed with
confidence, small lesions affecting a single tendon
may remain concealed when standard physical
signs are considered. At present the most useful
test for rupture of the posterosuperior rotator cuff is
considered to be the Jobe sign,'* which assesses
the ability (i.e., strength) of the affected shoulder to
maintain the arm in a position of 90 ~ elevation in
internal rotation againt a force applied by the
examiner. False-positive results may occur because
From the Department of Orthopaedic Surgery, Inselspital, Universi~/of Berne.
Dr. Lambert is supported by grants from AO Internationa), Davos,
Switzerland, The Wellington Foundation Fund, London, U.K.,
The St. Mary's Special Trustees Fund, London, U.K., The Ethicon
Foundation Fund at The Royal College of Surgeons of England,
London, U.K.
No benefits in any form have been received or will be received
from a commercial party related directly or indirectly to the
subject of this artide.
Reprint requests: R. Hertel, MD, Department of Orthopedic Surgery, Inselspital, 3010 Bern, Switzerland.
Copyright 9 1996 by Journal of Shoulder and Elbow Surgery
Board of Trustees.

1058-2746/96/$5.00 + 0 3211171631

pain may interfere with strength testing. For assessment of subscapularis function the lift-off sign, as
originally described, 2 is most useful, although incomplete ruptures may remain concealed. Positive
clinical diagnosis of a rotator cuff tear, its location,
and its size may therefore be difficult and imprecise.
We have assessed the value of three new signs,
which we decided to call the external rotation lag
sign, the drop sign, and the internal rotation lag
sign, designed to enhance our clinical diagnosis of
rotator cuff lesions. A concept of the assessment of
the malfunction of musculotendinous units at near
maximal contraction, leading to visible lags, is
presented.
PATIENTS AND METHODS
From March 1992 to December 1993, 100
consecutive patients with unilateral subacromial
impingement syndrome (stages 1 to 3)~ were entered into a prospective study. Only patients who
subsequently underwent open or arthroscopic rotator cuff exploration were included. Those with
any impairment of the passive range of glenohumeral motion were excluded.
The patients included 74 men and 26 women.

307

308

Hertel et al.

The median age was 51 years (range 16 to 79

years). The diagnosis at operation was a complete
rupture of the supraspinatus tendon in 17 patients
and a partial-thickness rupture in 19 patients.
Eleven had a rupture of both the supraspinatus and
infraspinatus tendons. Eight patients had an isolated rupture and five a partial rupture of the
subscapularis tendon. Sixteen had combined lesions consisting of a complete or partial rupture of
the supraspinatus, the infraspinatus, and the subscapularis tendons. Twenty-four patients had a
chronic impingement syndrome associated with a
hooked acromion but no structural lesion of the
rotator cuff.
For evaluation of the results we divided the
patients into two partially overlapping groups:
those with supraspinatus and infraspinatus disease
and those with subscapularis disease.

The lag signs: Description and

clinical application
The external rotation lag sign (ERLS).The

patient is seated on an examination couch with his

or her back to the physician. The elbow is passively flexed to 90 ~, and the shoulder is held at
20 ~ elevation (in the scapular plane) and near
maximal external rotation (i.e., maximum external
rotation minus 5 ~ to avoid elastic recoil in the
shoulder) by the physician. The patient is then
asked to actively maintain the position of external
rotation in elevation as the physician releases the
wrist while maintaining support of the limb at the
elbow. The sign is positive when a lag, or angular
drop, occurs (Figure l J. The magnitude of the lag
is recorded to the nearest 5 ~. For small ruptures the
movement may be subtle with a magnitude of as
little as 5 ~. With practice this movement can be
clearly appreciated, particularly when compared
with the (normal) contralateral side.
Testing and interpretation are complicated by
pathologic changes in the passive range of motion. When the passive range of motion is reduced
because of capsular contracture or increased because of a subscapularis rupture, for instance,
false-negative and false-positive results, respectively, must be expected. The ERLS is designed to
test the integrity of the supraspinatus and infraspinatus tendons.
The drop sign. The patient is seated on the
examination couch with his or her back to the
physician, who holds the affected arm at 90 ~ of

J. Shoulder Elbow Surg.

July~August 1996

elevation (in the scapular plane) and at almost full

external rotation, with the elbow flexed at 90 ~
(Figure 2). In this position the maintenance of the
position of external rotation of the shoulder is a
function mainly of the infraspinatus. 7 The patient is
asked to actively maintain this position as the
physician releases the wrist while supporting the
elbow. The sign is positive if a lag or "drop"
occurs. The magnitude of the lag is recorded to the
nearest 5 ~. Limitations in testing and interpretation
are the same as for the internal rotation lag sign
(IRLS). The drop sign is designed to assess the
function of infraspinatus.

The ERLS and the drop sign in suprascapular palsy. In complete suprascapular

palsy the function of both supraspinatus and infraspinatus will clearly be absent. Because in effect
this condition is indistinguishable from the clinical
presentation of a massive tear of the supraspinatus
and infraspinatus tendons, both the ERLS and the
drop sign will be positive. However, we assessed
the value of the ERLS and the drop sign by correlation of these signs with intraoperative findings,
and no lesions of the posterosuperior cuff were
present in four cases of complete isolated suprascapular palsy examined. We therefore adopted
the convention of defining both signs as "falsepositive" for the presence of a cuff tear after
negative operative findings when a peripheral
neuropathy was present.
The internal rotation lag sign. The patient
is seated on an examination couch with his or her
back to the physician. The affected arm is held by
the physician at almost maximal internal rotation.
The elbow is flexed to 90 ~, and the shoulder is
held at 20 ~ elevation and 20 ~ extension. The
dorsum of the hand is passively lifted away from
the lumbar region until almost full internal rotation
is reached. The patient is then asked to actively
maintain this position as the physician releases the
wrist while maintaining support at the elbow. The
sign is positive when a lag occurs (Figure 3). The
magnitude of the lag is recorded to the nearest 5 ~.
An obvious drop of the hand may occur with large
tears. A slight lag indicates a partial tear of the
cranial part of the subscapularis tendon. Limitations applied to the testing and interpretation of the
ERLS also apply in testing for the IRLS. The IRLS is
designed to test for the integrity of the subscapularis tendon.
For the purposes of this investigation the clinical

J. Shoulder Elbow Surg.

Volume 5, Number 4

Hertel et al.

309

Figure I Externalrotation lag sign. A, Arm is held at slight elevation and near
full external rotation; patient is then asked to maintain this position. B, Sign is
positive when Lag occurs. Magnitude of lag is judged in degrees.
examination included evaluation of the Jobe sign,
the lift-off sign, the ERLS, the drop sign, and the
IRLS. The magnitude of the angular lag was estimated to the nearest 5 ~, and was correlated with
the size of the tear measured at operation.
The sensitivity (Tp/Tp + Fn), specificity (Tn/Tn +
Fp), accuracy (Tp + Tn/total), positive predictive
value (Tp/Tp + Fp), and negative predictive value
(Tn/Tn+Fn) were calculated for each sign
(Tp = true-positive, Tn -- true-negative, Fp = falsepositive, and Fn = false negative). The significances of the measured differences between the signs
were also calculated with the critical ratio, z,
without the continuity correction.1

RESULTS
Supraspinatus and infraspinatus signs

(Table I). For the diagnosis of a tear in the pasterasuperior rotator cuff, the Jobe sign was more sensitive than the ERLS (p = 0.05). The external rotation lag sign was more sensitive than the drop sign
(p < 0.001 ). The external rotation lag sign was as
specific as the drop sign and was significantly
more specific than the Jobe sign (p = 0.002). No
significant difference was seen between the accuracy of the Jobe sign and the ERLS (p = 0.84). The
Jobe sign was more accurate than the drop sign
(p < 0.001). The external rotation lag sign was
more accurate than the drop sign (p < 0.001).

310

Hertel et al.

J. Shoulder Elbow Surg.

July~August 1996

Figure 2 Drop sign. A, Arm is held at 90 ~ elevation and near full external
rotation; patient is then asked to maintain this position. B, Sign is positive when
lag or drop occurs. Magnitude of lag is judged in degrees.
The Jobe sign had a positive predictive value of
84% for the presence of a posterosuperior cuff
tear. The external rotation lag and drop signs both
had a positive predictive value of 100% for the
presence of a posterosuperior cuff tear. The negative predictive values were 58% for the Jobe sign,
56% for the ERLS, and 32% for the drop sign.
Subscapularis signs (Table I). The IRLS was
significantly more sensitive than the lift-off sign
(p = 0.002). The IRLS and the lift-off sign were
equally highly specific (p = 1.0). The IRLS was
significantly more accurate than the lift-off sign
(p = O.OO7).
The lift-off sign had a positive predictive value of
100% for the presence of a tear of the subscapularis, and the IRLS had a positive predictive value
of 97%. The IRLS had a negative predictive value

of 96% for the absence of a tear of the subscapularis, and the lift-off sign had a negative predictive
value of 69%.

Correlation between the magnitude of

the lag signs and the extent of the rotator
cuff tears. No lag was observed for either the

ERLS,the drop sign, or the IRLSwhen the rotator cuff

was intact. Only one patient of 19 with partial rupture of the supraspinatus had a lag of 5 ~ A lag of 5 ~
to 10 ~ was observed in 16 of 17 patients with an
isolated tear of the supraspinatus tendon. A lag
between 10 ~ and 15 ~ was seen in all 11 patients
with complete ruptures of both supraspinatus and
infraspinatus and in 15 of 16 patients with massive
tears comprising ruptures of the supraspinatus, the
infraspinatus, and the subscapularis.
The drop sign was negative in all patients with

J. Shoulder Elbow Surg.

Volume 5, Number 4

Hertel et al.

311

Figure 3 internal rotation lag sign. A, Arm is held at slight extension and near
full internal rotation; patient is then asked to maintain this position. B, Sign is
positive when lag occurs. Magnitude of lag is judged in degrees.
intact rotator cuff and with partial rupture of the
supraspinatus. One of 17 patients with a complete
supraspinatus rupture had a drop of 10 ~ and 4 of
11 with complete supraspinatus and infraspinatus
rupture had a drop between 15 ~ and 20 ~ Eight of
the 16 patients with tears involving three tendons
had a drop between 15 ~ and 20 ~
Four of five patients with partial subscapularis
tears had an internal rotation tag of 5 ~, and all
eight patients with complete subscapularis rupture
had lags between 5 ~ and 10 ~.

DISCUSSION
Force measurements and observation of the
range of motion are useful for assessment of insufficiency of musculotendinous units. It was apparent
that estimation of force and measurement of the

lag were highly dependent on the position in

which these were observed. Any weakness of a
musculotendinous unit under observation is emphasized the closer it is tested to the position of
maximum possible shortening, because at this
length the muscle acts at greatest biologic and
mechanical disadvantage. Testing for the force
developed by an incompetent musculotendinous
unit is therefore most sensitive at a position close to
the shortest length of the unit. Besides testing for
strength, observation of the active range of motion
near to the maximum shortening of the musculotendinous unit afforded the most critical information.
The discrepancy between the active and passive
ranges of motion, manifest as the lag signs described, was therefore an indicator of the magnitude of the insufficiency.

312

Hertel et al.

J. Shoulder Elbow Surg.

July~August 1996

Table I Positive clinical tests and condition (diagnosis) of the rotator cuff

Diagnosis
Sensitivity
2
4
5
7
2, 4, 5, 7
3
6
7
3,6or7
Specificity
1
Accuracy
1, 2, 4, 5, or 7
1,3,6or7
Positive predictive
values
2, 4, 5, or 7
3, 6, or7
Negative predictive
values
2, 4, 5 or 7
3, 6, or7

Jobe

ERLS

Drop

15/19
14/17
10/11
7/8
53/63

1/19
16/17
I
1
44/63

0
1/17
4/11
1/2
13/63

Lift-off

7/8

11/16
18/29
14/24
67/87

53/63

14/24

1
37/87

24/43

1
68/87

24/74

IRLS

4/5
1

1
28/29

23/24

42/53

51/53

28/29

24/35

23/24

The denominator is the number of patients tested for the sign given. Where a value of 1 is given, all patients tested were positive
for the sign.
Diagnosis 1, intact rotator cuff; diagnosis 2, partial rupture of the supraspinatus; diagnosis 3, partial rupture of the subscapuloris;
diagnosis 4, supraspinatus rupture; diagnosis 5, supraspinatus and infraspinatus rupture; diagnosis 6, subscapularis rupture;
diagnosis 7, supraspinatus, infraspinatus, and subscapularis rupture.

By holding the patient's elbow passively in the

position of abduction, the confounding effect of the
interaction of the prime movers (deltoid, pectoralis
major) with the rotator cuff was reduced.
For assessment of the supraspinatus and infraspinatus the ERLS was less sensitive but more
specific than the Jobe sign, which indicates that the
Jobe sign might be a better screening test. However, the Jobe sign is biased by the fact that it often
enhances pain and thus may lead to reflex muscle
weakness even in the presence of an intact cuff.
One of the reasons for the success of the ERLS,we
believe, is the low level of pain felt by the patient
during the test when compared with the Jobe sign.
By contrast with the Jobe sign, the ERLSwas only
rarely false-positive and therefore maintained its
specificity in severely painful shoulders.
The position of the shoulder for the Jobe test (90 ~
of elevation) is held by the supraspinatus and the
deltoid. 3 Biomechanical evidence strongly suggests that holding the position for the ERLS (near
full external rotation with 20 ~ elevation) is more
dependent on the fully competent supraspinatus
than on the deltoid. 7 The moment arm for the
supraspinatus changes during the range of

rotation; as external rotation increases, more of the

muscle is located behind the longitudinal axis of
motion. 5 Its abductor moment-arm decreases, 7 and
the external rotation moment-arm increases. In this
position the deltoid cannot fully compensate for
rotator cuff insufficiency. The ERLS was therefore
considered to be a valid test for supraspinatus and
infraspinatus ruptures.
As the size of the rupture increased, the angular
drop also increased. With the ERLS, therefore, an
estimate of the extent (i.e., dimensions) of the
rupture could be made, thus providing the physician with a useful tool to assess the feasibility of the
planned repair.
In the position held for the drop sign the infraspinatus is as active as the supraspinatusS; extrapolation of data from Otis et al. z suggests that
external rotation in the position of 90 ~ elevation is
maintained substantially by the infraspinatus. The
drop sign was associated with a poor sensitivity
but had a high specificity. It should therefore be
used to validate the extent of the rupture into the
infraspinafus after the diagnosis has been established with the ERLS. We have observed a patient
with a traumatic lesion of the infraspinatus branch

J. Shoulder Elbow Surg.

Volume 5, Number 4
of the suprascapular nerve who had a negative
ERLS but a positive drop sign, further indicating
that the drop sign might be specific for the infraspinatus. Patte et al. 8 described a similar sign,
which they called "rappel automatique," tested at
90 ~ of abduction. 8 They associated this sign with
large ruptures of the supraspinatus and infraspinatus.
Little biomechanical data for the rotator cuff exist
for the position maintained during testing for the
IRLS. Clinical observation shows that the pectoralis
major is not able to contribute to the maintenance
of a position near full internal rotation because it
has maximally shortened before that position has
been reached. The maintenance of the position for
testing the IRLS is almost entirely a function of the
subscapularis. For complete and partial ruptures of
the subscapularis, the IRLS was significantly more
sensitive and as specific as the lift-off sign. The
negative predictive value was greater for the IRLS,
because it could detect partial ruptures of the
subscapularis tendon, now identified with increasing frequency in degenerative rotator cuff tears7
Positive predictive values were high for both signs.
The magnitude of the angular lag correlated with
the extent of rupture.
Testing and interpretation of lag signs is complicated by pathologic changes in the passive range
of motion. When the passive range of motion is
reduced, for example, because of capsular contracture, a potential lag will be reduced or completely masked, leading to false-negative results.
Conversely, a more than normal passive range of
motion (for instance, when a pathologically increased amplitude of external rotation is present
because of a subscapularis rupture) may result in a
false-positive external rotation lag sign. In addition, as previously discussed, neurologic deficits
can lead to folse-positive lag signs, as was the
case in four patients not in this series who had
complete suprascapular nerve palsies. All had a
positive ERLS and a positive drop sign.
On the basis of our findings we suggest the
following evolution of clinical findings in a primary
cuff rupture, progressing from a small to a large
rupture. At the stage of partial supraspinatus rupture the Jobe sign may be positive, mainly because

Hertel et al.

313

it enhances pain. The external rotation lag sign

becomes positive with as little as 5 ~ or less when a
transmural supraspinatus rupture has evolved. As
the rupture progresses, involving both the supraspinatus and infraspinatus, the external rotation lag increases from 5 ~ to 15 ~ or more. At this
stage the drop sign may be positive depending on
the compensation possibilities of the deltoid.
When the cranial part of the subscapularis tendon
starts to tear, the internal rotation lag sign becomes
positive. Eventually the lift-off sign becomes positive when the subscapularis has completely failed.
The authors conclude that the external and internal rotation lag signs and, to a lesser degree, the
drop sign, were helpful in the clinical evaluation of
rotator cuff rupture and were superior to the Jobe
and lift-off signs. Partial ruptures of the supraspinatus may remain concealed to the external rotation
lag sign, whereas partial ruptures of the subscapularis may be confidently diagnosed by the internal
rotation lag sign.
The authors acknowledge the help of the Mrs. Shu Fan
Hsu and Mrs. M. Hostettler from the Institute for Mathematical Statistics, University of Berne, Switzerland, in
the preparation of the statistical analysis of the results of
this study.
REFERENCES

1. Colton T. Statistics in medicine. Boston: Little, Brown and

Company, ] 974:164-8.
2. Gerber C, Krushell RJ. Isolated tears of the subscapularis
muscle. Cbinicalfeaturesin sixteencases.J BoneJoint Surg Br
1991 ;73B:399-94.
3. Howell SM, ImoberstegAM, SegerDH, Marone PJ.Clarification of the roleof the supraspinatusmusclein shoulderfunction.
J BoneJoint SurgAm 1986;68A:398404.
4. JobeI-W, MoynesDR. Delineationof diagnosticcriteriaand a
rehabilitationprogramfor rotatorcuff injuries.AmJ SportsMed
1982;10:336-9.
5. Kronberg M, N6meth G, Br6stromL,~. Muscle activity and
coordination in the normalshoulder.Clin Orthop 1990;257:
76-85.
6. Neer CS. Anterioracromiopiastyfor the chronic impingement
syndrome in the shoulder.J BoneJoint Surg Am 1985;51A:
41-50.
7. Otis JC, Jiang C-C, Wickiewicz TL, et ak Changes in the
moment arms of the rotator cuff and deltoid muscleswith
abduction and rotation.J Bonejoint SurgAm 1994;76A:66776.
8. Patte D, Goutallier D, Monpierre H, DebeyreJ. Etude des
16sions6tendues.Rev Chit Orthop 1988;744:314-8.