You are on page 1of 110



Centre Livet, Department of Orthopaedic Surgery - F-69300 Caluire

We owe our knowledge in this field to Henri Dejour. Our grateful thanks are due to Maîtrise Orthopédique for providing us with
the opportunity to translate what has so far been an oral tradition into a written record.

This review article is intended to familiarize orthopaedic surgeons with the methodology required for a systematic and logical
examination of the knee joint.

° The knee needs to be examined systematically, since so much can go wrong with the knee, and so many signs and symptoms may
be produced, that only a systematic technique will ensure that nothing of importance is missed.

° Equally, a logical approach is required, since the best way to remember a technique is to know why it is being applied. We will,
therefore, attempt to explain what the various signs and symptoms mean in terms of knee pathology or pathophysiology.

The detailed examination of the knee will need to be preceded by taking the patient's history. This aspect of the clinical approach is
vital, since it will often be found that a properly taken history will permit at least a presumptive diagnosis to be made. In fact, we
might go so far as to say that good history-taking is "non-invasive arthroscopy."

This article also presents the IKDC (International Knee Documentation Committee) Knee Evaluation Form, under the various
headings, in order to allow the more junior surgeons to familiarize themselves with international rating systems. The IKDC form
itself was explained by P. Chistel, in the Revue de Chirurgie Orthopédique (1993, 79:473). It is used mainly in anterior cruciate
ligament (ACL) surgery, to facilitate comparisons of results obtained in different studies of knee ligament trauma, in Europe and in
the United States.


After establishing the nature of the principal complaint (or reason for consulting a doctor), a systematic history should be obtained, with
questions grouped under three headings: History of present complaint; signs and symptoms; and life style (level of activity).


As a rule, three questions should suffice to obtain all the information required under this heading. The first two concern the onset of the
present complaint.

1.1 When?
This first question allows the examiner to distinguish between two patterns - complaints with an identifiable starting point, a known
accident (traumatic or post-traumatic knee disorders); and complaints that appear to have come on more insidiously, which would be more
suggestive of inflammatory or degenerative disease.

1.2 How?
The examiner should try to elicit the exact circumstances of how the disorder occurred - obtaining a description of how the accident
happened or of how the problem has developed over time. Thus, in a trauma case, the examiner would seek to find out the nature of the
contact (violent/non-violent; in valgus; in varus; in hyperextension) or whether the knee "went" after kicking a ball or "missing a kick." It
would also be important to establish whether there was an audible "pop" in the joint, whether the knee swelled rapidly, and whether the
athlete needed help to come off the pitch: - An affirmative answer to these questions would be highly suggestive of a torn ACL (Figs. 1-5).
Figure 2 Figure 3

Figure 4 Figure 1 Figure 5

1.3 What happened after that?

This question needs to be asked in order to obtain information on any medical or surgical treatments already applied, on rehabilitation (if
any), and on the course of the condition up to the time of consultation.


A complete picture of the patient's signs and symptoms should be obtained, since anything elicited under this heading may be of diagnostic

There are four categories of cardinal signs and symptoms in the knee joint:-

2.1 Pain
The examiner should establish the way in which the pain developed, its character, and its severity; the patient should be asked to point to
the site of the pain.

Pain at night suggests an inflammatory cause, while pain that gets worse towards evening, or during/after exercise, would be more likely to
be mechanical in origin.

Pain when going up or down stairs, or aching in positions where the knee is kept flexed for prolonged periods of time (car journeys, visits
to the cinema), are indicative of patellar problems, while pain that occurs when the knee is hyperflexed is usually caused by meniscal

The patient should also be questioned about pain in other parts of the body (low back pain, hip pain).

Bar- or vice-like pain below the patella is highly suggestive of a low-riding patella.

2.2 Laxity
A certain amount of confusion has occurred in the international literature, since the French use the term instabilité to denote what in
English would be called laxity, and speak of laxité in the sense of the English term instability.

° "Going out": This is the term used by many lay persons to describe what will usually be found to be a torn ACL or a dislocation of the

° "Giving way": This term is used to describe the sensation of the knee suddenly failing to provide proper support, especially when
walking on uneven ground. The symptom may be due to three mechanisms:
- Interposition: If, during weight-bearing, a third structure (meniscus, synovial membrane, cartilage, etc.) is placed between the opposing
cartilage surfaces of the joint, a protective reflex will be triggered. This reflex will make the quadriceps relax and unlock the knee, to allow
the joint to clear itself.

- Cartilage damage: If one or both of the cartilage surfaces are damaged, and the surfaces come into contact, the quadriceps may also be
made to relax.

- Muscle weakness: This may occur in quadriceps wasting, in polio, after surgery, etc.

2.3 Locking
A proper knowledge of this clinical feature is vital, since patients have often been misdiagnosed because of the examining physician's
imperfect understanding of this symptom. There are two types of locking, which must be carefully distinguished:

° Meniscal locking (true locking): This is what a physician would consider to be locking. It is the impossibility fully to extend the knee
for an appreciable period of time (more than a few minutes). This "passive flexion deformity" is brought on by a mechanical obstacle which
makes the knee stop short of full extension (Fig. 6). The cause may be a bucket-handle tear of the meniscus, or a bulky flap that has
dislocated forwards in the joint; a loose body or an ACL stump may also be to blame.

° Patellar catching (false locking): This is what the patient would consider to be locking. It is a momentary "sticking" of the knee, during
a flexion-extension movement, with the knee incapable of flexing or extending beyond that particular point. Catching is relieved as soon as
weight is transferred to the other side (Fig. 7). Usually, patellar cartilage damage will be found to have caused this fleeting episode of
Figure 6 Figure 7

2.4 Effusion
The knee swells up. Swelling of the knee is always indicative of a genuine lesion of the joint. Sometimes, its character will already have
been established by aspirating the joint. The features of the aspirated fluid (colour, viscosity, protein content, cellularity) make it possible to
ascertain whether the condition is mechanical or inflammatory; a search for microcrystals should always be made.

° Hydroarthrosis: The accumulation of clear, straw-coloured fluid is the result of irritation of the synovial membrane, which may be
primary (inflammatory disease) or secondary to cartilage damage (osteoarthritis), meniscal lesions, or the presence of a loose body
(osteochondritis dissecans; osteochondral fractures). Hydroarthrosis may also be seen following ligament lesions.

° Blood in the joint (haemarthrosis) without a history of trauma could mean two things: haemophilic arthropathy, or pigmented villonodular
synovitis. In the latter condition, the fluid may be amber-coloured rather than frankly bloody (Fig. 8).
Figure 8

2.5 Other clinical features

° Subjective sensation of internal derangement: The patient feels that there is something moving in the knee (joint mouse); something
"funny"; "a lump." This feature suggests a meniscal lesion or a loose body.

° Noises in the joint: Crepitus is a faintly audible and often palpable sensation of grating during flexion-extension. Clunks and clicks are
much louder, and suggestive of meniscal lesions.

For evaluation with the IKDC score, the patient will need to indicate the highest activity level that his or her knee will tolerate, even if (s)he
does not practise the activities at that level. The levels are as follows:-

I - Strenuous activity
(contact sports involving pivoting and cutting)

II - Moderate activity
(pivot sports without contact; manual work)

III - Light activity

(jogging, running)
IV - Sedentary activity

Patients who report symptoms at level I but not at level II are rated B = "nearly normal."

The lowest grade within a group determines whether the patient is group-graded A, B, C, or D in the last column in Table 1.

Table 1


History-taking also involves obtaining information on the patient's present and desired levels of activity. This information is important for
two reasons: Firstly, it gives an idea of the degree of disability produced by the knee disorder; and, secondly, it shows what the patient
would like to be able to do with his or her knee.

The questions to be asked will be a function of the patient's age:

° Young, active, or athletic patients should be questioned about the sports they practise, and about their ability to run, jump, or cut.

° Elderly or sedentary subjects should be asked about the use of walking aids, their walking distance, ability to go up and down stairs
without holding on to a handrail, and whether they can get up from a sitting position without using their hands.

The patient should be examined standing up, walking, and lying supine. It is essential that a comparison be made throughout with the
unaffected side.

1 - In the STANDING PATIENT, a study should be made of

° Lower limb pattern: The lower limb is said to be in normal alignment if, when seen head-on, with the patellae pointing forwards, the
medial malleoli and the femoral condyles touch. Genu varum describes a condition in which the femoral condyles are apart when the feet
are together, while in genu valgum the ankles are separated when the knees touch. The distances should be measured; they may be
expressed in centimetres (or in fingerbreadths), with the knees in extension (Figs. 9-10). With the knees in hyperextension, a varus
deformity may be seen to worsen. Genu varum and genu valgum are not, in themselves, abnormal conditions. This is why "normally
aligned" is not synonymous with "normal."

Figure 9 Figure 10

° Muscle wasting: With the patient standing "to attention",. the muscle bulk is checked. Quadriceps wasting shows up as wasting of the
vastus medialis. Looking at the vastus medialis will give an immediate idea of whether the quadriceps is wasted. Wasting may be the result
of under-use of the knee joint. It may be quantified by measuring the circumference of the thigh.

2 - In the WALKING PATIENT, the following features are checked:-

° The toeing angle is the angle between the axis of the foot and the direction in which the subject is walking. Normally, the axis will be
seen to point in a slightly lateral direction, enclosing an angle of 10° to 15°. In the normal postural pattern, this angle will be the same on
both sides.

° Tilting of the knee with single-leg stance: This is seen where bone wear has occurred as a result of osteoarthritis (OA). The feature does
not manifest itself until several years have elapsed. Where the tilt is due to ligamentous problems, it will be seen early on.

(a) With wear in the medial compartment, the tilt will chiefly be into varus, in a position of near-extension. It is best seen from behind the
patient (Fig. 11).

Figure 11

(b) With lateral compartment wear, on the other hand, the tilt will be into valgus and flexion; it is best appreciated from in front.

(c)A tilt into recurvatum is rare; it is not well tolerated by the patient.


3.1 General appearance

There are two cardinal features with which the examiner has to be familiar, since they are as important in the examination of the knee joint
as the temperature and pulse rate are in the general examination of a patient. They make it possible for a "spot diagnosis" of a genuine,
organic knee lesion to be made by the examiner.
(a) Looking for an effusion: The examiner's hands are placed on either side of the patella, with the thumb and middle to little fingers
stroking the synovial fluid towards the patella, while the index finger is used to elicit the patellar tap: The patella is at first pressed down
and submerged under the synovial fluid, and will strike the trochlea, producing a tap. As the pressure is relieved, the patella will bob up like
an ice cube in a drink (Fig. 12).

(b) Looking for a fixed flexion deformity: The patient is positioned supine and made to relax. The examiner grasps both the patient's
heels and supports them at a height of 10 cm above the examination couch. This is the best position for screening for a flexion deformity,
which is a major feature of knee pathology. This sensitive and straightforward method is ideal for screening purposes. It does not, however,
lend itself to a quantification (in degrees) of the deformity. Also, since the patient's feet are braced against the examiner's abdomen, the
examiner may seek to reduce the flexion deformity by pressing down on the patient's knees (Fig. 13).

Figure 12
Figure 13

- Knee pattern in the supine patient: As the patient goes from two-legged stance into the supine position, the deformity may or may not
correct itself. If a lower limb malalignment was found with the patient standing on both feet, the examiner should check whether the
deformity can be reduced in the supine patient, which would suggest that the deformity is articular (rather than extra-articular and bony) in

Another way of checking for a flexion deformity consists in positioning the patient prone, on a firm table, with the knees supported on the
table and the legs protruding beyond the table's edge. One heel is seen to be higher than the other. The distance between the two heels may
be measured. It provides direct evidence of the flexion deformity. The test is reproducible, and the result is numerical (in cm). However, it
requires a firm surface, and the patient must be positioned prone (Fig. 14).
Figure 14

In order to be able to tell whether a finding is abnormal, the affected side will have to be compared with the presumably healthy limb (the
reference knee). This comparison will show whether there is a fixed flexion deformity or a genu recurvatum.

- The range of movement (ROM) is tested at this stage of the examination (which also involves screening for quadriceps wasting). Three
figures are used to denote the ROM. The first indicates flexion; the second, full extension; and the third, recurvatum.

Thus, 140, 0, 5 means that flexion is 140°; the knee can be extended fully; and there is 5° of recurvatum. 120, 5, 0 would mean that flexion
is 120°, there is a flexion deformity of 5°; and no recurvatum.

For evaluation with the IKDC score, the passive ROM is recorded in both knees. Values are given in 3-figure form: Extension/Flexion /
Lack of extension (from 0°) / Lack of flexion (Fig. 15) (Table 2).
Figure 15

Table 2

3.2 Extensor apparatus

° Inspection
- The bayonet sign: This term describes the pattern produced by the patella, the patellar tendon, and the tibial tubercle. The line resembles
a bayonet fixed to a rifle.
Pathophysiology: The bayonet results from a lateral position of the tibial tubercle. It was thought for a long time that this was an organic
factor contributing to patellar instability. It is now known that it is the resultant of the forces produced by the lateral position of the tibial
tubercle that tends to pull the patella sideways.

Reliability: Not very good. A more reliable picture may be obtained from measuring the distance from the tibial tubercle to the inside of
the trochlea on CT scans.

- Squinting of the patella

° Palpation
- Looking for tender points is particularly useful if the patient has also reported symptoms of pain (Fig. 16).

* Tibial tubercle: In adolescents, the tibial tubercle may be affected by apophysitis, with fragmentation of the apophysis (the accessory
ossification centre). The patient will complain of pain in the tibial tubercle, and the examiner will be able to elicit tenderness over that
structure (Osgood-Schlatter disease).

* Patellar tendon: This structure may be the site of repetitive strain injuries (jumper's knee); the patient will complain of pain in the
patellar tendon, and extension against resistance will be painful.

* Apex of the patella: In adolescents, inflammation of the distal pole of the patella may occur as a very rare condition (Sinding-Larsen-
Johansson syndrome).

* Medial facet of the patella (Fig. 17a): The medial facet of the patella may be tender to palpation, and the patient may report pain at that
site. The feature is usually part of an anterior knee pain syndrome.

* Lateral facet of the patella (Fig. 17b): The examiner pushes the patella in a lateral direction, and palpates the lateral facet.
Patellofemoral dysplasia will produce lateral impingement and pain and tenderness along the lateral border.
Figure 16 Figure 17

* A search should also be made for a mediopatellar plica, which appears as a tender cord that rolls under the palpating finger, on the
medial femoral condyle. It sweeps along the condyle during flexion, and may give rise to pain symptoms.

* Patellar tilt: The examiner holds the edges of the patella between the thumb and index finger, thereby establishing the axis of the patella,
which should differ only slightly (10° lateral tilt) from the horizontal plane of the knee seen head-on.
The patella may be said to squint (convergent or divergent squint) (Fig. 18). Broadly speaking, a convergent squint tends to occur in
anterior knee pain syndrome, while a divergent squint would be more likely in recurrent dislocation.
Figure 18

° Patellar tests

- Apprehension sign
The patient is positioned supine, with the knee flexed between 0° and 30°. The examiner firmly pushes the patella in a lateral direction. The
patient, who knows and apprehends the dislocation that will be produced by this manoeuvre, will stop the examiner. Results are recorded as
+ or 0.

Pathophysiology: Between 0° and 30° of flexion, the patella is at its highest point in the trochlea. Pressure from the medial side will push
the patella in a lateral direction, causing it to dislocate from the trochlear groove. This will cause not only pain, but apprehension on the
part of the patient (Fig. 19).

This sign may be elicited in recurrent dislocation; it is highly suggestive, and particularly useful in patella alta. As Henri Dejour puts it,
"You can't get near their kneecaps." (Fig. 20)
Figure 19 Figure 20

The test must be stringently performed and analyzed.

- Patellar grind test

The examiner's hand is placed on the front of the knee. The patient performs flexion-extension. The examiner will feel a crepitus, and may
even notice the patella catching. The crepitus is difficult to interpret. If there is nothing more than a positive grind test, a diagnosis of OA or
of cartilage damage cannot be made.

Patellofemoral joint crepitus should be sought over the entire ROM from flexion to extension, against slight resistance. Crepitus in the
tibiofemoral compartments is sought from flexion to extension, against resistance, as well as in valgus-flexion-external rotation (to test the
lateral compartment) and in varus-flexion-internal rotation (to elicit medial compartment crepitus). Gradation depends on the loudness of
the crepitus and on the pain produced by the manoeuvre (Table 3).
Table 3

3.3 The menisci

Broadly speaking, the menisci should be examined with the knee in flexion. There must be tenderness (i.e. the patient must respond to
palpation with pain). There are various ways in which the sensitivity of the tests can be enhanced. However, all the tests for meniscal
lesions rely on the same principle: Stressing an injured medial or lateral meniscus will cause pain.

° Tenderness to palpation is elicited with the knee flexed 90° and the patient's foot resting on the table. The examiner's index finger
probes the meniscus along the joint line. The most frequently encountered sites of tenderness are over or behind the medial collateral
ligament, at the medial meniscal tender point (Fig. 21). Less often the tender point will be anterior, in which case the phenomenon may be
part of a patellar disorder, a bucket-handle tear of the medial meniscus, or a lesion of the anterior horn of the lateral meniscus. The lateral
meniscal tender point may be anywhere along the joint line (Fig. 22).
Figure 21

Figure 22

° Meniscal tenderness on mobilization: Compression of the different parts of the meniscus by the femoral condyle occurs as the meniscus
glides backwards on the condyle during flexion, and forwards during extension.
This means that the posterior horn will be compressed when the knee is in hyperflexion (Fig. 23); while the anterior horn will be
compressed in hyperextension.

The diagnostic accuracy of the tests is improved by adding a component of tibial rotation to the simple flexion-extension manoeuvres. This
rotation tends to bring the posterior horns forward. Medial pain will be elicited in external rotation, and lateral pain in internal rotation.

- McMurray's test: Forced flexion and external rotation with compression of the medial joint line will elicit pain in the medial meniscus.
The hand pressed over the joint line will feel a click. The test may be reversed, to examine the lateral meniscus.

- Apley's grinding test: For this test, the patient is positioned prone, with his or her knee flexed. Compression and external or internal
rotation may be painful, showing that the medial or the lateral meniscus are torn. This test is always checked, by performing rotation
without compression. This manoeuvre should not cause discomfort, unless the collateral ligaments are affected (Fig. 24).

Figure 23 Figure 24

- Cabot's manoeuvre: The heel is placed on the tibial crest of the opposite leg. The knee is gradually flexed, while the heel runs along the
tibial crest. This movement may produce lateral pain, when the knee is in 90° of flexion with the heel resting on the other leg (Cabot's
position). The lateral compartment is distracted by pressure on the medial side of the knee; this, too, may be painful.

° Cysts of the lateral meniscus will be seen in extension, and disappear in flexion. They are on or near the lateral joint line. They are best
seen in semiflexion. The cysts will disappear in hyperflexion, and reappear as the knee is gradually extended; in full extension, they will
once again be out of sight.

(In children, a malformation of the lateral meniscus may give rise to snapping when the knee is taken through flexion and extension. This
abnormal movement is associated with a clicking noise, which may be very pronounced.)

3.4 Stability testing

3.4.1. Medial/lateral instability in extension

(a) Medial instability in extension

The examiner grasps the patient's heel (not the ankle or the leg) with one hand, while the other hand is placed against the lateral aspect of
the patient's knee. A brisk valgus stress is imparted and immediately released. Medial instability is demonstrated if the medial joint line
opens up (Fig. 25). Sometimes the most characteristic phenomenon is a little click as the knee reduces after the stress test. Sometimes, it is
difficult to decide whether there is instability.

Figure 25

Several points need to be borne in mind:

° The abnormal feature is the asymmetrical pattern of the instability. The examiner may also ask the patient, "Do you think your right knee
is different from your left knee?"

° The leg may be supported halfway between the knee and the ankle, by pressing it against the examiner's body. That way, a greater valgus
thrust may be applied. The third point of support produces better leverage.

° The instability will be due to a lesion of the ligaments on the medial side and/or to medial tibiofemoral compartment wear.

(b) Lateral instability in extension

The examiner grasps the patient's heel with one hand, while exerting pressure against the inside of the knee with the other hand. The varus
stress applied will cause lateral gaping in the laterally unstable knee. Lateral joint gaping is physiological. It is the asymmetry of the gaping
that constitutes the abnormal finding.

3.4.2. Medial/lateral instability in 30° flexion

Description: The leg is held as described above, but the knee is unlocked by putting it in 20°-30° flexion.

(a) Medial instability

The movement imparted is one of valgus-flexion rather than valgus-flexion-external rotation. Instability in valgus-flexion-external rotation
would be a sign of an injured medial collateral ligament.

(b) Lateral instability

Varus-flexion-internal rotation is used to investigate the lateral collateral ligaments. Once again, asymmetry would have to be demonstrated
to qualify the result as abnormal (Fig. 26).
Figure 26

In the United States, examiners prefer to sit on the couch, between the patient's knees, for the performance of this test (Fig. 27).

The lateral collateral ligament is palpated with the knee in Cabot's position (see above), where it will be felt as a tense band. It is possible to
explore both knees simultaneously, with the patient in the "frog position" described by Henri Dejour (Fig. 28).
Figure 27 Figure 28

3.4.3. Anterior instability

(a) Lachman-Trillat test

It is important to ensure that the patient is relaxed. This is all the more vital in recent trauma cases. In order to obtain relaxation, the patient
is made to rest his or her head on the couch. It may be useful to roll the thigh in and out, to get the muscles to relax (Fig. 29).

For the test, the knee is unlocked in 20° flexion. The patient's heel rests on the couch. The examiner holds the patient's tibia, with the thumb
on the tibial tubercle. The examiner's other hand is placed on the patient's thigh, a few centimetres above the patella. The hand on the tibia
applies a brisk anteriorly directed force to the tibia (Fig. 30).
Figure 29 Figure 30

The quality of the endpoint at the end of the movement is described as either "firm" or "soft." Grading depends on the quality of the
endpoint observed, and on whether there is a difference of 3-5 mm between the affected and the unaffected knee. A soft endpoint will make
the grading "abnormal" rather than "nearly normal."

If the movement of the tibia on the femur comes to a sudden stop, this is described as a firm endpoint. If it does not, the endpoint is
described as soft. A soft endpoint is pathognomonic of a torn ACL. It is easier to demonstrate a firm endpoint, which is also recognized by
the patient. If the ACL is torn in one knee, the patient will be perfectly aware of the difference between the firm endpoint in the healthy, and
the soft endpoint in the cruciate-deficient knee. Sometimes, the endpoint will be firm, but translation will be seen to be asymmetrical. This
is known as a "delayed firm endpoint" (increased excursion and a good endpoint). It is indicative of a torn and partially healed ACL (ACL
adherent to PCL), of a stretched ACL graft, or a torn PCL (changing the "starting point" of the test). A firm endpoint results from the
sudden tensioning of the ACL (Fig. 31). The test is of lesser value in knees affected by OA, with a large number of osteophytes.
Figure 31

Lachman used to perform the test with the thumb of the distal hand on the medial joint line, so as to feel the displacement of the tibia on the

In Trillat's modification of the test, the thumb is placed on the tibial tubercle, so as to get visual evidence of the translation. In our practice,
we prefer Trillat's method.

The Lachman-Trillat test, the drawer test in 70° flexion, and the medial and lateral joint opening tests may be performed manually or using
an arthrometer (KT-1000, KT-2000) or stress radiography. If an arthrometer is used, a force of 134 N (30 lbs.) will need to be applied.
Measurements are performed on both sides. The difference between the affected side and the opposite side is recorded. Usually, only one
value is recorded. The absolute value is also of interest, and should be recorded for prospective study purposes.

If, after ACL surgery, the operated side is found to be tighter than the healthy side, the graft will be at increased risk of failure.

Active resisted extension

In obese patients or subjects with bulky muscles, it may be difficult for the examiner to encircle the patient's thigh with his or her hand. In
such cases, the examiner may place a fist under the knee, hold the ankle against the couch with the other hand, and ask the patient to lift the
leg against resistance (Fig. 31b). This resisted quad setting will move the tibial tubercle forward. This is a useful screening test, which will
not, however, be positive unless there is major instability. It would be preferable to obtain stress radiographs (134 N stress) on which to
base the diagnosis.
Figure 31b

(c) Pivot shifts

Tests screening for pivot shift were first described by M. Lemaire, in 1968. Since then, many such tests have been devised, of which we
shall only list the main ones. According to Noyes, the phenomenon is potentiated when the hip is abducted. At present, attempts are being
made to substantiate the diagnostic value of these tests. However, they should not be expected to provide more than they can: A shift means
that the ACL has gone. Sometimes, though, the ACL may be deficient without a pivot shift occurring during the relevant tests.

Screening for pivot shift must be done systematically, using the customary techniques. The IKDC form records only the greatest shift
found. The scoring system is conventional: + = glide; ++ = clunk; +++ = gross.

- Pivot shift in extension (Dejour's test) (1978)

(1) The patient's foot is wedged between the body and the elbow of the examiner. The examiner places one hand flat under the patient's
tibia, pushing it forwards (force applied in an anterior direction), with the knee in extension. The other hand is placed against the patient's
thigh, pushing the other way (force applied in a posterior direction) (Fig. 32).
Figure 32

(2) The lower limb is taken into slight abduction, by the examiner's elbow, with the examiner's body acting as a fulcrum to produce the

(3) The examiner maintains the anterior tibial translation and the valgus, and imparts flexion.

At 20°-30° flexion, pivot shifting will occur, with a clunk as the lateral tibial plateau suddenly reduces (Fig. 33).
Figure 33

Significance: The valgus stress associated with the anterior tibial drawer makes the lateral tibial plateau sublux on the lateral femoral
condyle and compresses the structures. The sudden reduction of the convex lateral tibial plateau compressed under the lateral condyle is
responsible for the clunk. Sometimes, a clunk may be elicited with compression, rather than any major valgus stress. The pivot shift is easy
to produce, and causes no discomfort. It is a mixture of shift and Lachman, and provides evidence of ACL tears and damage to
posteromedial structures.

- The glide pivot shift test described by Henri Dejour is the equivalent of Noyes' test. It produces a glide (a minor form of shift) rather
than a proper clunk. The patient is unaware of the slip of the plateaus on the femoral condyles. The valgus component is less pronounced
than are the compression and anterior drawer applied by the examiner. It produces joint play rather than a pivot shift. It can be seen in
patients with a torn and partially healed ACL, as well as after ACL grafts or other ACL surgery (Fig. 34).
Figure 34

- The pivot shift test of MacIntosh

"When I pivot, my knee shifts." This is how a hockey player described his symptoms - hence the name of the test.

MacIntosh realized that the sensation of shifting or slipping was related to rupture of the ACL.

He devised a test to reproduce the sensation reported by patients, involving stress applied to the knee in valgus and flexion, with or without
internal rotation (Fig. 35).
Figure 35

Description of the test: The patient is positioned supine, with the examiner standing on the affected side. The examiner uses one hand to
hold the patient's foot in very slight internal rotation. With the other hand, (s)he applies a valgus stress to the posterolateral aspect of the
proximal calf. At this point, flexion is started. The lateral tibial plateau will be seen to sublux forwards during the first degrees of flexion.
As flexion progresses, the anterolaterally subluxed tibia will suddenly reduce, at 30° of flexion. This reduction is associated with a
characteristic clunk, which the patient will readily recognize.

- Hughston's jerk test

The patient is positioned supine, with the examiner holding the lower limb in such a way as to have the hip flexed 45°, the knee flexed 90°,
and the leg in internal rotation. The distal hand grasps the foot and places it in internal rotation, while the left hand applies valgus stress to
the upper end of the leg.

A jerk is defined as a sudden change in the relationship between the joint surfaces. The phenomenon occurs as the subluxation reduces near
full extension.

Hughston thought that rotational phenomena were more important (hence his defence of the concept of rotary instability).

- Slocum's ALRI test (1976)

The patient is placed in a semilateral position, with the knee unsupported, and the lower limb resting only on its heel. This allows internal
rotation of the foot, to produce translation of the lateral tibial plateau. The examiner stands behind the patient's back, with the distal hand
holding the upper end of the leg, and the proximal hand around the lower end of the thigh. The test is started in extension, using vertical
pressure. Next, flexion is commenced. Translation occurs at 20°, and the pivot shifts at 40° (Fig. 36).

Figure 36

- Losee's test (1969)

Over the years, Losee has described five tests: The first is identical with the MacIntosh test; the second uses external rotation of the tibia to
elicit the pivot shift; the third (deceleration test) uses the quadriceps slingshot effect; while the fourth and the fifth test the anti-pivot shift
effect of extra-articular ligament reconstructions.

(c) Anterior drawer in 90° flexion, or direct anterior drawer

The examiner sits on the patient's foot, which has been placed in neutral position. The knee is in 90° flexion. The index fingers are used to
check that the hamstrings are relaxed, while the other fingers encircle the upper end of the tibia and push the tibia forwards (Fig. 37).
Figure 37

If a direct anterior drawer is obtained, the ACL will be torn. However, for this sign to be elicited, peripheral structures such as the medial
meniscus or the meniscotibial ligament must also be damaged. This ligament forms a wedge, in 90° flexion, preventing anterior tibial
translation. The finding of an anterior drawer is conclusive evidence of an ACL tear. However, not every ACL tear will be associated with a
positive anterior drawer test.

° Drawer in external rotation (foot in external rotation). This test examines the posteromedial structures (posteromedial corner, posterior
horn of medial meniscus). The results are recorded as +++/0.

° Drawer in internal rotation (foot in internal rotation). The diagnostic value of this test is less well established.
Table 4

° One-leg hop functional test (Dale Daniel)

This test investigates the greatest distance the patient can hop on one leg.

The test is performed three times, and the mean value achieved is recorded. It is a comparative test, the result of which is expressed as a
percentage of the value achieved on the healthy side.

Table 5

3.4.4. Posterior instability

- Posterior tibial sag in 90° flexion

In the IKDC form, the posterior tibial sag is recorded in 70° flexion; a comparison is made with the opposite side, looking sideways at the
knees to establish the amount by which the tibial tubercle has dropped backwards. Posterior tibial translation may be confirmed if it is seen
that quadriceps contraction pulls the tibia forward.

- Straight posterior drawer

With the patient supine, the test is performed with the knee in 70°-90° flexion and the foot in neutral rotation. The examiner sits on the
patient's foot, placing both thumbs on the tibial tubercle, and pushing the tibia backwards. A positive test (i.e. one in which the tibial
plateau moves in a posterior direction) means that the PCL is torn.

Oddly enough, this is a difficult test to perform, since the patient's tibia may often sag spontaneously, requiring this posterior translation to
be reduced before the test is started, since otherwise the result may be erroneously given as an anterior drawer. The quality of the endpoint
(firm or soft) is of no significance in the posterior drawer test.

To check the knee pattern, the knees should be inspected side-on, to see whether the tibial tubercle has sagged backwards. The knees should
be in 90° flexion, with the patient's feet resting on the couch. It is preferable to use Godfrey's drop back test.

There are three tests that can be used to demonstrate a posterior drawer, and which may be resorted to in case of doubt:

- Godfrey's drop back test: The patient is supine, with the thighs and knees flexed 90°, legs horizontal, and heels held by the examiner in
such a way as to have the legs parallel to the table. The test is positive if the upper end of the tibia on the affected side is seen to drop
backwards (Fig. 38).

Figure 38
- Muller's test: Patient positioning is identical to that for the posterior drawer test in 90° flexion. Tibial sag is observed. The patient is
asked to set his or her quadriceps: Before the heel will have had time to lift off the couch, the posterior displacement of the tibia will have

- Active extension against resistance: This is the same as Muller's test, but in 20° of flexion. This test is of lesser practical value.

- Posterior drawer in external rotation

The foot is placed in external rotation. This test is equivalent to the straight posterior drawer test in patients with PCL deficiency. The result
is more pronounced in cases of posterolateral lesions.

- Posterior drawer in internal rotation

The foot is placed in internal rotation. Translation is usually 4 mm less than in the straight posterior drawer test. If the same value is
obtained, this would be indicative of a lesion of the medial meniscofemoral or even the medial collateral ligament.

- Whipple's test: In order to rule out tibial translation as a result of gravitational sag, the patient should preferably be examined prone.
"This test is difficult. It has not yet gained wide acceptance." (B. Moyen). The information provided is the same as that obtained with the
conventional posterior drawer test performed with the patient supine, feet resting on the couch. Apart from the fact that sag is eliminated,
this test is completely unconstrained.

- Posterior translation in 20° flexion: Posterior displacement may be observed with the knee in 20° of flexion. If the result is the same as
that of the straight posterior drawer test, there will be associated posterolateral lesions.

3.4.5. Posterolateral instability

(a) External rotation recurvatum test

Description: The patient is supine. The examiner stands at the foot of the couch, and grasps the patient's big toes, lifting the feet off the
couch. The affected knee will go into varum-plus-recurvatum (Fig. 39).
Figure 39

According to Hughston, this test provides information on major disabling peripheral instabilities; it cannot be positive unless the PCL is
also torn: A positive recurvatum test shows that at least one cruciate is torn and that there is a posterolateral lesion. Of the cruciates, the
ACL is more often involved than the PCL.

(b) Jakob test (reverse pivot shift test)

The patient lies supine. The examiner stands at the patient's feet, and places his or her distal hand on the patient's ankle, with the patient's
leg braced against the examiner's pelvis. The proximal hand supports the upper third of the calf on the lateral side, and imparts a valgus
force so as to compress the lateral compartment. The lateral tibial plateau will drop back under its own weight, with the foot in external

This test starts with the knee in flexion. The joint is gradually taken into extension by its own weight.

At a given point, the subluxation will reduce with a snap, and the foot will go into neutral rotation. The test involves a pivot shift
manoeuvre, but since it starts from a subluxed position of the lateral tibial plateau, with reduction in extension, it is known as a reverse test.

The jolt is produced by the reduction of the posterior subluxation through the action of the lateral head of the gastrocnemius, the capsule,
and the pull of the iliotibial band, which, past 40° of flexion, moves from a flexor to an extensor function. All these factors will tend to pull
the tibial plateau forward, out of its posteriorly subluxed position. Since there cannot be any rolling-sliding, the plateau will snap back into
Significance: This test provides evidence of a posterolateral instability which may or may not be associated with a lesion of the PCL. The
test is not very specific.

Posterior subluxation of the lateral tibial plateau may occur without any PCL deficiency, since, in external rotation, the PCL is relaxed and
will allow a certain amount of posterior displacement of the tibial plateau. However, where the PCL is torn, the pivot shift will be very
much more pronounced.

(c) Increased external rotation

It is indicative of lateral lesions. The test is performed as a side-to-side comparison:-

° External rotation in 20° flexion: The examiner stands at the foot of the couch, and looks for unequal rotation. The difference found may
be expressed in degrees (Fig. 40).

Figure 40

° External rotation in 90° flexion.

(d) Lateral hypermobility

This test has been described by Gilles Bousquet. It is performed with the knee flexed at 90°, with the examiner's hands around the top of
the tibia (tibial tubercle), applying external rotation (Fig. 41). A positive test shows that the posterolateral structures have been injured.

To complete the test, the examiner should

° Look for stiffness of the hamstrings,. With the hip flexed at 90°, the popliteal angle is measured (Fig. 42). The examiner tries to take the
knee into extension, maintaining the foot in dorsiflexion.

Figure 41 Figure 42

° Look for stiffness of the rectus femoris. For this, the patient is positioned prone. The heel is brought towards the buttock, with the hip in
extension. The heel-to-buttock distance is measured (Fig. 43). While the patient is supine, a check should also be made for popliteal
(Baker's) cysts. This test should always form part of the general examination of a knee patient; in particular, it must always be borne in
mind that knee pain may be due to problems in the hip or the spine.
Figure 43


The following techniques are available for the further workup of knee patients:

1 - Scout radiographs

A.p. single-leg stance; lateral in 30° of flexion; axial views in 30° of flexion.

Instability should be investigated using the radiological Lachman test. For the instrumented test, a Telos device may be used (varus,
medial/lateral instability, a.p. instability).

Long films may be used to measure limb length and study the axes of the lower limb.

Rosenberg views (p.a. weight-bearing views in flexion) may be used to detect incipient OA or to screen for osteochondritis dissecans of the
medial femoral condyle.

2 - Other imaging techniques

° Arthrography: This is a useful technique in peripheral meniscal detachment, or when there is a suspicion of recurrent meniscal lesions.

° CT is essential in cases of patellar instability and of bone disease (tumour, trauma).

° MRI shows meniscal lesions, injuries of the cruciates, villonodular synovitis, and necrosis of the femoral condyles.

° Radionuclide bone scans should be requested if it is thought that the patient may be suffering from a tumour, from incipient avascular
necrosis of a femoral condyle, or from reflex sympathetic dystrophy; or if an infection is suspected.

3 - Blood tests

Sed rate, blood count, CRP, and rheumatological tests (latex fixation, Rose-Waaler, complement, ANA, antimitochondrial antibodies, Lyme
serology, etc.) should be ordered in inflammatory knee conditions.

5 - Arthroscopy and biopsy

These modalities are rarely indicated in the diagnostic workup of knee patients. They may occasionally be used in inflammatory conditions.

The various signs observed may be grouped together to distinguish a number of knee disorder patterns:-

- Complete isolated anterior instability = Lachman-Trillat +; pivot shift +

- Incomplete isolated anterior instability = Lachman-Trillat: Delayed firm endpoint +; glide pivot shift +

- Advanced anterior instability = Lachman-Trillat +; pivot shift +, straight anterior drawer +

- Posterior instability = Lachman-Trillat: Delayed firm endpoint +; posterior drawer +, tibial sag +, pivot shift 0

- Organic patellar instability = Apprehension +

- Anterior knee pain syndrome = Medial facet tenderness + (quad/hamstring stiffness ±)

- Medial meniscus lesions = Effusion +, medial joint line tenderness +, McMurray +, Grinding +

- Lateral meniscus lesions = Effusion +, lateral joint line tenderness +, Grinding +, Cabot +
- Isolated peripheral lesions:
Medial collateral ligament = Instability in valgus-flexion-external rotation

- Associated peripheral lesions:

Medial triad (ACL + MCL + IM ±)
Lachman-Trillat +, direct anterior drawer +, valgus-flexion-external rotation +,
instability in extension-valgus +

- Posterolateral lesions: Posterior drawer +, reverse pivot shift +

- Pentads:
Posterior drawer +, straight anterior drawer +, Lachman-Trillat +, pivot shift +, external rotation recurvatum +
(in lateral pentad);
Medial instability in extension, valgus-flexion-external rotation, varus-flexion-internal rotation +
(in medial pentad)

PS: The authors would appreciate hearing from other colleagues, in order to expand and augment the present
list of tests. Please send your suggestions, descriptions of tests, etc. to
Prof. Philippe Neyret
Centre Livet
8, rue de Margnolles
F-69300 Caluire
Pathophysiology and clinical examination «The third parameter»
Department of Orthopaedics, Bicêtre University Hospital

Our knowledge of the mechanisms underlying instability has been greatly enhanced in recent years, mainly as a result of the
consideration of a new parameter - ligamentous laxity.

In the work-up of instability, the clinical examination is of paramount importance: in the overwhelming majority of the cases, it will
allow a diagnosis to be made; further investigations will only provide confirmation. The clinical examination will allow the physician
to recognize an associated inherent multidirectional laxity.(10,20,34) In some, very rare cases, examination under general
anaesthesia will be required. Sometimes, the definitive diagnosis can only be made in the light of the arthroscopic findings


Five anatomical features of the shoulder are crucial to an understanding of the problem:
1) The shoulder joint is not congruous.
2) There are no discrete ligaments.
3) No one ligamentous structure is taut in all positions of the joint.
4) All the ligaments of the shoulder are slack in the resting position.
5) There is a weak point.

The ligaments of the glenohumeral joint are not discrete structures. The joint is surrounded by a sheet of fibrous tissue, which
constitutes both the capsule and the ligaments. This sheet inserts all around the glenoid, except for the rotator interval, and on the
humerus. Back in the 19th century, Schlemm described the structure as the "broad ligament of the shoulder."(30) In this paper, the
term inferior glenohumeral ligament (IGHL) will be used to denote the ligamentous structure inserted on the glenoid between 2
o'clock and 6 o'clock, which is the zone involved in instability. This zone includes the middle glenohumeral ligament and the anterior
band of the inferior glenohumeral ligament, as described in anatomical studies of the shoulder.(7,22)

The shape of this capsuloligamentous apparatus changes with the position of the glenohumeral joint; this is why the physiology of
this structure is much more difficult to understand than is that of a linear ligament that extends between two "point" attachments.
The kinematic analysis of the joint is made difficult by the fact that no one ligament is taut in all positions. This fact also accounts
for the difficulty encountered in trying to devise a simple and specific ligament test that would show abnormal ligamentous laxity; in
other words, there is nothing in the examination of the shoulder that would equate to the Lachman test, or to varus/valgus stress
testing, of the knee.

The relaxed condition of the ligaments at rest was demonstrated by Kumar and Balasubramaniam,(19) who showed that puncturing
the capsule resulted in a downward displacement of the humeral head by more than 1 cm.

When the arm is in 90° of abduction and in external rotation, the subscapularis muscle is pulled cranially, leaving the head of the
humerus covered anteroinferiorly only by the IGHL, without any other muscle providing a check to anterior displacement.(33) This is
the weak point through which the head may dislocate.

Mechanisms of glenohumeral joint stabilization

There are many and complex ways in which the glenohumeral joint is stabilized.

The fibrous labrum adds some depth to the glenoid socket; however, its role in the stabilization of the joint is very limited.

The rotator cuff used to be considered the only significant stabilizer of the shoulder. Recent research (7,8,22,23) has redressed the
balance by showing the importance of the ligaments. Between the resting position (with the arm hanging at the side) and maximum
scapular-plane abduction, the articular ligaments go from complete relaxation to a state of uniform tension. Thus, it could be said
that the stabilization of the joint gradually increases throughout abduction. The rotator cuff, in turn, has a very important role to
play at the start of elevation, becoming progressively less important as the movement continues.

This discrete involvement of the two anatomical structures is very clearly reflected in the clinical pattern: shoulder instability is
virtually only seen at the extremes of abduction and external rotation, i.e. when the ligaments have come into play. Conversely,
destabilization of the humeral head in rotator cuff lesions occurs chiefly at the start of abduction. This instability can be
demonstrated by Leclercq's resisted-abduction manoeuvre.

The vacuum phenomenon described by Kumar & Balasubramaniam (19) and by Habermeyer et al (13) is virtually unique in the
human body. In the cadaver shoulder, all the muscles stabilizing the glenohumeral joint may be cut without affecting the position of
the humeral head within the glenoid, providing that the joint capsule has not been breached,. As soon as the capsule is punctured,
the head will translate downwards by more than 1 cm. This shows that the shoulder depends to a considerable extent on negative
pressure for the maintenance of the humeral head in the glenoid fossa. It is difficult to assess the importance of this mechanism in
the prevention of shoulder instability.



Findings at surgery

In order to understand the mechanisms of instability, it is useful to consider the

main findings at surgery using the Bankart technique, i.e. with separation of the
subscapularis from the capsule, to provide a detailed and comprehensive picture
of the ligamentous lesions.

In the overwhelming majority of cases, the ligamentous apparatus will be

continuous, and inserted on, or at a distance of less than 1 cm medial to, the
glenoid rim. The labral lesion1 will vary from a complete disappearance of the
structure and bone wear, through greater or lesser detachment, to discreet
surface damage. In some patients, there will be evidence of a glenoid rim

These findings allow several conclusions to be drawn:

a) a Broca-Hartmann pouch, in the strict sense of the term, is extremely rare;

b) the initial ligamentous lesion will heal, regardless of its shape and site;
c) lesions of the anterior rim vary greatly in size. Their significance in the
production of chronic instability is difficult to assess, since the movement of the
humeral head forward of the glenoid rim during repeated episodes of instability
must progressively change and aggravate the lesional pattern.

Fig. 1a-c :
1a Dislocation following a direct blow to the back of the shoulder
A "historical misunderstanding"
1b Formation of a Broca-Hartmann pouch. The persistence of In France, the mechanism underlying recurrent shoulder dislocation is still
this pouch is thought to be the main factor responsible for explained in terms of the formation of a Broca-Hartmann pouch.4 Broca and
recurrent dislocation.
1c Mechanism of indirect dislocation
Hartmann published the first description of the stripping of the capsule and the periosteum from the anterior aspect of the scapula,
in 1889. The persistence of this detachment leaves an anterior pouch, which accounts for recurrent dislocation: the humeral head
will "drop" into the pouch (Fig. 1b). That publication attracted general attention. However, the authors' second paper,3 published one
month later, went unnoticed. In that paper, the authors stated that the lesion described previously was produced by a direct blow
from behind (Fig. 1a), and had nothing to do with the very much more frequently observed so-called indirect dislocations (Fig. 1c).
Thus, the Broca-Hartmann pouch should be seen for what it is: a very rare injury mechanism, which cannot account for recurrent
anterior dislocation. (The rarity of the lesion is shown in our clinical material, where, in a consecutive series of 100 cases, there was
only one pouch.)

Experimental studies

Many papers published in the international literature over the past few years have allowed us to formulate a more coherent concept
of how chronic instability of the shoulder is produced.

In order to dislocate a shoulder, at least the anteroinferior portion of the capsuloligamentous apparatus must be divided. (In the
right shoulder, the division should extend from 3 o'clock to 6 o'clock.) The division of the capsuloligamentous apparatus is not
enough to cause dislocation.9,33 The rotator cuff must also be damaged. Some authors have obtained dislocation following division
of the supraspinatus tendon. We have shown that the lesion required may consist in a partial detachment of the deep portion of the
cuff insertion. The anti-dislocation role of the rotator cuff can also be shown the other way round, by looking at the rate of cuff
lesions found in shoulder dislocation in patients over the age of 40 years.21 It is thus clear that the rotator cuff provides a second
line of defence to protect the shoulder from dislocating. The experimental creation of an isolated Bankart lesion32 has also been
found to be insufficient for the production of shoulder instability.

La lésion de Bankart
Il est actuellement couramment admis que la lésion typique de l'instabilité chronique de l'épaule est la lésion de Bankart. Elle
associe une lésion du bourrelet qui peut être usé, désinséré, voire presque entièrement détruit, à un décollement capsulo-périosté
antérieur d'importance variable. Nous avons vu que le décollement antérieur était le plus souvent très limité. Il paraît douteux que la
lésion du bourrelet entraîne une perte importante de stabilité. De plus on sait maintenant que les lésions antérieures s'aggravent
progressivement avec le nombre de récidives (12) ce qui peut laisser un doute quant à l'importance des lésions initiales et fait que
la question est posée : la lésion de Bankart suffit elle à expliquer l'instabilité chronique ?

The Bankart lesion

The Bankart lesion is nowadays accepted as the typical pattern in chronic instability of the shoulder. This lesion combines damage to
the glenoid labrum, which may be worn, avulsed, or almost entirely destroyed, and anterior soft-tissue stripping, of varying extent.
We have found this anterior detachment to be usually very limited. It is doubtful that the labral lesion will lead to a major loss of
stability. It is also now known that anterior lesions tend to get worse with recurrences.(12) Thus, the importance of the initial lesion
may not be as great as hitherto assumed; and the real question is whether the Bankart lesion as such can account for chronic
The Hill-Sachs lesion
Unlike the reverse Hill-Sachs lesion, which plays a major role in recurrent posterior dislocation, the implication of the Hill-Sachs
lesion15 (known in French as the Malgaigne notch) in the generation of recurrent anterior instability has never been conclusively

Laxity of the inferior glenohumeral ligament (IGHL) complex

Over the years, several publications have stressed the importance of ligamentous laxity in shoulder instability. Neer and Foster(20)
showed that some subjects have extremely hypermobile shoulders, which could cause multidirectional instability, in a context of
inherent, non-trauma-related laxity of the shoulder capsule. Other authors have stressed the presence of ligamentous laxity in some
cases of atraumatic shoulder instability.(31) Others found that, in traumatic dislocation, the ligament could undergo plastic
deformation at the time of the traumatic event, and remain permanently stretched, with further stretching caused by repeated
episodes of instability.(2)

In a paper to be published soon, we have shown that IGHL laxity is also consistently seen in traumatic dislocation.

This is an interesting concept, since it shows how shoulder laxity can be acquired
without a discrete macrotraumatic event. Repeated stress (cumulative microtrauma)
from sports involving extremes of abduction and external rotation (throwing, pitching)
may result in gradual stretching of the capsuloligamentous apparatus of the shoulder.
(25) In the position required by these sports, the ligament is tented and stretched over
the unyielding convexity of the humeral head.

Arthroscopic repair of chronic anterior instability Fig. 2 The ligament complex restricts glenohumeral abduction to
Major progress in the arthroscopic management of chronic anterior instability came 85° .
with the advent of devices for the fixation of sutures in bone. This has resulted in a
simplification of the technique, and in a reduction of the recurrence rate, from 30%
following the Caspari procedure,(5) to a level of 10-15%.(14,17,18) However, this rate
is still well above the 3-5% rate seen after conventional surgery. The other major
advance is the recognition of the importance of IGHL laxity, which is seen as a factor
adversely affecting the outcome of arthroscopic repair.(29) Thus, developments in
arthroscopic shoulder surgery have served to under line the role played by IGHL laxity.

Fig. 3 Cutting the ligament provides an additional 10-15° of

Voluntary instability abduction. The tuberosities are not involved in restraining this
Voluntary subluxation or dislocation should be considered as a separate entity. While, movement.
obviously, the patients concerned will have generalized ligamentous laxity, voluntary instability also involves an element of volitional
control of muscle forces. We wish to emphasize that surgery is absolutely contra-indicated in this subset of patients.
The different forms of chronic instability of the shoulder can be brought together in an overall picture. On the one hand, there is
instability with laxity brought on by trauma and made progressively worse by recurrent episodes of instability, and instability with
laxity resulting secondarily from gradual stretching of the capsule (overuse); on the other hand, there is instability in the context of
congenital excessive laxity.

Chronic anterior instability of the shoulder

is, thus, characterized by three main parameters:

- ligamentous laxity, which is a consistent feature, regardless of whether it is congenital or

- a labral lesion, which may vary greatly in size, and which will worsen with every
dislocation of the humeral head;
- anterior soft-tissue stripping, which will often be very slight.

This suggests that anterior instability of the shoulder is an instance of joint instability
resulting from ligamentous laxity. What distinguishes the pattern from that seen in other
joints is the lack of bony constraints in the glenohumeral joint, and the fact that
stabilization is provided by one single ligament. The "labrum syndrome" may be accounted
Fig 4 : Clinical measurement of passive for by the fact that the laxity is great enough to allow the humeral head to ride up the
abduction. The examiner stands behind the glenoid slope, without being great enough to allow it to ride over the rim.
patient, and gradually lifts the upper limb, strictly
in the coronal plane, while pushing down on the
top of the shoulder. The patient's elbow is flexed A constant physiological value
to 90°, while the forearm is held horizontal. The
patient must be fully relaxed throughout the test. Can ligamentous laxity of the shoulder be directly analyzed?

In order to analyze laxity directly, one would need a test performed in a position where the IGHL is taut. This is the only way in
which the presence of stretching can be assessed.

In an anatomical study, it was shown that glenohumeral abduction is consistently limited to 85° by the IGHL, with no other
restraints involved (Fig. 2). The tuberosities have nothing whatsoever to do with restricting the range of this movement (Fig. 3):
dividing the ligament allows an additional 15° of abduction to be obtained.

The clinical measurement of passive abduction may be performed as shown in Figure 4. The examiner stands behind the patient,
and exerts firm downward pressure on the top of the shoulder. The examiner's other hand is used to gently lift the patient's upper
limb in the strict coronal plane, with the elbow flexed to 90° and the forearm held horizontal. The patient must, of course, be
completely relaxed throughout the test. A study in 100 healthy volunteers showed that the range of abduction never exceeded 90°,
in 95% of the patients. This finding was bilateral. In 5% of the subjects, passive abduction was greater than 105°, bilaterally; this
suggested generalized ligamentous laxity. Radiographic checks showed that the scapula did not rotate by more than 5° during the
test. Thus, scapular rotation during the test would account for at most 6% of the range of movement. It follows that the
measurement of passive abduction will show only the movement that occurs in the glenohumeral joint. This movement is limited by
the IGHL, and has a constant range. We therefore hypothesized that IGHL laxity should lead to an increased range of passive
abduction. In the light of the results obtained in that study, we wish to propose a test for the direct assessment of IGHL laxity, which
will be described in the section concerning the clinical examination.


This classification is important from the point of view of prognosis and of the management policy to be adopted.

Instability may be traumatic or atraumatic. It may occur with or without evidence of excessive ligamentous laxity. Instability may
take the form of recurrent dislocation (in which case, reduction may be performed by the patient himself or herself, or may require
the assistance of a third party). Equally, it may manifest itself as a brief sensation of painful instability (the condition described by
Patte et al(24) as a "painful and unstable shoulder," and by Rowe(28) as "recurrent transient anterior subluxation). Also, each
patient should be checked to see whether the instability is voluntary or involuntary.

The two parameters combine to form a spectrum of instability, as in the classification proposed by Silliman and Hawkins:(31)
instability following initial trauma, without any prior laxity; instability following initial trauma and laxity acquired from overuse;
instability following trauma, with prior laxity; and instability without a history of trauma, with prior laxity.

In patients with typical dislocation and radiographic evidence (radiographs showing the shoulder in the dislocated position), the
diagnosis is fairly straightforward; the only pitfall being an associated hyperlaxity that may be missed. As more and more people
practise sports intensively, an increasing proportion of clinic patients will present with a less well-defined pattern. In these patients,
a methodical examination will be of the utmost importance.


Often, the patient will spontaneously report the initial traumatic event. The index event should always be carefully elicited.
Sometimes, the first episode of dislocation would appear to have been atraumatic; however, detailed history taking may reveal fairly
major shoulder trauma in the past, which must be taken into account. There are situations where trauma may cause all the lesions
required for instability, without any dislocation occurring. In other patients, a history of intensive sports practice involving abduction
and forced external rotation will provide useful clues.

On the other hand, the existence of an initial trauma should not stop the physician from searching for an associated multidirectional
hyperlaxity. The nature of the trauma is not always easy to identify. The only important pattern is a direct posterior blow to the
shoulder, or indirect trauma (a blow or fall on the elbow or on the outstretched, externally rotated arm). Indirect trauma is by far
the most frequent cause involved. In this analysis of the first parameter, there is much that is not clear-cut.
It is important to obtain a description of the episodes of instability, trying to ascertain how many there have been, and how easily
they tend to occur. If there have been many episodes of instability after very minor trauma, surgery should, obviously, be
considered. Equally, though, a diver or a mountaineer with a first recurrence may be a candidate, given the risk involved in his or
her sport.

Physical examination
This examination is performed in three stages, and involves a search for three broad patterns: apprehension, during dynamic
manoeuvres designed to reveal instability; laxity, which will be discussed in greater detail below; and evidence of associated
multidirectional hyperlaxity.

Apprehension tests

All apprehension tests are designed to place the humeral head in a position of imminent
subluxation or dislocation, which makes the patient recognize the familiar pattern of
instability, and react with anticipated fear.

Crank test and fulcrum test

This test is designed to reproduce the position of instability. It is the oldest of the
apprehension tests. The examiner places the arm in extreme abduction and external
rotation, which may cause apprehension (Fig. 5). This is the most commonly used test. It
has a high specificity. A negative test does not rule out shoulder instability. The test may
be performed in the sitting or standing patient (crank test), or with the patient supine
(fulcrum test).
Fig 5 Crank test.
This test serves to place the shoulder in a position of
maximal instability (extremes of abduction and
external rotation). The test is positive if the patient
expresses pain or apprehension.
Relocation test
This is a more sensitive variant of the test described above. The patient is positioned
supine. The first part of the test is a classic fulcrum test, in which the humeral head is pushed forward to elicit apprehension. In the
second part of the test, a posteriorly directed force is applied to the humeral head. This prevents anterior subluxation, and produces
a negative apprehension test (Fig. 6a and b).

Inferior apprehension test

This test was initially described by Feagin, and further refined by Itoi et al,16 who suggested the name ABIS (abduction inferior
stability). For this test, the upper limb is held in abduction, with the patient's forearm resting on the examiner's shoulder. The
examiner exerts downward pressure over the neck of the humerus. If the shoulder is unstable, the head will be pushed down, and a
groove will appear; also, the patient may show apprehension (Fig. 7).
Fig 6 : Relocation test. This test is performed with the patient supine.
6a Pressure over the back of the humeral head causes apprehension, while
6b pressure over the front of the humeral head prevents the head suluxating anteriorly, and does
not cause apprehension.

Tests for overall laxity

These tests are designed to show abnormal mobility of the humeral head. Since none
of the articular ligaments is taut in the position used for these tests, the procedures
should not be looked upon as ligamentous laxity tests. What is provided is global,
and difficult-to-interpret, information on excessive joint mobility, covering not only
laxity of the capsuloligamentous apparatus, but also the control of muscle tone.
These procedures are tests of excessive mobility.

Sulcus test
The patient is told to relax, while the examiner exerts gentle downward traction on
the patient's arm (Fig. 8). The test is positive if traction makes the humeral head
Fig 7 : Abduction inferior stability (ABIS) test. The patient's move down; this distal movement of the humeral head manifests itself as a groove
arm is in abduction, with the forearm resting on the or sulcus below the lateral border of the acromion.(20) The amount of downward
examiner's shoulder. The examiner exerts pressure on the
arm, gradually pushing the humeral head downwards. The movement of the humeral head may be measured and graded.
test is positive if there is downward displacement of the
head, or if the patient shows apprehension.
Drawer tests
These tests, too, should be performed in a relaxed patient. The patient is asked to lean forward slightly, with both arms hanging
down. The examiner holds the patient's shoulder girdle with one hand, while cupping the other around the humeral head, and sliding
the head backwards and forwards to detect any abnormal mobility(27) (Fig. 9). This test may be performed with the patient
sitting(26) or supine.(11)
Fig 8 : Sulcus test. In the Fig 9 : Drawer test.
relaxed patient, the examiner The patient is made to relax and slightly
gently pulls the humerus lean forward. The examiner holds the
downwards. The test is humeral head between his or her thumb
positive if the humeral head and index finger, and tries to make the
descends, with formation of a head slide backwards and forwards. This
groove or sulcus under the test demonstrates overall hyperlaxity
lateral border of the (without being specific of any particular
acromion. The amount of ligament), and may provide information on
downward movement can be the direction of the instability.
measured. A positive test is
indicative of abnormal

Is there a specific test for laxity?

Since passive abduction has a constant range, and since the range is controlled by the IGHL, we suggest that laxity of the IGHL will
be associated with an increase in the range of abduction. The passive abduction test was performed in patients with post-traumatic
shoulder instability without any associated hyperlaxity. In 85% of the cases, the range of passive abduction was at least 105°, while,
on the healthy side, it was limited to 90°. In 15% of the cases, the test caused acute apprehension, making it impossible to
measure passive abduction. In such cases, the test works as an apprehension test, along the lines of the procedure initially devised
by Feagin, and proposed by Itoi et al as the ABIS test.

The test was performed under general anaesthesia, immediately prior to surgery. In all the cases, passive abduction was at least
105°, while, on the contralateral side, it was restricted to 90°.

Providing that the test is performed strictly in the coronal plane, it furnishes objective evidence of excessive IGHL length, and gives
a direct demonstration of the laxity of the ligament. Thus, this test of passive hyperabduction is positive if the range on the affected
side is greater than 105° (Fig. 10). This is the first test that allows shoulder ligament laxity to be directly assessed; however, it will
need to be used more widely, in a prospective study, to establish its specificity and sensitivity.
Neurological examination

This part of the general work-up must not be overlooked: in almost 15% of cases of chronic
shoulder instability, the axillary nerve is affected.

Looking for evidence of generalized ligamentous laxity

Multidirectional hyperlaxity affects the outcome of instability treatments.(29) On examination,

there will be a groove of more than 2 cm in the sulcus test, as well as major anterior and
posterior drawer movements. External rotation of the upper limb of more than 90° is also
considered to be a sign of abnormal laxity. The wrists should be examined for increased palmar
flexion, as should the elbow for marked hyperextension, the knees for a recurvatum deformity,
and the trunk for enhanced forward bending (palms of hands to floor). In patients with
Fig 10 : Positive hyperabduction test. generalized ligamentous laxity, the passive hyperabduction test will be bilaterally positive.
Marked asymmetry between the affected
and the healthy side is characteristic of
laxity of the ligament complex. Where these tests are positive, the diagnosis will be one of instability associated with
multidirectional hyperlaxity. Evidence of true multidirectional instability should be carefully
sought. The most important feature is episodes of posterior instability when the arm is in forward elevation and internal rotation.

The patient should be questioned about episodes of posterior instability of the shoulder. Special attention should be devoted to
eliciting previous incidents of voluntary shoulder dislocation.

In some of the more difficult cases, especially when trying to confirm or exclude multidirectional instability, CT arthrography may be
helpful. Arthroscopy may be indicated, to obtain objective evidence of laxity, as described by Detrisac and Johnson.(6).


1. BANKART A.S.B. : The pathology and treatment of recurrent dislocation of the

shoulder joint. Br J Surg 23, 1938.(Abstract)
MOW, V. : Tensile properties of the inferior glenohumeal ligament. J Orthop Res
10:187, 1992.
3. BROCA, A. and HARTMANN, H. : Contribution à l'étude des luxations de
l'épaule (luxations anciennes et luxations récidivantes). Bull Soc Anat 416, 1890.
4. BROCA, A. and HARTMANN, H. : Contribution à l'étude des luxations de
l'épaule (luxations dites incomplètes, décollements périostiques, luxations
directes et luxations indirectes). Bull Soc Anat 312, 1890.
5. CASPARI, R.B.: Reconstruction for anterior shoulder instablity. Tech Orthop
3:59, 1988.
6. DETRISAC, D.A. and JOHNSON, L.L.: Arthroscopic shoulder capsulorrhaphy
using metal staples. Orthop Clin North Am. 24:71, 1993.
7. FERRARI, D.A.: Capsular ligaments on the shoulder. Anatomical and functional
study of the anterior superior capsule. Am J Sport Med 18:219, 1990.
8. GAGEY,O.J., BONFAIT, H., GILLOT, C., and MAZAS, F.: Anatomie fonctionnelle
et mécanique de l'élévation du bras. Rev Chir Orthop 74:209, 1988.
9. GAGEY, O.J., GAGEY, N.F., BOISRENOULT, P., HUE, E., and MAZAS, F.: Etude
expérimentale des luxations antéro-internes et érecta de l'articulation scapulo-
humérale. Rev Chir Orthop 79:13, 1993.
10. GERBER, C.: Les instabilités de l'épaule. Paris, L'Expansion Scientifique
Française, 1989,
11. GERBER, C. and GANZ, R.: Clinical assessment of instability of the shoulder
with special reference to anterior and posterior drawer tests. J Bone Joint Surg
66-B:551, 1984.
12. HABERMEYER, P., GLEYZE, P., and RICKERT, M.: Evolution of lesions of the
labrum-ligament complex in posttraumatic anterior shoulder instability : a
prospective study. J Shoulder. Elbow.Surg 8:66, 1999.
13. HABERMEYER, P., SCHULLER, U., and WIEDEMANN, E.: The intra-articular
pressure of the shoulder : an experimental study on te role of the glenoid labrum
in stabilizing the joint. Arthroscopy 8:166, 1992.
FUKUSHIMA, S.: Arthroscopic Bankart suture repair for traumatic anterior
shoulder instability: analysis of the causes of a recurrence [see comments].
Arthroscopy. 14:295, 1998.
15. HILL, H.A. and SACHS, M.D.: The grooved defect of the humeral head. A
frequently unrecognized complication of dislocation of the shoulder joint.
Radiology 35:690, 1940.
16. ITOI, E., MOTZKIN, N., MORREY, B., and AN, K.: Scapular inclination and
inferior stability of the shoulder. J Shoulder Elbow Surg 1:131, 1992.
MAEDA, A., and IZAWA, K.: Modified Caspari technique for traumatic anterior
shoulder instability: comparison of absorbable sutures versus absorbable plus
nonabsorbable sutures. Arthroscopy. 15:400, 1999.
and SWARD, L.: Bioabsorbable tacks for arthroscopic treatment of recurrent
anterior shoulder dislocation. Scand.J Med. Sci. Sports 8:411, 1998.
19. KUMAR, V.P. and BALASUBRAMANIAN, P.: The role of atmospheric pressure in
stabilizing the shoulder. J Bone Joint Surg 67B:719, 1985.
20. NEER, C.S. and FOSTER, C.R.: Inferior capsular shift for involuntary and
multidirectional instability of the shoulder. A preliminary report. J Bone Joint Surg
62A:897, 1980.
21. NEVIASER, R.J., NEVIASER, T.J., and NEVIASER, J.S.: Concurrent rupture of
the rotator cuff and anterior dislocation of the shoulder in the older patients. J
Bone Joint Surg 44A:523, 1996.
anatomy and histology of the inferior glenohumeral ligament complex of the
shoulder. Am. J. Sport Med. 18:449, 1990.
The contribution of the glenohumeral ligaments to anterior stability of the
shoulder joint. Am.J.Sport Med. 18:449, 1990.
24. PATTE, D., BERNAGEAU, J., and GARDES, P.: Epaules douloureuses et
instables. Rev.Chir.Orthop. 66:157, 1980.
25. PERRY, J.: Anatomy and biomechanics of the Shoulder in throwing,
swimming, gymnastics and tennis. Clin Sport Med 2:247, 1983.
26. ROCKWOOD, C.: Fractures. Tome 1 Rockwood and Green Ed, Lipincott,
Philadelphie, 1984.
27. RODINEAU, J., COURROY, J., and KUMAR, A.: Epaules douloureuses et
instables par lésion du bourrelet et du rebord glénoïdiens. Médecine du sport
54:343, 1980.
28. ROWE, C.R.: Recurrent transient anterior subluxation of the shoulder. The
«dead arm» syndrome. Clin Orthop 11, 1987.
29. SCHENK, T.J. and BREMS, J.J.: Multidirectional instability of the shoulder:
pathophysiology, diagnosis, and management. J Am Acad.Orthop Surg 6:65,
30. SCHLEMM, F.: Uber die Verstarkungbander am Shultergelenk. Arch Anat 48,
31. SILLIMAN, J. and HAWKINS, R.: Classification and physical diagnosis of
instability of the shoulder. Clin Orthop 291, 1993.
WARREN, R.F.: Biomechanical evaluation of a simulated Bankart lesion. J Bone
Joint Surg Am 76:1819, 1994.
33. TURKEL, S.J., PANIO, I.M., MARSHALL, J.L., and GIRGIS, R.G.: Stabilizing
mechanisms preventing anterior dislocation of the gleno-humeral joint. J Bone
Joint Surg 63A:1208, 1981.
Instabilité antérieure récidivante avec hyperlaxité multidirectionnelle de l'épaule.
Rev Chir Orthop 81:682, 1995.
Clinical Examination of the Elbow
C. Dumontier
Hôpital St-Antoine, 184 rue du faubourg St-Antoine, F-75012 Paris
Institut de la Main, 6 square Jouvenet, F-75016 Paris

The elbow complex is made up of three separate articulations, the humero-ulnar joint, the humeroradial (radiocapitellar) joint,
and the superior radio-ulnar joint. These joints are covered by the same capsule. The elbow allows flexion and extension, as
well as pronation and supination, and thus enables the hand to be placed in a variety of positions in space. Elbow flexion
brings the hand to the chest, the mouth, or the face, thereby allowing the performance of most of the activities associated with
feeding, dressing, and body care; elbow extension, on the other hand, takes the hand away from the body, and enables it to
grasp objects.

Elbow injuries are rare; however, they may be difficult to diagnose. This problem may be resolved to some extent or simplified
by a full and systematic clinical examination. The joint is superficial, and hence readily accessible to clinical examination. As
with other structures in the body, the examiner must be thoroughly familiar with the anatomy of the joint and with the
abnormal conditions that may be encountered.

This article deals with the broad principles of clinical examination, and will highlight only some of the disorders of the elbow.



The anatomy of the elbow joint and the surrounding structures has been the subject of much research. In this article, only the main points
that have emerged from recent studies will be summarized.

• The bones (Figs. 1-4 )

Figure 1 Diagrammatic AP view of elbow joint
Figure 2 Diagrammatic lateral view of elbow joint. Note
that the elbow is slightly twisted in respect of the axis of
the ulna.
The trochlea is shaped like a pulley; its medial lip is more prominent than the lateral one. The groove of the pulley runs obliquely
downwards and outwards; it courses in a helical manner, forming an arch of about 330°. The distal joint surface of the humerus is in about
30° anterior rotation with respect to the long axis of the humerus, in the sagittal plane; the condyles have 3-8° of internal rotation with
respect to a line joining the epicondyles, in the axial plane; while, in the frontal plane there is a 6-8° valgus tilt of the condyles with respect
to the long axis of the humerus(32, 34, 42). Elbow rotation is virtually around a single centre, which coincides with the condylotrochlear
axis(48). On a true lateral radiograph of the elbow, the flexion-extension axis is at the centre of three concentric circles formed, respectively,
by the projection of the edges of the condyles, the ulnar groove at the back of the medial epicondyle, and the medial lip of the trochlea (Fig.
4b). This flexion-extension axis is on a vertical line drawn down from the anterior cortex of the humeral shaft. It is an essential landmark
for the implantation of a total elbow joint replacement (TEJR).

Figure 3 AP Figure 4a: Radiograph Figure 4b: True lateral radiograph of the
radiograph of the of the elbow taken at humerus. The centres of the three circles
elbow right angles to the axis formed by the edge of the condyle, the
of the forearm. Note ulnar groove, and the medial lip of the
medial rotation of the trochlea coincide; this point is the flexion-
humerus, as shown in extension axis of the elbow. The axis is on
the diagram in Figure a vertical line drawn down from the
2. anterior cortex of the humeral shaft.
There is a valgus angle of about 4° between the trochlear (greater sigmoid) notch of the ulna and the ulnar shaft (32, 34). The opening of the
trochlear notch is angled ca. 30° posteriorly with respect to the long axis of the ulna; this allows better approximation in flexion to the 30°
anterior rotation of the humeral articular condyles (34). The joint surface of the trochlear notch forms an arc of about 180°; however, it is
not entirely covered with cartilage: in over 90% of individuals, a bare area covered by fibro-adipose tissue extends transversely across the
mid-portion of the trochlear notch; this feature accounts for the frequency of fractures at this site, and permits a trans-olecranon approach to
the joint (34).

The radial head is covered with cartilage over four fifths of its circumference. The 15° angulation between the neck and the shaft of the
radius leaves an excursion of 180° for forearm rotation(34).

• The joint capsule

The capsule is attached around the articular surfaces, and blends with the annular ligament. It covers the tip of the olecranon, the coronoid
process, and the radial fossa. The fibres are arranged in such a way as to provide stabilization in flexion and in full extension(23). When the
elbow is stiff, the capsular capacity will be reduced by more than 50%; equally, the capsular compliance of the stiff elbow will be very
poor, which shows that the capsule itself has been compromised(14). The position of minimum intracapsular pressure ("resting position") is
around 60-70° of flexion, which means that prolonged immobilization of the elbow in this position - as practised since the days of
Ambroise Paré - will increase the risk of capsular contraction(14).

• The ligaments

The medial collateral ligament is a strong and well-demarcated structure that consists of three bundles (Fig. 5):

Figure 5 Diagrammatic view of the medial

collateral ligament, with its three bundles.
The anterior bundle is the most important
functionally, since it provides valgus and
anteroposterior stability.

Figure 6 Diagrammatic view of the lateral

ligament complex. It would appear that the
most import structure is the lateral collateral
ligament, which blends with the annular
ligament. The lateral ulnar collateral ligament
is indissociable from the lateral collateral
ligament, at its attachment to the lateral
epicondyle. Distally, it branches off, and
attaches to the supinator crest. The role of
the accessory lateral collateral ligament is
poorly understood.

Figure 7 Diagrammatic view of the origin

and insertion of anconeus, which covers the
capsule and collateral ligaments on the
lateral side.
• x The oblique anterior bundle is wide (5 mm) and thick.
Its apex is attached to the front and the medial aspect
of the medial epicondyle, of which it covers two
thirds, and its base to the medial aspect of the ulna,
just below the coronoid process(9, 32). This bundle is
taut in flexion and in extension; its mean length is 27
mm(31, 32).

• x The oblique posterior bundle is less well defined. This

broader structure is also attached low on the medial
epicondyle, and inserts in a fan-shaped pattern over
virtually the entire margin of the trochlear notch of
the olecranon. It is one of the structures that make up
the cubital tunnel floor(9). This bundle is taut in
flexion only; it is absent in many primates, which
suggests that it is not a main stabilizer of the
elbow(31, 53). The two bundles of the medial collateral
ligament insert slightly posterior to the centre of
rotation, which accounts for the fact that they are
tauter in flexion than in extension(23).

• x The oblique transverse ligament (sometimes referred

to as the ligament of Cooper) is short, and does not
appear to have a stabilizing role(31, 32). It extends
from the posterosuperior portion of the trochlear
notch to the coronoid process; its origin blends
together with that of the anterior bundle(9). Like the
posterior bundle, it contributes to the floor of the
cubital tunnel.

On the lateral side, there is no discrete collateral ligament in the strict sense of the term. Anatomical patterns vary widely, which is why the
description of the structures involved has been difficult, and why disorders of the lateral collateral complex may be hard to understand(31,
32, 42, 43). There are five ligamentous structures involved in the lateral stabilization of the elbow joint (Figs. 6, 7):
• x The lateral (radial) collateral ligament attaches to the
medial aspect of the lateral epicondyle and inserts into
the annular ligament(31). Its origin overlies the centre
of rotation of the elbow, which means that the
ligament will be taut throughout flexion and
extension(23, 32). Its mean length is 21 mm(31). The
fan-shaped base of the ligament blends with the fibres
of the annular ligament(43).

• x The annular ligament is a thick structure that attaches

to the anterior and posterior margin of the radial
(lesser sigmoid) notch. It makes up four fifths of the
fibro-osseous ring that encircles the head of the

• x The lateral ulnar collateral ligament is difficult to

distinguish as a discrete structure. It attaches to the
lateral epicondyle, like the lateral collateral ligament,
with which it blends at this site; it inserts into the
superficial and posterior fibres of the annular
ligament, but attaches to the supinator crest of the
ulna(31, 42, 43). It forms, roughly, the posterior part of
the lateral collateral ligament complex.

• x The accessory lateral collateral ligament has its origin

in the distal fibres of the annular ligament, and
attaches distal to the attachment of the lateral ulnar
collateral ligament. It is an inconstant structure. Since
it is taut only with varus stress, it may act as a
stabilizer of the annular ligament when a varus stress
is imposed upon the joint(32, 42).

• x The anconeus muscle, whose physiological role is still

the subject of controversy, appears to be chiefly a
joint stabilizer, serving as an active collateral
ligament(7, 53). This would account for the fact that it
is often torn when the lateral collateral ligament
complex is ruptured as a result of elbow dislocation.


About 60% of the axial loads imposed on the elbow will be transmitted through the humeroradial joint, as compared with only 40%
through the humero-ulnar articulation(34). The stresses imposed on the elbow vary; they depend on the load applied, the resultant force
vector, and the length of the lever arm. The loads may amount to 2-3 times the body weight, and to 8-10 times the lifted weight. This
accounts for the compressive loads observed during simple activities such as dressing or feeding(48, 51). The use of crutches will transfer
between 40% and 50% of the body weight onto the upper limb(51).

The elbow is a very congruous joint, and, hence, inherently very stable. In flexion, the coronoid process locks into the coronoid fossa,
while the medial rim of the radial head engages in the trochleocapitellar groove(23). In extension, the apex of the olecranon is held in the
olecranon fossa. Elbow stability is enhanced by the perfect congruency between the radial head and the radial notch of the ulna. Roughly
speaking, the bony surfaces contribute 50% of the mediolateral stability of the elbow, while the other 50% comes from the ligaments(34).
One important thing to bear in mind is that the role of each of these structures varies with the degree of flexion or extension of the elbow.

Seventy-eight per cent of the valgus stability of the elbow is contributed by the medial collateral ligament; the bony surfaces, including the
humeroradial joint, have only an accessory function in the constraint to valgus stress of the elbow, although experimental studies have not
as yet provided unequivocal evidence(19, 29, 33, 42, 48). In experiments, the insertion of a Silastic implant was seen to leave valgus stability
unaltered(19). By and large, valgus stability comes from the bony structures below 20° and above 120° of flexion, and from the anterior
bundle of the medial collateral ligament over the in-between range(11). The flexor-pronator group is bulky, but does not appear to provide
dynamic support of the medial aspect of the elbow(16). The radial head has only a secondary role, providing about 30% of the stability on
the lateral side(19, 23, 33). The minor importance of the radial head as a lateral stabilizer is illustrated by the fact that radial head excision
will not adversely affect the joint, providing that the medial collateral ligament is intact. However, a distinction must be made between
elbow valgus stress, which is checked by the medial collateral ligament, and external rotation (or supination) stress, which is checked by
the lateral collateral ligament(43).

There is much less agreement concerning the roles of the different ligamentous structures in varus stability. Initially, it was thought that the
annular ligament was chiefly responsible for resistance to varus stress between 40° and 60° of elbow flexion(49). According to this author,
the lateral collateral ligament serves to stabilize the annular ligament. This idea was contested by several authors(48), which prompted
Søjbjerg et al. to reinvestigate this subject. They showed that the isolated division of the lateral collateral ligament resulted in 15° of varus
(at 110° of flexion), and that the division of the lateral ulnar collateral ligament had little influence on the instability observed(43). Thus, it
is the lateral collateral ligament complex, and in particular the lateral collateral ligament, that stabilizes to varus and extension loads(34, 43,
49). The sole function of the annular ligament appears to be the stabilization of the radio-ulnar joint.

The anterior joint capsule resists distraction, and, under those conditions, provides 85% of the resistance observed(33).

In the sagittal plane, stability also depends on the medial collateral ligament(33). Loss of less than 50% of the olecranon will not interfere
with function, providing that the collaterals are intact(4). Valgus stability is provided largely by the proximal portion of the trochlear notch
of the ulna (85%), while varus stability is chiefly a function of the distal part of this notch (65%)(4, 23). In the sagittal plane, bony stability
in extension comes from the coronoid process(23, 47). This bony and ligamentous stability is enhanced, in the sagittal plane, by the powerful
action of the muscles around the elbow.


In addition to the standard orthopaedic history, the following items of information should be obtained from patients presenting with elbow
dysfunction: age, duration of the complaint, or time since onset of the elbow-related symptoms. The dominant side needs to be ascertained;
specifically, it must be established whether there has been a recent reversal of the natural dominance, which would show that function has
been severely impaired.

The severity of the patient’s pain is assessed using a visual analogue scale. The site of the pain may provide valuable clues. Conditions
involving the lateral compartment (radiocapitellar joint) provoke pain that extends over the lateral aspect of the elbow, with radiation
proximally to the midhumerus and distally over the forearm; this pain may be deep. Diffuse pathological conditions, on the other hand,
cause pain that is described as periarticular in distribution(55).

The patient should be questioned about locking, pain and/or instability during throwing movements, joint swelling, or fleeting inability to
extend the elbow, which would suggest a joint effusion.

Paraesthesiae of the hand may, in some cases, be related to ulnar nerve compromise at the level of the elbow.

A note should also be made of any previous treatments of the elbow (synoviorthesis, intra-articular injections, surgery).

Since the elbow is a superficial joint, many of its disorders can be readily detected by simple inspection. The patient should be suitably
undressed to the waist for examination. Since, at the front, the muscle masses obscure the joint, much of the examination will need to be
conducted with the examiner sitting or standing behind the patient. The patient should be standing, with shoulders slightly braced back, to
display the elbow.

When the forearm is in full extension and supination, there will be a physiological valgus ("carrying angle") of 9-14°; in women, the angle
will be 2-3° greater (8, 18, 35, 55). This angle has been found to be 10-15° greater in the dominant arm of throwing athletes (5). This angle
allows the elbow to be tucked into the waist depression above the iliac crest; it increases when a heavy object is being lifted (Fig. 8). Any
increase in, or loss of, this physiological angle is indicative either of major elbow instability or of malunion. However, the angle varies
from valgus in extension to varus in flexion, and its measurement is not of any practical importance(48).

Figure 8 The physiological

valgus (“carrying angle”) of the
elbow is increased when a load
is being carried. Normally, the
angle is between 9 and 14°
when the elbow is extended and
the forearm is supinated.
Sometimes, on the side of the elbow, bulging in the para-olecranon groove will be seen; such a swelling is produced by an effusion or by
synovial tissue proliferation (55). On the back, prominence of the olecranon is a sign of posterior subluxation of the elbow, a feature
commonly found in RA (55).

Rheumatoid nodules are extremely common; they are usually found on the posterior aspect of the elbow, mainly on the medial aspect of the
extensor surface. The nodules should be counted and their volume noted: large nodules may cause skin ulceration and harbour infection. A
note should also be made of their site, since they may cause problems if they are over an intended surgical approach to the elbow.

Bursitis is also a frequently encountered pathology, especially in RA patients. As with nodules, the volume of the lesion and the quality of
the overlying skin should be noted.

Inspection may also show skin atrophy at steroid injection sites, or scars from previous surgery. These features must be noted, since they
may affect the surgical approach to be employed.


Palpation starts at the posterior aspect, with the patient standing with his or her shoulder braced backwards. The three palpation landmarks -
the two epicondyles and the apex of the olecranon - form an equilateral triangle when the elbow is flexed 90°, and a straight line when the
elbow is in extension (Figs. 9, 10).
Figures 9, 10 Three bony landmarks - the medial
epicondyle, the lateral epicondyle, and the apex of the
olecranon - form an equilateral triangle when the elbow is
flexed 90°, and a straight line when the elbow is in extension
Since the elbow is a very superficial joint, it can be readily palpated from behind and from the sides. The posterior aspect has the olecranon
mid-way between the medial and the lateral condyle. Slight elbow flexion will bring the olecranon out of the olecranon fossa, in which it
lodges in extension; in this position, the proximal portion of the fossa on either side of the triceps tendon may be palpated (Fig. 11).

Figure 11 Flexing the elbow Figure 12 Anatomical

allows palpation of the olecranon landmarks on the
fossa on either side of the triceps lateral aspect of the
tendon. elbow: The lateral
epicondyle continues
proximally in the
supracondylar ridge.
Two centimetres
distally, the main
landmark is formed by
the radial head.
The olecranon bursa is not in communication with the synovial cavity. This is why the elbow may be mobilized in bursitis, and why even
massive bursitis will not be tender. In chronic bursitis, a boggy globular mass may be palpated; the overlying skin will be thickened. Flat,
hard nodules may be felt under the palpating fingertips(12). In infected bursitis, the skin will be tight and shiny; streaks of lymphangitis will
be commonly seen; while in 25% of the cases, the axillary lymph nodes will be enlarged(12).

On the lateral side, the main landmarks are the lateral epicondyle proximally and the radial head distally. The supracondylar ridge is also
very accessible to palpation; its chief value is that of a landmark for surgical approaches (Fig. 12). Sometimes, palpation may be carried out
all the way up to the deltoid tuberosity. The radial head is palpated with the examiner’s thumb, while the other hand is used to pronate and
supinate the forearm (Fig. 13). The head is about 2 cm distal to the lateral epicondyle(5). Inside the triangle formed by the bony
prominences of the lateral epicondyle, the radial head and the olecranon, the joint itself is palpated, to detect even very minor effusions or
low-grade synovitis (Fig. 14).

Figure 13 Anatomical landmarks on

the lateral aspect of the elbow: The
radial head is palpated with the
thumb, while the examiner’s other
hand is used to pronate and supinate
the forearm.
Figure 14 The elbow joint may be
palpated inside a triangle formed by
the bony prominences of the lateral
epicondyle, the radial head, and the
olecranon. This palpation will reveal
even minor effusions or mild
synovitis. Puncture for joint
aspiration is performed inside this
triangle. Similarly, an arthroscopy
portal may be placed there
(posterolateral portal).

The muscles on the lateral side may be palpated individually. For the palpation of brachioradialis, the patient is asked to clench his or her
fist and flex the elbow with the forearm in neutral position (mid-way between pronation and supination) and with the fist blocked under a
table (Fig. 15). The wrist extensors are palpated by asking the patient to extend the forearm at the elbow against resistance (Fig. 16).
Extensor carpi radialis longus produces both flexion and abduction of the wrist. Anconeus may be palpated behind the radial head, on the
side of the olecranon; it increases in bulk when the forearm is extended against resistance(35).

Figure 15 Palpation and testing of Figure 16 Palpation and testing of the

brachioradialis, a forearm flexor. wrist extensors
From the medial side, the joint is not very accessible to palpation, and the small amount of synovial tissue on the medial border of the
olecranon makes joint palpation difficult(35). Palpation of the ridge that provides insertion for the intermuscular septum is useful mainly as
a guide for surgical approaches. Also, the supracondylar lymph nodes may be palpated at this site (Fig. 17). Over, and slightly anterior to,
the supracondylar ridge, a bony excrescence may be palpated; this outgrowth may irritate the median nerve(5). This supracondylar process
is present in 1-3% of the population, and is seen at a distance of 5-7 cm above the joint line(32). Behind the septum, the ulnar nerve may be
palpated; in patients with a very mobile nerve, it may be seen to roll on the medial condyle(10) (Fig. 18). Ulnar nerve instability is more
easily tested with the arm in slight abduction and external rotation, with the elbow flexed between 20 and 70°.

Figure 17 Palpation of the Figure 18 The ulnar nerve is palpated behind the
medial aspect of the elbow. intermuscular septum. It may sometimes sublux or
Above the medial roll on the epicondyle. Ulnar nerve instability is
epicondyle is the ridge on more readily demonstrated if the elbow is flexed 60°
which the intermuscular and the upper limb is abducted and externally
septum inserts. Two rotated.
centimetres above the
epicondyle is the site used
for lymph node palpation.
Anteriorly, the bulk of the flexor-pronator group restricts the extent of joint palpation. The flexor-pronator muscles must be tested as a unit,
by asking the patient to perform wrist adduction and flexion against resistance (Fig. 19). Next, each one of these muscles should be tested
individually. The anterior aspect does not lend itself to palpation, since it is tucked away behind the muscles. Laterally, brachioradialis will
be felt; and in the middle, the biceps tendon is readily accessible if the patient is made to flex the forearm against resistance. Lacertus
fibrosus is palpated medial to the biceps tendon; the pulse of the brachial artery will be felt deep to this aponeurosis (Fig. 20). Sometimes
anterior protrusion cysts produced by herniated synovial membrane may be felt(52).

Figure 20 Palpation of
the medial biceps
expansion (lacertus
fibrosus), which courses
over the brachial vessels
and the median nerve.
Figure 19 Diagrammatic view of the pattern of the flexor-pronator group: The thumb
represents pronator teres; the index, flexor carpi radialis; the middle finger, palmaris
longus; and the ring finger, flexor carpi ulnaris.

The main function of the elbow is to bring the hand to the mouth; this is why the investigation of the elbow range of movement (ROM) is
an important part of the examination process. Any difference between passive and active mobility is usually due to reflex inhibition from
pain(55). The end-feel - the feeling transmitted to the examiner’s hands at the extreme range of passive motion - must also be assessed
(Table 1). If the feel is abnormal, there is usually something wrong with the joint.

Bony Two hard surfaces meeting, bone to bone

(elbow extension)
Capsular Leathery feel, further motion available
(forearm pronation and supination)

Soft tissue Soft tissue contact (elbow flexion)


Spasm Muscle contraction limits motion

Springy block Intra-articular block; rebound is felt

Empty Movement causes pain, pain limits


Table 1 Classification and description of end-feels (modified from TS Ellenbecker & AJ Mattalino)(12a)

• Flexion-extension

The normal flexion-extension range is 0 to 140° (+ 10°). Mobility is measured with a goniometer placed on the side of the arm and
forearm; the measurement thus obtained will be reliable to within 5° of accuracy(35) (Fig. 21). This ROM is well in excess of what is
needed for the majority of activities of daily living (ADLs). The useful arc of motion is between 30° and 130° of elbow flexion; most ADLs
require an arc of only 60-120°(30, 33, 55) (Fig. 22).

Figure 21a-b Flexion-extension is measured with a goniometer applied to the lateral

aspect of the elbow. The normal range is 0-140° (+ 10°).

Figure 22 Arcs of flexion required for some activities

of daily living (adapted from Morrey, 1981)
Door / Chair / Fork / Telephone
Loss of extension provides a very sensitive clue to intra-articular elbow pathology, since extension is the first sector of the ROM to be
affected, and the last to recover(55). However, since the extension deficit shortens the lever arm, it is well tolerated up to a loss of 45°(55).
At the end of flexion, there will be a soft-tissue approximation end-feel as the movement is blocked by the bulk of the arm and forearm
muscles. At the end of the normal extension movement, there will be a bony end-feel, as the olecranon locks into the olecranon fossa(5).
• Pronation and supination

Pronation and supination cannot be complete unless the proximal and distal radio-ulnar joints are in correct anatomical relationship, the
two bones are of normal length relative to each other, and the interosseous membrane is intact(55). The arc of motion varies widely in
different individuals; the mean values are 70° pronation, and 85° supination (Figs. 23, 24). However, with only 50° pronation as well as
supination, most ADLs can be readily performed(30, 55). At the end of pronation and supination, there will be a capsular end-feel(5).

Figure 23 Measuring pronation: The Figure 24 Measuring supination: The

vertical limb of the goniometer is placed horizontal limb is placed on the anterior
parallel to the long axis of the humerus, aspect of the wrist. The mean value is 85°.
while the horizontal limb is placed on the
back of the wrist (to eliminate additional
motion at the radiocarpal joint). The mean
value is 70°.

• Mediolateral stability

Stability testing is performed with the patient standing, shoulder braced backwards; the examiner is behind the patient. The elbow is
slightly flexed, to bring the apex of the olecranon out of the fossa. Varus stability is checked with the humerus in full internal rotation,
while valgus stability is tested in full external rotation (Figs. 25, 26). The physiological laxity of the elbow between 10 and 20° of flexion,
in varus and in valgus, does not exceed 5°. In rotation (pronation and supination), it does not exceed 3°(49).

Figure 25a-b Testing 25b: To eliminate Figure 26 Testing 26b: Valgus instability.
mediolateral elbow stability interfering movements mediolateral stability in The physiological
25a: To eliminate interfering during valgus instability an RA patient with a mediolateral instability is
movements during varus testing, the humerus is TEJR < 5°.
instability testing, the placed in full external 26a: Varus instability
humerus is placed in full rotation. Valgus testing is
internal rotation and the done with the forearm
forearm in pronation. pronated, to test the
medial collateral ligament,
followed by testing in
supination, to check the
lateral collateral complex.
Patients may also be examined lying supine. In this case, the humerus is held with one hand, while the examiner’s other hand places the
forearm in valgus (or in varus), with the elbow flexed 20-30° (to remove the olecranon from the fossa)(5, 39) (Fig. 27). With the patient’s
abducted and externally rotated arm tucked under the examiner’s shoulder, the medial collateral ligament may be palpated at the same
time(11) (Fig. 28). As we shall see in the section on instability, it is important for mediolateral stability to be tested in pronation and in

Figure 27 Assessing mediolateral stability Figure 28 With the abducted and

with the patient lying supine. The elbow externally rotated arm tucked under the
should be slightly flexed, to disengage the examiner’s shoulder, the medial collateral
olecranon from the fossa. Testing must be ligament may be palpated.
done in pronation and in supination.
• Anteroposterior stability

Anteroposterior stability is controlled exclusively by the collaterals. Removal of the olecranon will not result in instability if the collaterals,
and above all the medial collateral ligament, are intact. In RA patients, a search should be made for anteroposterior translation, which
shows the extent of joint destruction. The forearm is flexed to 90° and held by the examiner with one hand, while the other hand holds the
humerus, as anteroposterior stress is applied to the joint.


This examination forms part of the examination of the elbow; depending on the patient’s symptoms, a rough screen or a more detailed
investigation will have to be performed.

• Ulnar nerve

At the elbow, the ulnar nerve may be damaged at several different levels: at the arcade of Struthers; in the ulnar groove behind the medial
epicondyle; under the fascial band bridging the two heads of flexor carpi ulnaris (arcade of Osborne); and even under the deep aponeurosis
of flexor carpi ulnaris(3) (Figs. 29, 30). Tinel’s sign, elicited at different levels, will give a clue as to the site of the compression (Fig. 31).

Figure 29 Diagrammatic view of the

four zones around the elbow in which
the ulnar nerve may be compressed
1 - Under the deep aponeurosis of
flexor carpi ulnaris
2 - Under the fibrous arcade formed
by the two heads of flexor carpi
ulnaris (arcade of Osborne). This is
the most frequently encountered
entrapment site.
3 - Behind the medial epicondyle
4 - At the arcade of Struthers

Figure 30a-b Ulnar nerve compression in an 30b: With the nerve pulled out of the way, the
RA patient. inflamed synovia is seen at the bottom of the
30a: Ulnar nerve (on loop) flattened by tunnel.
synovial proliferation, and destruction of the

Figure 31 Eliciting Tinel’s sign.

Paraesthesiae in the territory of the
ulnar nerve allow an assessment of
the likely site of compression
In the ulnar groove, the ulnar nerve may be affected by synovial proliferations herniating on the medial side of the joint; by bony lesions
with spicules causing irritation along the course of the nerve behind the epicondyle; or by ischaemic events. Paraesthesiae of the ulnar
border of the hand and the fingers are usually the first signs of ulnar neuropathy. Pain is a less frequent complaint; where it is encountered,
it is usually localized to the elbow or along the medial edge of the forearm. These symptoms are commonly triggered or exacerbated by
attempts to flex the elbow. Prolonged elbow flexion may produce the paraesthesiae reported by the patient. This test - known as the elbow
flexion test, and analogous to Phalen’s test for carpal tunnel syndrome - can be made more discriminating by putting the wrist in extension,
so as not to perform a Phalen test at the same time. In the more advanced stages, there will be ulnar nerve palsy as well.

• Posterior interosseous nerve

The posterior interosseous nerve is the motor branch of the radial nerve. It can be found, at the back of the arm, using the three-finger
method described by Henry: the index, middle, and ring fingers of the examiner’s hand opposite the examined side are placed on the
posterior aspect of the radius, with the ring finger at the junction between the neck and the head of the radius. The nerve will then be under
the tip of the index finger.

• Anterior interosseous nerve

This motor branch of median nerve origin may be compressed where it courses between the two heads of pronator teres(17). Compression
of this nerve will lead to weakness and even paralysis of the flexor digitorum profundus muscle of the index finger and the flexor pollicis
longus muscle of the thumb. The patient will be unable to create an OK sign (pinching the tips of the index finger and the thumb together).

• Global muscle evaluation

Testing the different muscles around the elbow forms part of the standard workup of patients with elbow complaints. Each muscle must be
tested separately. Details of the technique to be employed will be found in the standard textbooks, such as Kendall’s(22). The global
evaluation gives a picture of the extent of a nerve trunk or plexus lesion. Table 2 lists the actions, nerve supply, and nerve root derivations
of the different muscles.

Movement Muscles Nerve Supply Nerve Root

Brachialis Musculocutaneous C5 C6 (C7)

Biceps brachii Musculocutaneous C5 C6

Elbow flexion Brachioradialis Radial C5 C6 (C7)

Pronator teres Median C6 C7

Flexor carpi Ulnar C7 C8


Elbow Triceps Radial C6 C7 C8

extension Anconeus Radial C7 C8 (T1)

Forearm Supinator Posterior C5 C6

supination (radial)

Biceps brachii Musculocutaneous C5 C6

Forearm Pronator Anterior C8 T1

quadratus interosseous

pronation Pronator teres Median C6 C7

Flexor carpi Median C6 C7

Wrist flexion Flexor carpi Median C6 C7


Flexor carpi Ulnar C7 C8


Extensor carpi Radial C6 C7

radialis longus

Wrist Extensor carpi Posterior C7 C8

radialis brevis interosseous
extension (radial)

Extensor carpi Posterior C7 C8

ulnaris interosseous

Table 2 Action and innervation of the muscles around the elbow

• Reflexes

Testing the reflexes forms part of the neurological examination of the elbow. The biceps reflex is a C5 function (although the muscle is
supplied by C5 and C6); the brachioradialis reflex is C6; and the triceps reflex, C7 (Figs. 32, 33, 34).
Figures 32, 33, 34 Testing the biceps reflex (Fig. 32 - C5 root), brachioradialis reflex (Fig.
33 - C6 root), and triceps reflex (Fig. 34 - C7 root)

• Functional investigation of the elbow

Pronator, supinator, and extensor strength is largely the same on the dominant and the non-dominant side; this is why a comparison with the
unaffected side is the best test(13). However, flexor strength is always greater on the dominant side. A manual assessment of elbow strength,
at 90° of flexion and with the forearm in neutral position, may appear somewhat crude. However, this isometric measurement is very
reliable, yielding values close to the results of isotonic measurements, which are much more difficult to perform(6). Male subjects are about
twice as strong as females, and the dominant limb is about 6% stronger than the non-dominant one(6). Extensor strength is about 60-70% of
flexor strength(6, 35, 55). Supinator strength usually exceeds pronator strength by 15%(6, 35, 55). Strength must be assessed, especially in
RA, because of the magnitude of mechanical stresses: a 1 kg weight at the hand produces a reactive force, at the 90° flexed elbow, of the
order of 10 kg(55).

• Functional investigation of the upper limb

The elbow is situated between two highly mobile joints - the shoulder and the wrist. It can compensate only for flexion-extension deficits.
As regards flexion, the overall capability for feeding (hand to mouth), body care (hand to face), and hair care (hand to back of head) should
be examined. The activities looked at with regard to extension are opening doors, and reaching to grasp objects.

• Global functional investigation: the lower limbs - walking aids

A search must be made for coexisting conditions affecting the lower limbs. The use of walking aids puts increased functional loads on the
elbow, and may rule out elbow joint replacement. On the other hand, a hip or knee condition may have to be treated with surgery first,
before doing elbow replacement, so as not to jeopardize the outcome of the elbow arthroplasty.

Clinical Examination of the Elbow

C. Dumontier
Hôpital St-Antoine, 184 rue du faubourg St-Antoine, F-75012 Paris
Institut de la Main, 6 square Jouvenet, F-75016 Paris


Rheumatoid arthritis (RA) frequently affects the elbow; however, it is the presenting complaint in only 3% of cases(15). The clinical and
radiological course of the disease is very similar to the pattern seen in the shoulder and the metatarso-phalangeal joints(15). Usually, the
symptoms occur late in the course of the disease, after a duration of 11 to 15 years(15, 52). In Hämäläinen’s study(150, only 33% of the
elbows were causing symptoms after 17 years of disease, and only 20% had major radiological compromise. On the other hand, in Porter’s
earlier series of 225 patients(46), only 28% of the patients had no clinical evidence of elbow involvement; 25% had severe disability in one
or both limbs resulting wholly or partly from disease of the elbow. In another study, reported in(52), only 21% of 191 RA patients had
unilateral disease.

RA of the elbow presents four different patterns:

1. Nodules and bursitis, which are very common superficial lesions that rarely require treatment.

2. Synovial hyperplasia, which marks a more advanced stage of the disease, but is rarely noticed by the patients themselves.
The condition is detected by palpation, especially of the lateral aspect of the elbow. This hyperplasia may be so exuberant as to
cause synovial herniation, usually in the ulnar groove, with possible nerve compression by the herniated tissue.

3. Ulnar nerve compression, which is due either to synovial proliferation or to joint destruction (Fig. 30). Ulnar nerve
translocation is indicated if there is entrapment neuropathy. Compression of the posterior interosseous nerve by synovial cysts
arising from the anterior aspect of the humeroradial joint has been reported in RA(25, 27). While this is a rare lesion, the
possibility of this entrapment neuropathy should be considered in patients presenting with loss of active extension of the
metacarpo-phalangeal (MCP) joints of the triphalangeal digits. The two differential diagnoses to be considered are rupture of
the extensors, and irreducible extensor dislocation into the intermetacarpal valley. Often, several features will need to be taken
into account. Thus, long-standing and/or severe RA of the elbow, gradual loss of extension, tenderness on palpation of the ulnar
groove, and a tenodesis effect on the extensor tendons with wrist flexion, would argue in favour of compression neuropathy.
Extension at the interphalangeal joints will still be possible, through the action of the intrinsic muscles of the hand. Wrist
extension can still be accomplished, albeit with radial deviation, since only the extensor carpi ulnaris muscle is paralyzed. With
incomplete involvement, there will be lack of MCP extension in the fourth and fifth digits only, producing a ‘false ulnar claw
hand’, which differs from a true ulnar claw by the absence of MCP joint hyperextension.

4. The chief problem in the management of RA of the elbow is the gradual destruction of ligamentous and/or articular structures
by the disease. The workup of the RA elbow will, therefore, have to include an assessment of the stability of the joint in the
coronal and the sagittal plane, and an evaluation of the functional capabilities of the elbow and of the other joints of the upper
limb (Fig. 26). Patients who cannot ambulate without aids should not be considered for elbow joint replacement.


In 1873, Runge first described this condition, noting that “its aetiology is varied, its pathogenesis unknown, and its treatment uncertain.”
Since then, much has been written about what is now popularly known as tennis elbow. Readers interested in more detailed information on
the subject should consult the extensive literature survey given by Narakas and Bonnard(36).

The disorder has a nearly equal gender incidence, and is most frequently seen in the 35-55 age group(38). Often, a history of elbow overuse
can be obtained (change of tools or equipment, different rate of elbow use, etc.). The clinical examination must, in the first instance,
establish the actual site of the pain (anterior or posterior aspect of the epicondyle; humeroradial joint; annular ligament; common extensor
origin). The radial nerve should be palpated along its course, especially where it passes in front of the annular ligament(36). The muscles
are pinched laterally, while the nerve is felt by direct palpation(36).

Lateral epicondylar pain may be caused by humeroradial joint disorders, radial nerve compression, or lesions of the tendinous attachments
on the epicondyle. The clinical examination should establish which of these causes is responsible for the symptoms in the particular patient
(Table 3).

"Tendinous" Epicondylar pain

epicondylitis from nerve damage
Pain Over the epicondyle Yes Diffuse

Over lateral elbow +++ +++

Nocturnal (+) +++

Exercise-related +++ +++

Dorsum of wrist rare frequent

Lat. aspect of arm rare frequent

Shoulder rare frequent

Along radial nerve track NO Yes

Painful Maximum elbow ++ +++

manoeuvres extension

Resisted wrist extension +++ ++

Maximum finger rare frequent


Fist-clenching ++ +

Resisted supination + ++

Tenderness Epicondyle +++ rare

Humeroradial joint often NO

Annular ligament ++ +++

Extensor muscles sometimes frequent

Lat. bicipital groove NO frequent

Supinator (+) +++

Table 3 Tests for the determination of radial nerve involvement in "epicondylitis" (From [36])

Primary radial nerve lesions account for only 5 to 10% of all cases of lateral epicondylar pain. Secondary compromise, from the
peritendinous spread of inflammatory lesions, is probably more common, although no exact figures can be given. Nocturnal pain is
suggestive of nerve damage. Stretching of the radial nerve may be accomplished by asking the patient to supinate the forearm against
resistance, with the arm elevated forward at the shoulder, the elbow in full extension, the forearm pronated, and the wrist flexed(36).
According to Lister et al.(24), three features are typical of radial nerve involvement (radial tunnel syndrome): the pain is well localized to
the extensor mass, just distal to the radial head; resisted extension of the middle finger with the elbow extended produces pain at the site of
the previously elicited tenderness; and similar pain is produced on resisted supination of the extended forearm. Most of my patients with
lateral epicondylitis were found to have pain over and in front of the radial head; I do not think that this is due to nerve involvement.
Nirschl(38) is of the opinion that this pain is indicative of degeneration of the origin of the extensor carpi radialis brevis, in the deeper

In humeroradial joint disease, the epicondyle and the tendon attachments are non-tender to palpation, whereas the humeroradial joint space
and the annular ligament will be tender(36). Compression of the humeroradial joint by placing the elbow into valgus may also be painful.
In pathophysiological terms, it is probably rare for humeroradial joint compromise to be the sole source of pain in the lateral epicondylar

The most frequent cause of pain in this region are lesions of the tendinous insertions on the lateral epicondyle; in particular, the attachment
of the extensor carpi radialis brevis tendon, which is partly proximal to the elbow’s axis of rotation, may be responsible for the patient’s
symptoms. Any tests stressing the wrist and finger extensors will incite pain(38), which limits the diagnostic value of such tests (Figs. 15,
16). The main disadvantage of the diagnostic tests devised for the workup of lateral epicondylitis is their great sensitivity allied to poor


Medial epicondylitis, commonly known as golfer’s elbow or medial tennis elbow, occurs five to seven times less frequently than (lateral)
tennis elbow, and is predominant among men(38). In over 50% of the cases, there is associated ulnar nerve damage. The pain is either over
the medial epicondyle, or 1 cm distal to the epicondyle, in the flexor-pronator muscle mass (Fig. 35). Passive stretching of these muscles by
extension of the elbow and the wrist, with the forearm supinated, will be painful in patients with tendon attachment lesions. Similarly, the
resisted contraction of the flexor-rotator group will incite medial epicondylar symptoms(38). All patients presenting with this condition
must be examined for possible ulnar nerve irritation.
Figure 35 Medial epicondylitis, with
intratendinous degeneration. View at
In order to differentiate medial epicondylitis from chronic lesions of the medial collateral ligament in throwing athletes, the patient’s arm
should be placed with the elbow extended, the forearm supinated, but the wrist flexed. In this position, a valgus stress on the elbow will not
produce pain in patients suffering from medial epicondylitis.


(excluding radio-ulnar instability)
While elbow instability is rare, it has been known for a long time. Instability has been reported as a sequela in 15 to 35% of elbow
dislocation patients(20, 26). However, it was not until 1991 that a comprehensive theory of elbow instability was advanced, taking into
account recent work and studies of the anatomy and the biomechanics of the elbow. In the light of this theory, O’Driscoll et al. proposed a
clinical examination designed to detect the probable lesions within a spectrum comprising valgus instability, varus instability, anterior
instability, and posterolateral instability(41).


This instability may be acute or chronic.

The acute form is due to tearing of the anterior bundle of the medial collateral ligament (often associated with other soft tissue lesions); it is
found mainly in elbow dislocation(54). According to O’Driscoll et al.(41), it is often associated with a fracture of the head of the radius.
When, after reduction of the dislocation, the elbow is valgus tested with the forearm pronated, this instability is commonly seen(54). The
condition rarely becomes chronic, since the ligament tends to heal well, probably because it is surrounded by muscles; it should, however,
be noted that, in 50% of the cases, the muscles will be torn as well(41, 54). Outside the context of dislocation, acute elbow instability is
rare, and is mainly confined to athletes practising throwing sports (javelin throwers, baseball pitchers)(11, 53). Sharp pain on palpation
along the medial collateral ligament (rather than over the ulna) exacerbated by valgus stress with the forearm pronated is usually
diagnostic. The question of the exact amount of flexion required to improve the value of the test has not yet been settled. It would appear
that the ideal angle is somewhere around 60°. If the resisted flexion of the fingers or the wrist is painless, the (even rarer) condition of
isolated flexor-rotator muscle rupture may be ruled out (Fig. 36).

Figure 36 In valgus testing, to look for a

recent lesion of the medial collateral ligament,
the absence of pain on resisted wrist flexion
rules out a flexor-rotator muscle tear.
Complete tears of the medial soft tissues (capsule, ligament, and flexor-rotator muscles) have also been reported(39). In addition to the
signs of acute instability, there will, very often, be bruising and swelling over the medial epicondyle (Fig. 37), and signs of nerve irritation.

Figure 37a-d Fresh traumatic lesion of the

capsuloligamentous and muscular structures.

37a Bony avulsion of 37b At 3 days after the accident

the medial collateral
37c Partial tear of the flexor-rotator muscles 37d Complete avulsion of the medial
ligament complex, with gross detachment of
the capsule and the periosteum
Chronic instability has been studied mainly by American researchers, since this form of instability specifically affects throwing athletes
such as baseball pitchers(11). Valgus elbow deformity has also been found to occur in more than 30% of professional baseball pitchers(53).
Chronic stretching of the medial collateral ligament causes pain on throwing, and will result in a flexion deformity in two thirds, and ulnar
nerve damage in 40%, of the patients(11). Palpation of the medial collateral ligament will elicit tenderness, while valgus testing with the
forearm pronated will show instability in two thirds of the elbows examined(11) (Fig. 38). These chronic instabilities will lead to
impingement of the medial tip of the olecranon process on the wall of the olecranon fossa, with production of loose bodies (Fig. 39).

Figure 38 Chronic Figure 39 Medial olecranon

elbow instability, with process impingement as a
gaping in valgus, after result of instability.
avulsion of the medial Repeated impingement of
collateral ligament the medial tip of the
olecranon on the
posteromedial trochlear
surface will lead to the
exfoliation of loose bodies,
which may cause joint
Loose bodies in the humeroradial joint will only subsequently be induced by this impingement(53). Olecranon impingement is searched for
by placing the patient’s arm in full extension, exerting valgus stress, and simultaneously palpating the posteromedial surface of the
olecranon(5). If there is impingement, there will be tenderness and, sometimes, crepitation either from loose bodies or from a synovial
reaction over osteophytes (Fig. 40). This instability-related impingement must be differentiated from pain without instability felt over the
medial surface of the olecranon when there are radiologically demonstrable loose bodies and/or osteophytes. These lesions, which are
marked by pain and gradual stiffening with loss of the last degrees of extension, are seen in boxers and in patients with primary
osteoarthritis; they may also occur in “ageing athletes”(21).

Figure 40 Olecranon impingement is looked for by putting the forearm

into maximum extension and applying valgus stress, while palpating the
posteromedial aspect of the olecranon. Impingement with instability
must be distinguished from pain without instability as seen chiefly in
“ageing” athletes.

Varus instability is also seen in elbow dislocation. Chronic instability is rare, perhaps because the elbow is rarely stressed with a pure varus
force(41, 42). For this instability to occur, there must be a complete tear of the lateral collateral complex. For mechanical reasons,
examination for varus instability must be performed in pronounced flexion (to at least 70°), since it is in this position that the ligament is
maximally relaxed.


Anterior instability is seen in olecranon fractures, which pose a management rather than a diagnostic problem.


Posterolateral instability is thought to be the most frequent form of elbow instability. The pattern is one of rotational displacement of the
ulna (and the radius) on the humerus, leading to supination (or external rotation) of the ulna in relation to the humerus(41). The first
studies, in particular those by Josefsson et al.(20), showed elbow dislocation to be consistently associated with a rupture of both collateral
ligaments. However, the comparative rarity of instability after dislocation did not fit in with the fundamental role (at least in experimental
studies) of the anterior bundle of the medial collateral ligament(50). Recent studies by O’Driscoll et al.(41) and Regan et al.(48) showed
that dislocation may occur around an intact medial collateral ligament if the dislocating force acting on a forearm in external rotation (or
supination) exerts an axial stress in flexion and in valgus. These studies confirm the results of earlier experimental work by Søjbjerg et al.
(50), and contradict the previously held idea that elbow dislocation results from hyperextension. Witvoet and Tayon have given a good
description of these rotational subluxations that occur with the elbow flexed when carrying a load. These authors, as well as others(45, 56),
also describe posterolateral capsular distension, which provides a rationale for the treatment of recurrent dislocation by posterolateral
stabilization. This recognition of posterolateral instability as a distinct entity provides an explanation of “valgus instability” with lateral
lesions. Posterolateral instability is associated with external rotation (supination) of the ulna during testing in valgus, with the forearm
supinated (Fig. 41). This instability disappears during valgus testing with the forearm pronated, because, under these conditions, the
structure being tested is the medial collateral ligament(41, 44).
Figure 41 Posterolateral instability following elbow dislocation, in a 25-year-old
patient. Radiographs taken under general anaesthesia show posterolateral
subluxation of the radial head during forearm supinating stress.
O’Driscoll et al. describe three stages of instability. Stage 1 is posterolateral rotatory instability. According to these authors, the lateral ulnar
collateral ligament is torn at this stage; however, there is considerable controversy concerning this idea. Patients with this lesion will have a
positive lateral pivot shift test. The lesion may occur as a result of dislocation, but is also seen after the elbow has been subjected to a varus
stress(37). Stage 2 is characterized by incomplete dislocation, with the ulna perched on the trochlea. Stage 3 shows complete dislocation,
with the coronoid process behind the humerus(41). O’Driscoll et al. also distinguish between a Stage 3a, in which the anterior bundle of the
medial collateral ligament is intact, which means that the elbow will be stable in valgus following reduction; and a Stage 3b, in which this
ligament is torn, resulting in gross instability after reduction(41).

Patients suffering from posterolateral instability will have pain and report a sensation of snapping or catching(37, 41). These symptoms are
particularly marked in near-extension with the forearm supinated(41). An extension deficit is seen in one third of the patients(37). Recently,
there have been several reports of instability in post-traumatic cubitus varus(1, 2, 28). While there is still some controversy on the subject,
it would, in the light of the most recent studies, appear that the main structure that has been damaged is the lateral collateral ligament(44).
The lateral pivot shift is enhanced if the lateral collateral ligament is transected, or if all the collateral ligaments are cut(44). The division of
the other ligaments, in particular that of the lateral ulnar collateral ligament, would appear to cause less destabilization. The posterolateral
capsular distension seen in chronic instability would appear to be the result of the lesions, rather than a lesion required for the instability to

The lateral pivot shift test, which reproduces the instability noted by the patients, may be performed in two ways. The most efficient
method consists in placing the extremity over the patient’s head (Fig. 42). The examiner holds the wrist and the elbow. The forearm is fully
supinated, and valgus stress is applied as the elbow is moved from the fully extended position to a flexed position(41). The method
constitutes an apprehension test, since, in the awake patient, the manoeuvre is very uncomfortable, while it is virtually impossible to elicit
frank subluxation(37). General anaesthesia (or the intra-articular instillation of a local anaesthetic) would be required in order to produce a
snap at about 40° of flexion, where the subluxation reduces(37, 41). Prior to this reduction, a posterolateral prominence is produced by the
subluxation of radial head, and a dimple may be seen in the soft spot area between the radial head and the capitulum(41). The test may also
be performed with the arm held internally rotated at the side, with the patient recumbent(40). The forearm is supinated, and the elbow is
taken from full extension into flexion while applying valgus moments and axial compression force(40) (Fig. 43). The results are the same
as in the “overhead” test described above.

Figure 42 Performance of lateral pivot shift test, as proposed by

O’Driscoll et al. The examiner holds the wrist and the elbow. The
forearm is supinated, and a valgus stress is applied as the elbow is
taken from extension into flexion. The “snap” noted by the patients can
only be reproduced under general anaesthesia; it occurs around 40° of
elbow flexion.
Inset: Diagrammatic representation of rotatory subluxation of ulna on
humerus around the pivot of the medial collateral ligament

Figure 43 Performance of lateral pivot shift test on a recumbent patient.

The arm is placed alongside the body, in full internal rotation. The forearm
is supinated, and axial compression and valgus stress are applied as the
elbow is moved from the fully extended to a flexed position.
I am indebted to Prof. Thierry Bègué for his comments and corrections; to the radiology team at the Clinique Jouvenet, for their
assistance; and to Ms Angèle Travadel, for help with the illustrations.

1. Abe M., Ishizu T., Nagaoka T., Onomura T. Recurrent posterior dislocation of the
head of the radius in post-traumatic cubitus varus. J.Bone Joint Surg. 1995 ;77B :

2. Abe M., Ishizu T., Morikawa J. Posterolateral rotatory instability of the elbow after
posttraumatic cubitus varus. J. Shoulder Elbow Surg. 1997 ;6 :405-409.

3. An K-N., Morrey B.F. Biomechanics of the elbow. In The elbow and its disorders.
B.F. Morrey ed, W.B. Saunders, Philadelphia, 1985, pp43-61.

4. An K.N., Morrey B.F., Chao E.Y. - The effect of partial removal of proximal ulna on
elbow constraint. Clin. Orthop. Rel. Res. 1986 ;209 :270-279.

5. Andrews J.R., Wilk K.E., Satterwhite Y.E., Tedder J.L. Physical examination of the
thrower’s elbow. J.Sports.Phys.Ther. 1993 ;17 :296-304.

6. Askew L.J., An K-N., Morrey B.F., Chao E.Y.S. Isometric elbow strength in normal
individuals. Clin.Orthop. 1987 ;222 :261-266.

7. Basmajian J.V., Griffin W.R. Function of the anconeus muscle. An

electromyographic study. J.Bone Joint Surg. 1972 ;54A :1712-1714.

8. Beals R.K. - The normal carrying angle of the elbow - a radiographic study of 422
patients. Clin. Orthop. 119, 194-196. 1976.

9. Berg E.E., Dholl D. Radiography of the medial elbow ligaments. J. Shoulder Elbow
Surg. 1997 ;6 :528-533.

10. Childress H.M. Recurrent ulnar nerve dislocation at the elbow. J.Bone Joint Surg.
1956 ;38A :978-984.

11. Conway J.E., Jobe F.W., Glousman R.E, Pink M. Medial instability of the elbow in
throwing athlete. Treatment by repair or reconstruction of the ulnar collateral
ligament. J.Bone Joint Surg. 1992 ;74A :67-83.

12. Durey A. L’hygroma du coude du sportif. J. Traumatol.Sport 1987 ;4 :96-101

13. Gallagher M.A., Cuomo F., Polonsky L., Berliner K., Zuckerman J.D. Effects of age,
testing speed, and arm dominance on isokinetic strength of the elbow. J.Shoulder
Elbow Surg. 1997 ;6 :340-346.

14. Gallay S.H., Richards R.R., O’Driscoll S.W. Intrarticular capacity and compliance
of stiff and normal elbows. Arthroscopy 1993 ;9 :9-13.

15. Hämäläinen M., Ikävalko M., Kammonen M., Lehtimäki M., Niskanen R., Veli-Matti
N., Repo A., Kaarela K. Epidemiology of the elbow joint involvement in rheumatoid
arthritis. In Rheumatoid surgery of the elbow, Hämäläinen M., Hagena F-W. eds,
Karger, Basel, 1991, vol 15, pp1-5.

16. Hamilton C.D., Glousman R.E., Jobe F.W., Brault J., Pink M., Perry J. Dynamic
stability of the elbow : Electromyographic analysis of the flexor pronator group and
the extensor group in pitchers with valgus instability. J. Shoulder Elbow Surg. 1996 ;
5 :347-354.
17. Hartz C.R ., Linscheid R.L., Gramse, R.R., Daube J.R. The pronator teres
syndrome : compressive neuropathy of the median nerve. J. Bone Joint Surg 1981 ;
63A :885-890.

18. Hoppenfeld S. Examen clinique des membres et du rachis. Masson, Paris, 1984,

19. Hotchkiss R.N., Weiland A.J. Valgus stability of the elbow. J. Orthop.Res. 1987 ;5 :

20. Josefsson P.O., Johnell O., Gentz C.F. long-term sequelae of simple dislocation of
the elbow. J.Bone Joint Surg. 1984 ;66A :927-930.

21. Judet T., Bon T., Garreau de Loubresse C., Piriou P. Coude raide d’hyperutilisation
et conflit péri-olécranien. Communication SOFCOT 1997 ; abstract n°3.

22. Kendall H.O., Kendall F.P., Wadsworth G.E. Les muscles : bilan et étude
fonctionnelle. Maloine, Paris, 1971.

23. King G.J., Morrey B.F., An K-N. Stabilizers of the elbow. J. Shoulder Elbow Surg.
1993 ;2 :165-174.

24. Lister G.D., Belsole R.B., Kleinert H.E. The radial tunnel syndrome. J. Hand Surg.
1979 ;4 :52-59.

25. Marmor L., Lawrence J.F., Dubois E.L. - Posterior interosseous nerve palsy due to
rheumatoid arthritis. J. Bone Joint Surg. Am. 49-A, 381-383. 1967.

26. Mehlhoff T.L., Noble P.C., Bennett J.B., Tullos H.S. Simple dislocation of the elbow
in the adult. J.Bone Joint Surg. 1988 ;70A :244-249.

27. Millender L.H., Nalebuff E.A., Holdsworth D.E. - Posterior interosseous-nerve

syndrome secondary to rheumatoid synovitis. J. Bone Joint Surg. Am. 55-A, 753-757.

28. Mondoloni P., Vandenbussche E., Peraldi P., Augereau B. Instabilité du coude
après cal vicieux supra-condylien huméral en cubitus varus. Rev.Chir.Orthop. 1996 ;
82 :757-761.

29. Morrey B.F., Chao E.Y.S., Hui F.C. Biomechanical study of the elbow following
excision of the radial head. J.Bone Joint Surg. 1979 ;61A :63-68.

30. Morrey B.F., Askew L.J., An K-N., Chao E.Y.S. A biomechanical study of normal
functional elbow motion. J.Bone Joint Surg. 1981 ; 63A :872-877.

31. Morrey B.F., An K-N. Functional anatomy of the ligaments of the elbow.
Clin.Orthop 1985 ; 201 :84-90.

32. Morrey B.F. Anatomy of the elbow joint. In The elbow and its disorders. B.F.
Morrey ed, W.B. Saunders, Philadelphia, 1985, pp7-42.

33. Morrey B.F., An K-N., Chao E.Y.S. Functional evaluation of the elbow. In The elbow
and its disorders. B.F. Morrey ed, W.B. Saunders, Philadelphia, 1985, pp73-91.

34. Morrey B.F. Applied anatomy and biomechanics of the elbow joint. The American
Academy of Orthopaedic Surgeons, vol 35, The C.V. Mosby, St Louis, 1986, pp59-68.

35. Morrey B.F. Physical examination of the elbow. In M. Post, Physical examination
of the musculoskeltal system, Year book medical publisher, Chicago, 1987 , pp56-

36. Narakas A.O., Bonnard C. L’épicondylalgie : son traitement conservateur et

chirurgical. In Traité de chirurgie de la main, R. Tubiana ed, Masson, Paris, 1991,

37. Nestor B.J., O’Driscoll S.W., Morrey B.F. Ligamentous reconstruction for
posterolateral rotatory instability of the elbow. J. Bone Joint Surg. 1992 ;74A :1235-

38. Nirschl R.P. Soft-tissue injuries about the elbow. Clin.Sports med. 1986 ;5 :637-

39. Norwood L.A., Shook J.A., Andrews J.R. Acute medial elbow ruptures. Am.J.Sports
Med. 1981 ;9 :16-19.

40. O’Driscoll S.W., Bell D.F., Morrey B.F. Posterolateral rotatory instability of the
elbow. J. Bone Joint Surg. 1991 ;73A :440-446.

41. O’Driscoll S.W., Morrey B.F., Korinek S., An K-N. Elbow subluxation and
dislocation. A spectrum of instability. Clin.Orthop. 1992 ;280 :186-197.

42. O’Driscoll S.W. Elbow instability. Hand Clin. 1994 ;10 :405-415.

43. Olsen B.S., Søjbjerg J.O., Dalstra M. Sneppen O. Kinematics of the lateral
ligamentous constraints of the elbow joint. J. Shoulder Elbow Surg. 1996 ;5 :333-

44. Olsen B.S., Søjbjerg J.O., Nielsen K.K., Vaesel M.T., Dalstra M., Sneppen O.
Posterolateral elbow joint instability : the basic kinematics. J. Shoulder Elbow Surg.
1998 ;7 :19-29.

45. Osborne G., Cotterill P. Recurrent dislocation of the elbow. J. Bone Joint Surg.
1966 ;48B :340-346.

46. Porter B.B, Richardson C., Vainio K. Rheumatoid arthritis of the elbow : the
results of synovectomy. J.Bone Joint Surg. 1974 ;56B :427-437.

47. Regan W.D., Morrey B.F. Fractures of the coronoid process of the ulna. J. Bone
Joint Surg. 1989 ;71A :1348-1354.

48. Regan W.D., Korinek S.L., Morrey B.F., An K.N. - Biomechanical study of
ligaments around the elbow joint. Clin. Orthop. 170-179. 1991.

49. Søjbjerg J.O., Ovesen J., Gundorf C.E. The stability of the elbow following excision
of the radial head and transection of the annular ligament. An experimental study.
Arch.Orthop.Trauma.Surg. 1987 ;106 :248-250.

50. Søjbjerg J.O., Helmig P., Kjaersgaard-Andersen P. Dislocation of the elbow : an

experimental study of the ligamentous injuries. Orthopedics 1989 ;12 :461-463.

51. Souter W. Le traitement chirurgical du coude rhumatoïde. Cahiers

d’enseignement de la SOFCOT 1987 :159-172.

52. Souter W. Surgery of the rheumatoid elbow. Ann.Rheum.Dis. 1990 ;49 :871-882.

53. Tullos H.S, Schwab G., Bennett J.B., Woods G.W. Factors influencing elbow
instability. The American Academy of Orthopaedic Surgeons, vol 30, The C.V. Mosby,
St Louis, 1981 ; pp185-199.

54. Tullos H.S., Bennett J.B., Shepard D., Noble P.C., Gabel. Adult elbow dislocations :
mechanism of instability. The American Academy of Orthopaedic Surgeons, vol 35,
The C.V. Mosby, St Louis, 1986, pp69-82.

55. Volz R.G., Morrey B.F. The physical examination of the elbow. In The elbow and
its disorders. B.F. Morrey ed, W.B. Saunders, Philadelphia, 1985, pp62-72.

56. Witvoet J., Tayon B. La luxation récidivante du coude (à propos de 6 cas).

Rev.Chir.Orthop 1974 ;60 :485-495. (mise à jour sur 11 cas aux entretiens de Bichat,

Physical examination of wrist instabilities

Christian Dumontier
Institut de la Main, 6 square Jouvenet, 75016 Paris, France
& Consultant, Upper limb surgery Dpt, hôpital St-Antoine,
184 rue du Faubourg St-Antoine, 75012 Paris, France

This version is an update of the original version as Lou A. Gilula was very kind to correct the grammatical as well as the orthographic
mistakes of the text. He also took of his time to partly rewrite this text. His advices and comments were very useful and I want to deeply
thank him for his contributions. Some remarks have also been made by David M. Lichtman who sent me more details on the performance
of the midcarpal shift test that he described. His drawings were not included in the original version of this work published in French in the
number 49 of Maîtrise Orthopédique.

Although 35 years ago McLaughlin stated that wrist sprains were rare injuries, Linscheid and Dobyns's publication of 1972 [1] was the
beginning of an enormous number of publications reporting either experimental or clinical work on wrist ligamentous injuries. Most of
these injuries are diagnosed late, as early clinical signs are not specific, and plain X-rays are usually considered normal. Late radiographic
deformations are no longer a problem for the orthopedic surgeon, but are indicative of advanced lesions whose treatment is still
controversial. The main difficulty remains the chronically painful wrist despite normal X-rays. To deal with this problem, as frequent as it
is frustrating, the surgeon must be aware that there may be a ligamentous wrist injury needing surgical treatment. Although all authors
stress the importance of physical examination of the wrist, precise clinical details are seldom available. In this paper we will only deal with
the clinical examination of ligamentous wrist injuries and will not discuss imaging techniques or therapeutic indications. Neither will we
discuss soft tissue injuries of the wrist, which are much more frequent and may lead to a differential diagnosis in each case.

Ligamentous wrist injuries may arise in four joints: distal radioulnar, radiocarpal, midcarpal and carpometacarpal joint. Here we will the
keep Anglo-Saxon term of wrist instability because it accurately describes the abnormal motion or alignment between the bones, although
the correct term is "severe sprain". Wrist instability may appear after multiple microtraumas as in athletes, but it usually appears after a
single trauma.
Physical examination is fundamental, as it focuses attention on ligamentous lesions which if left untreated, we believe will lead to
degenerative arthritis. The surgical alternatives for degenerative arthritis carry high morbidity, and long-term follow-up is not yet available.
Early surgical treatment allows "conservative" treatment, with lower morbidity, and satisfactory results. However, as pointed out by Gilula,
this is still debatable as indications for surgery are not clear, and may lead to unsatisfactory results. It is a belief that early surgery may be
easier to perform and will give better results.
Physical examination of an unstable wrist includes two types of evaluation: firstly, an objective evaluation of wrist function by measuring
mobility and strength and conducting functional tests, and secondly, an attempt to delimit the painful zone and/or to detect abnormal
motion between the carpal bones, in order to determine the diagnosis. Knowledge of wrist instability mechanisms and of bone and
ligamentous wrist anatomy is a prerequisite for clinical examination, but will not be discussed here.

Mobility :

To test pronosupination, the patient is asked to keep his or her elbows close to the body and to turn the palm up and down alternatively. One
arm of the goniometer is placed parallel to the axis of the humerus, and the other along the distal part of the forearm (Figure 1 & 2). One
should avoid measuring pronosupination with a stick in the patient's hands, as the pronosupination mobility is increased by the passive
rotatory mobility of the carpus, which may be as high as 40°. If the neutral prono-supination position is defined as zero (with the elbow
flexed and maintained against the chest, the thumb must be raised up), normal pronation varies between 60 and 90°, and normal supination,
between 45 and 80°. Flexion-extension mobility is measured by placing the goniometer on the palm for wrist extension, and along the
dorsum of the hand for wrist flexion, over the axis of the third metacarpal bone (figure 3 & 4). Normal values vary among individuals and
may reach 85° of flexion or extension. Ulnar inclination varies between 30 and 45°, and radial inclination, between 15 and 25°. Both
inclinations are measured with one arm of the goniometer along the axis of the forearm, and the other along the axis of the third
metacarpal, with the wrist in the neutral position of flexion or extension. These methods are simple and reproducible [2]

Figure 1: Figure 2: Figure 4:

Measurement of pronation: Measurement of Masurement of extension:
The vertical arm of the supination. The The goniometer is placed
goniometer is placed in the horizontal arm is anteriorly on the wrist.
axis of the arm and the placed on the volar
horizontal arm on the dorsal surface of the wrist.
surface of the wrist, but not
the hand.

Measurement of strength :

This should be done with a Jamar dynamometer, which is considered an international reference. Measurements should be done, either using
each of the five handle positions, which is time-consuming, or using only one handle position, with three successive measurements. There
are no standard values, and the contralateral hand serves as reference. The mean of three different measurements with maximum muscular
contraction is noted. Usually, the curve for a single handle position is horizontal or slightly descending. Rapid alternating measurements
changing from one hand to the other prevent patients from controlling their contraction and may reveal the absence of maximum
contraction [3,4]. The dominant hand is usually 5 to 10% stronger than the non-dominant hand. The principles of key-pinch measurements
are identical, but the measurements are less revealing for wrist instability. The functional tests will not be described in detail here as they
are not conducted by French surgeons. Functional tests are usually performed by physical therapists who, in France, usually have a private
practice. They are rarely done in practice, and never appear in scientific papers made by surgeons.


The normal wrist :

The key to correct examination of the wrist is precise location of the symptoms relating to the underlying anatomical structures, i.e., bones,
articular spaces, ligaments or tendons. As in all clinical examinations, the most painful area is examined last. Comparative wrist
examination is the rule, as there are no criteria of normality.

Conditions of examination :

The wrist must be examined with the forearm free of clothing and jewelry. For a satisfactory examination, the patient and the examiner
should be comfortably seated. The ideal solution is to place the patient's forearm on a narrow examination table whose height may vary. In
clinical practice, the easiest solution is to sit very close to the patient so that his or her hand rests on the examiner's knee, with the patient's
elbow resting on his thigh (Figure 5). As pointed out by Lou Gilula, this may be a problem for potential sexual accusations in the USA.
Buying a narrow table is probably less expensive than lawyer`s fees. Physical examination usually begins on the dorsal surface of the wrist,
with pronation of the forearm and wrist flexion, whereas the ulnar surface of the wrist is examined during maximum elbow flexion. To
move the wrist, the proximal hand of the examiner stabilizes the forearm while the distal hand moves the wrist. For palpation, the examiner
stabilizes the wrist with both hands and uses his (her) thumbs to palpate the anatomical structures.

Figure 5:
A "practical" position for wrist examination.

Cutaneous projection of the anatomical structures

A beauty (the richness) of wrist examination is due to the fact that almost all bony, articular, tendinous or vascular structures may be
palpated through the skin that covers it. To be compete, the physical examination should be methodical and whichever structure is
examined first, the examination should cover the entire wrist.

Dorsal surface: Proximal to the wrist, proceding from the radius to the ulna it is easy to identify the radial styloid. One cm proximal you
will palpate the sharp bony ridge which limits the first extensor compartment. More ulnar is a dorsal bump on the distal radius which is
Lister's tubercle, around which passes ulnarly the extensor pollicis longus tendon (figure 6 & 7). Closer to the ulna and ulnar to Lister`s
tubercle, one can feel the flat dorsal surface of the radius and the ulnar head which protrudes in pronation. On the ulnar side of the wrist,
the ulnar styloid can be palpated dorsally in supination, at the ulnar and volar surfaces in pronation and on the ulnar side of the wrist in
neutral rotation [4-6].

Figure 6: Figure 7:
To examine a wrist Main palpable bony structures on
correctly, one should the dorsal surface of the wrist
mentally project the (redrawn after [4].
bones onto the skin.

At the level of the carpus, the anatomical snuffbox is easy to locate radially: it is limited radially by the extensor pollicis brevis and the
abductor pollicis longus and ulnarly by the extensor pollicis longus. The scaphoid lies at the bottom of the snuffbox, with the radial artery
crossing over it. In radial deviation the scaphoid disappears dorsally and one can palpate the scaphotrapezial joint palmarly (figures 8 & 9).
Dorsally, at the distal end of the scaphoid there is a groove in which the examiner can place an index finger to palpate the trapezoid along
the axis of the second metacarpal, and the trapezium along the axis of the first metacarpal [4]. The radial part of this groove, just ulnar to the
extensor pollicis longus tendon, is what is termed the STT entry point (scaphotrapeziotrapezoidal) for mid-carpal arthroscopy (Figure 10).
In the middle of the dorsal surface of the carpus, one centimeter distal to Lister's tubercle, lies the scapholunate interval. As pointed out by
Lou Gilula, the scapholunate interval cannot be palpated unless the wrist is flexed. It would be usually better to say that the scapholunate
interval can be palpated just distal to the dorsal rim of the radius at the level of Lister`s tubercle, with flexion of the wrist. Flexion moves
the lunate dorsally out of the lunate fossa as shown figure 5. Just radial to that point, the proximal pole of the scaphoid can be palpated if
the wrist is in flexion. Ulnar and distal to the scapholunate space lies a concavity which corresponds to the neck of the capitate (Figure 11).
(French anatomists use the term “the crucifixion groove” as it represents the place where you should place your nails if you plan to crucify
somebody...) When the wrist is flexed, the lunate and the head of the capitate are more easily palpable (Figure 12). Slightly radial to the
neck of the capitate and one cm distal to the scapholunate interval is the radial entry point of the midcarpal space. The prominence of the
third metacarpal base, the third metacarpal styloid, is located one to one and a half cm distal to that point, between the capitate and the
trapezoid. It is more or less developed depending on the individual and may sometimes be hidden by the insertion of the extensor carpi
radialis brevis tendon. When the wrist is in neutral position, with the third metacarpal in the axis of the radius i.e. without flexion or
extension or radial or ulnar deviation, the ulnar head, triquetrum, hamate and fifth metacarpal form a continuous line on the ulnar side of
the wrist

Figure 8: Figure 9:
The scaphoid lies at the The cutaneous projection of the
bottom of the anatomical
snuffbox and distal to it lies
the scaphotrapezial joint.
Palpation of bony structures
varies during radial and ulnar
deviation. anatomical snuffbox.

Figure 11: The posterior surface of the waist

of the capitate is palpable through a Figure 12:
depression easily found in the midportion of Wrist flexion allows palpation of
the dorsal surface of the wrist. the head of the capitate and the
posterior horn of the lunate.

Figure 10:
The midcarpal joint can be palpated through
the groove between the scaphoid and the
trapezium and trapezoid bones.

The triquetrolunate joint and triquetrum may be palpated during radial deviation of the wrist. The triquetrum is palpated just distal to the
ulnar head and disappears with ulnar deviation. The triquetrohamate space whose mobility can be appreciated lies distal to the dorsal
tubercle of the triquetrum (Figure 13). On the ulnar side of the wrist lies the "ulnar snuffbox" between the extensor and the flexor carpi
ulnaris tendons. At the base of this snuffbox one can palpate the triquetrum during radial inclination, as well as the triquetrohamate joint
distal to it, which is a drainage portal for mid-carpal arthroscopy (Figure 14).

Figure 14:
The posterior side of the triquetrum is
easily found as one can palpate the
insertional crest of the posterior
radiotriquetral ligament. Just distal to
the crest lies the triquetrohamate space.

Figure 13: The ulnar "anatomical


The palmar surface : The bony structures on this surface are too deep to be palpated. However, it is possible to palpate not only the radial
and ulnar styloid processes but also, radially, the trapezial ridge which lies at the base of the thenar eminence, as well as the scaphotrapezial
space and proximal to the distal tuberosity of the scaphoid [5,6] , when the wrist is in extension (Figure 15). Ulnarly, the pisiform is easily
palpated, just distal to the distal wrist crease. The hamate hook (hamulus ossi hamatum) lies just along the radial edge of the pisiform, on a
line from the pisiform to the second metacarpal head [7] (Figure 16). The articular spaces of the carpus are not accessible to palpation, but
the radiocarpal joint is located at the level of the middle part of the proximal wrist flexion crease, while the midcarpal joint is located at the
level of the middle of the distal flexion wrist crease [8].

Figure 16:
The hamulus ossi hamatum
Figure15: (hook of the hamate) is
Main palpable bony palpated deeply, 2 cm below
structures on the anterior the pisiform bone, on a line
side of the wrist joining the pisiform to the head
(redrawn after [4]. of the second metacarpal bone.

Soft tissues :

Using the bony structures as landmarks, almost all the tendinous, vascular and neural structures can be palpated. Their palpation is part of
the normal examination of the wrist but will not be described in detail here. A complete examination is also performed as needed.


Most patients consult because of wrist pain, weakness or instability. These symptoms correspond to a wide pathological spectrum, from
mild wrist sprain to chronic dissociative instability. Clinical instability varies greatly from one individual to another [9]. Examination of
normal structures may sometimes be painful which explains why wrist examination should be compared to the opposite wrist. This
examination must be rigorous, methodical, and systematic and includes five steps:
The first step is the description by the patient of the circumstances of the injury because knowledge of the impact point and exact wrist
position during the impact may be indicative of the type of injury. Unfortunately, however, most patients are unable to recall the precise
position of their wrist at the time of injury.

The second step is the patient's description of symptoms. Pain is the most frequent complaint and its characteristics, times of occurrence
and alleviating or aggravating factors should be defined. Dynamic instability is painful during effort and may cause pain and/or synovitis
which usually resolves in one or two days [9]]. With dynamic instability pain is generally of the same intensity from one day to another and
may never completely disappear [7]]. Most authors consider that the exact location of pain is the most precise and specific sign for
diagnosis. The clinician should ask the patient to localize the pain with one finger. The correlation between arthrographic radial perforation
and radial pain is weak (49%) but is better for ulnar pain (88%) [10,11]]. This correlation is weaker still with bilateral wrist arthrograms, as
almost 75% of the arthrographic perforations are bilateral and symmetric [12]]. During this part of the examination a simple test for pain
can be conducted, like vigorously shaking patient's hand, but this is only of interest if the result is negative [13]. It is also a good sign if pain
is induced by passive movements performed with slight traction, or by provocative maneuvers, or if pain disappears after injection of local
anesthetics [3,14].
Loss of strength is usually the second most frequent complaint. It is often caused by fear of pain but if the pain is moderate, loss of strength
is a good sign of wrist instability.

The third step in wrist examination is palpation and testing of each articulation and will be discussed in detail later. This step comprises a
search for pain, abnormal motion, crepitus or swelling of the palpated zone.

The fourth step is an attempt to define any click or clunks reported by the patient. Benign wrist popping must be eliminated first. It is a
frequent symptom consisting of a sharp noise which occurs during active movements. It is transient, with a period of latency, and reflects
variations in the saturation of intra-articular gases. When a joint is stressed in certain positions gas in normal solution in the body can “pop”
in the joint, causing a “pop” sound, much like some people produce when they “pop” or “crack” their fingers [Lou A. Gilula, personal
communication]. Clicks or snaps are visible, audible and usually palpable, but there is a wide range of clinical signs which include a
sensation of internal derangement and instability, and clicks, snaps or clunks. The patient is asked to reproduce the triggering movement,
and the examiner then tries to reproduce the symptoms. With the patient's elbow resting on the table, the examiner holds the patient's hand,
and by exerting axial pressure, realizes movements of flexion, extension, radial and ulnar inclination, and rotation. The so-called specific
maneuvers described in the literature are designed to increase the mechanical constraints on ligaments in order to reveal pain, snaps or
abnormal motion between bones. However, most of these maneuvers have not been validated for specificity or sensitivity and are only
indicative of a potential pathology. During this stage of the examination the patient's laxity should be evaluated. This includes general
laxity, i.e. elbow recurvatum and elbow valgus, the possibility of touching the forearm with the thumb during maximum flexion of the
wrist, hyperextension of the MP joints, and also wrist laxity including:

(1) palmar sag of the wrist that you can see on the ulnar side which disappears if dorsal directed pressure is applied to the pisiform (Figures
17 a & b),
Figure 17a,b: Spontaneous ulnar sag of the wrist in a lax young girl. By pushing on the pisiform this ulnar sag will be reduced. This test
differentiates laxity from a posterior radioulnar dislocation.

(2) the anterior and posterior drawer tests in which the examiner exerts axial traction while holding the patient's forearm with one hand and
the metacarpal heads with the other. In that position, anterior (palmar) or posterior (dorsal) displacement is usually of little amplitude. Lax
patients display true subluxation which is usually located in the mid-carpal joint and is associated with a flexion deformity of the first row
(VISI) during the anterior drawer test and a dorsal (DISI) deformity during the posterior drawer test [15]Jeanneret has shown that in slight
ulnar deviation, most patients exhibit a positive anterior drawer test, which disappears in neutral, radial or full ulnar deviation [16]. As pointed
out by Gilula, the lunate tilts more dorsally in full ulnar deviation, however the anterior drawer test disappears in that position .

An anterior drawer test (18a). Usually the posterior drawer test is

always more important than the anterior. The first row is placed in
flexion during the anterior drawer test (18b, left side) and in
extension during the posterior drawer test (18c, right side). In lax
patients the distal carpal row subluxes palmar to the proximal
carpal row.

(3) A mid-carpal, painless snap is present in one out of four individuals. This snap occurs when a wrist in slight flexion subjected to axial
compression, is moved from radial to ulnar deviation. This movement places the first row in flexion and anterior translation. The mid-
carpal snap occurs with the sudden displacement of the first row from the flexion position to the extension position (which also includes a
posterior translation) that it should have in ulnar deviation. Some patients can reproduce this snap with active movements. The Lichtman's
test or midcarpal shift test has been described by Lichtman [34]. To perform this test on the right wrist, the forearm is stabilized with the
examiner's left hand and held in pronated position (figure 35a). With the patient's wrist in about 15° of ulnar deviation, the examiner's right
hand grasps the patient's right hand and, with the thumb, exerts palmarly directed pressure at the level of the distal capitate (figure 35b).
The distal row then palmarly translates and the proximal row assumes a VISI position (figure 35c). The wrist is then simultaneously axially
loaded and ulnarly deviated resultng in sudden reduction of the midcarpal joint (figure 35d). A similar test exists for ulnar to radial
deviation but its results are not usually as clear [8]. As pointed out by Lichtman, the midcarpal shift test is subjective and often difficult to
quantify. Therefore he proposed a grading system based on 3 criteria; the force required to produce palmar translation, the quality of wrist
clunk, and whether or not the patient can reproduce the clunk spontaneously. Grade I means no palmar translation and no clunk, grade II
minimal palmar translation and minimal clunk, grade III moderate palmar translation and moderate clunck, grade IV maximum palmar
translation and significant clunk, grade V self-induced palmar translation and self-induced clunk. Grade I-IV are considered to be
increasing degrees of "normal" midcarpal laxity and can only be reproduced by the examiner. A grade v score represents the pathologic
condition of midcarpal instability (Lichtman, personal communication). Hyperlaxity is common in patients with wrist instability and is
considered an aggravating factor. If the radiographs are normal, the fifth step is to repeat the entire examination after an infiltration of
local anesthetics in the painful zone. As in all clinical examinations, it may be necessary to trick a distressed or uncooperative patient. What
are known as substitution maneuvers and distraction techniques are very helpful. They include simultaneous testing of multiple or bilateral
muscles, and simultaneous exploration of different areas [13].

Before making a diagnosis, one should remember that although many lesions cause wrist pain, the frequency of ligamentous injury is small.
Jones has shown that the frequency of scapholunate lesions was not more than 5% after wrist trauma [17]. Even after a single trauma, many
painful wrists are due to a dorsal ganglion or to a dorsal impaction syndrome without instability [18]. On the other hand, although
intracarpal injuries are only asymptomatic in 2% of cases, suggestive symptomatology was only correlated to anatomical lesions in half the
cases reported by Truong et al. [10], and exhibited no statistical correlation according to Cantor [12]. Thus, out of 87 wrist arthroscopies for
pain suggestive of wrist instability, 51 had an intra-articular ligamentous tear and 18, a TFCC defect. However, more than 70% of these
defects were partial and involved no instability [19]. Others have found with arthroscopy that patients presenting with carpal instability
symptoms had 2.7 ligamentous defects in the wrist, suggesting that our actual classification of lesions is inadequate [20]. However it is still
uncertain which ligamentous defects are significant for symptoms or are just attritional due to age, prior injury, or degeneration. Lou A.
Gilula suggests that we should not use the term “tear” as that implies traumatic incident , which is not always the case..

Scapholunate instability:

The mechanism of scapholunate injury includes a fall onto a hyperextended wrist with the forearm in pronation and the impact point on the
thenar eminence [7,21]. Radial pain and progressive loss of strength are usual [21]. Loss of mobility appears much later. Patients may
sometimes complain of a snapping wrist which usually occurs during the passage from radial deviation to neutral with the wrist in flexion.
In ulnar deviation, the snap represents the action of the scaphoid on the lunate bone and the sudden correction of the proximal carpal row
into dorsiflexion. With wrist flexion, a snap may represent penetration of the capitate into the scapholunate interval (rare), or the dorsal
subluxation of the scaphoid on the posterior margin of the radius [ [5,6]. ]. Capitate popping is rare in Gilula's experience, and Gilula also
pointed out that in the great majority of the cases with popping that he sees, fluoroscopic exam is normal and he does not know what
ligaments or anatomic structures cause the popping. The popping seems to be related to moving of tendons or other soft tissue structures.
Ulnar inclination combined with anterior translation places a load on the dorsal part of the scapholunate ligament and a snap may suggest
partial tears [Masquelet, personal communication]. The snap may be reproduced during ulnar deviation combined with axial compression
[22]. The various provocative maneuvers reported in the literature include the following:

(1) The synovial irritation sign of the scaphoid. To elicit this sign, pain is induced by exerting pressure on the scaphoid through the
anatomical snuffbox [7,23] (Figure 19). This sign is usually positive in patients with scaphoid instability, but its specificity is very low.
Figure 19: Figure 20:
The synovial irritation test is The scapholunate ballottement test.
performed by palpating the
bottom of the anatomical

(2) The scaphoid bell sign. This is performed by palpation of the scaphoid tuberosity anteriorly through the radial groove while placing the
index finger in the anatomical snuffbox. With ulnar deviation of the wrist, the anterior protrusion of the distal scaphoid tuberosity
disappears and the proximal pole appears in the snuffbox. With radial deviation, the proximal pole disappears in the snuffbox and the
protrusion of the distal scaphoid tuberosity reappears in the radial groove. Any disruption of this normal mechanism is suggestive of
instability, but the sensitivity of this test seems very low [8].

(3) The scapholunate ballottement test. This test is designed to highlight any abnormal motion between the scaphoid and lunate bones.
With one hand the examiner holds the scaphoid between his thumb (placed distally over the scaphoid tuberosity on the palmar side) and
index finger (placed posteriorly and proximally over the proximal pole of the scaphoid). The other hand holds the lunate (Figure 20). The
hands then move in opposite directions and appreciate the ballotement between the two bones. It may be difficult to appreciate instability as
the normal laxity of the scapholunate joint varies greatly among individuals [5,6]. However, if the test induces pain, this is a good sign. This
test, as all tests, may be compared to the opposite wrist to appreciate normal variations. Scapholunate ballottement is more marked when
the wrist is in slight flexion, and, in this position, dorsal protrusion of the second row is sometimes visible [4,22]. Flexing the wrist also
brings the lunate more dorsal and distal to the dorsal rim of the radius making it easier to palpate the lunate. Another technique to palpate
the scapholunate interval is to place the index finger on the dorsal and distal pole of the lunate and then move the index finger radially
while moving the wrist in flexion and extension. One can sometimes feel a groove corresponding to the scapholunate interval, or more
often a slight protrusion of the proximal pole of the scaphoid [7]. The limitations of these tests are connected with the difficulty to hold the
lunate bone correctly.

(4) The wrist-flexion finger-extension maneuver was described by Watson. With the elbow resting on the table, the wrist is placed in
flexion and the patient is asked to extend the fingers. Application of pressure on the nails may reveal pain in the scapholunate interval
[10,13] (Figure 21).

Figure 21:
The wrist-flexion finger-extension maneuver.
This maneuver induces loads into the carpus
that arouses pain at the scapholunate space.

(5) Watson's test or the scaphoid shear test [9,24]. The examiner and patient face each other as for arm wrestling. The examiner's fingers are
placed dorsally on the distal radius, while the thumb is placed on the palmar distal tuberosity of the scaphoid. The other hand holds the
metacarpals. Firm pressure is applied to the palmar tuberosity of the scaphoid while the wrist is moved in ulnar deviation which places the
scaphoid in extension. While the wrist is moved in radial deviation the scaphoid cannot flex, as it is blocked from flexing by the examiner's
thumb. In case of scapholunate tear, or in lax wrist patients, the scaphoid will move dorsally under the posterior margin of the radius and
will reach the examiner's index finger, thus inducing pain (Figure 22). Sometimes this test may only be painful, without any perception of
dorsal scaphoid displacement. When pressure on the scaphoid is removed, the scaphoid goes back into position with what Watson described
as a "thunk" (a clunk) [24]. In certain patients, the absence of normal mobility compared to the uninjured wrist may be due to swelling
and/or synovitis [9]. To avoid false-positive testing, the examiner should first place his fingers on the posterior surface of the scaphoid to
detect spontaneous pain. Even though the Watson's test is the best known for scaphoid subluxation, its sensitivity and specificity are low. In
two studies, this test proved positive in 20% of normal individuals [23,24]. In another study, the prevalence of a painless snap when
removing thumb pressure was 32%, thus showing patient laxity. However, 14% of these lax patients were positive only in one wrist [25].
Lane suggested modifying the Watson's test by moving the scaphoid only from an anterior to a posterior position (he called it the Scaphoid
shift test). This modification would enhance the test's sensitivity by using simple movements [26].

Figure 22: The Watson's test.

Lunotriquetral instability:

Lunotriquetral instability may appear after a hyperpronation injury [27,28], but more often after a hyperextension injury with an impact on
the ulnar side [7,29,30]. Ninety per cent of patients complain of ulnar pain, and lunotriquetral joint palpation is usually painful [14,29,30].
Active prono-supination movements against resistance are painful if the resistance causes twisting of the carpus [29]. A feeling of
instability or loss of strength is present in rare cases. A snap or clunk may be observed in half of the patients during ulnar deviation or
extension [4-6,14,28,30]. The following tests are considered "specific":

The lunotriquetral ballottement test or Reagan's test (also called the Shuck or shear test, depending on the authors): as in the
scapholunate ballottement test, the clinician holds the lunate bone between his thumb and index finger with one hand, and moves the
triquetrum with the pisiform dorsal and palmar (Figure 23). The aim is to appreciate instability (very difficult) and above all the arousal of
pain [30-32]. The sensitivity of this test varies from 33 to 100%, depending on the authors, and its specificity is still unknown.

Kleinman's shear test (which some authors call the shuck test!) (Figure 24): ): With the patient's forearm in a vertical position, the
examiner places one finger on the posterior part of the lunate and with his contralateral thumb placed palmar, pushes the pisiform dorsal
which arouses pain in the lunotriquetral joint [31,32]. This test might be more sensitive and more specific than the Reagan's test [32].

Figure 23: Figure 24: Figure 25:

The lunotriquetral The Kleinman's test. The ulnar snuffbox compression
ballottement test (Reagan's test (Linscheid's test)

The ulnar snuff box compression test or Linscheid's test. This test may be the least specific according to Kleinman (Figure 25). The thumb
placed on the ulnar side of the triquetrum exerts an axial pressure directed toward the lunate, which arouses pain [31 ].

The raised triquetrum test was recently proposed by Zradkovic and Sennwald (personal communication). The examiner holds the
patient's hand proximal to the wrist and places his thumb on the triquetrum. From the neutral position, without flexion or extension, he
performs radial and ulnar deviation movements and appreciates the dorsal and palmar movements of the triquetrum, which should be
compared to those of the other wrist (Figures 26 a,b,c). The sensitivity and specificity of this test are still unknown, as are the anatomical
lesions which cause the test to be positive. As pointed out by Gilula, the triquetrum is very prominent or dorsal with radial deviation, and
moves palmarly and may even disapear with ulnar deviation. On plain radiographs, the triquetrum is located "onto" or proximal on the
hamate with radial deviation (superposed), and "lateral" or ulnar to it with ulnar deviation (juxtaposed) [Laredo, personal communication].

Figure 26: The raised triquetrum test (Photos G. Sennwald). In Fig 26a, the examiner places the wrist in radial deviation while palpating
the triquetrum. He then moves the wrist in neutral (26b) and ulnar (26c) deviation to appreciate the depression of the triquetrum with
ulnar deviation and prominence of the triquetrum with radial deviation that should be compared to the contralateral wrist.
Global first row instability:

This might reflect the spontaneous resolution of intra-carpal dislocations, thanks to the persistent efficacy of certain extrinsic ligaments [33].
Clinical examination reveals scapholunate and lunotriquetral signs of instability together with pain, loss of strength and mobility, and pain
during extreme movements of the wrist [33]. Associated lesions of the TFCC are frequent [33].

Midcarpal instability:

The classification of these lesions is still controversial and they probably include distinct entities. According to some authors, the main
lesion is ulnar and located on the hamatotriquetral joint [20,31,34,35]. Others believe that the lesion is located in the lunocapitate space
[36,37]. Lastly Schernberg and Masquelet both reported ligamentous lesions [8,15,38]. The mechanism of such injuries is still unknown, but it
seems that the radial, traumatic lesions are distinct from the central or ulnar lesions which are the most frequent, but do not seem to be
triggered by trauma [31,34]. Hyperflexion or hyperpronation with extension constitute an aggravating mechanism [15,39]

Ulnar or central mid-carpal instability :

Most patients present with bilateral laxity, and one out of three has a painless mid-carpal snap in the healthy wrist [34]. Pain is located on
the ulnar side and leads to functional impairment, loss of strength and loss of mobility [15,39]. The mid-carpal snap is the most revealing
sign. It appears when patients make wrist movements with axial compression and is usually visible and audible (figure 27 a & b).

Figure 27a,b :
Radiographs of a
midcarpal instability in
which the proximal
row suddenly moves
from a flexed (27a) to
an extended position
(27b). The proximal
carpal row is flexed in
Fig 27a as
demonstrated by the
ring sign of the
scaphoid; the lunate
with its posterior,
small and sharp horn is
clearly visible; and the
triquetrum is
superposed to the
hamate. In Fig 27b, the
entire length of the
scaphoid is visible in
extension and the
lunate presents its
anterior, large horn.
The triquetrum is
juxtaposed on the
It is the same snap as the one found in patients with wrist laxity, but it is pathologic because it is painful [5,6,35]. This snap disappears if one
places dorsally directed stress on the palmar surface of the pisiform, as this maneuver places the first row in extension. It also disappears
during contraction of the hypothenar muscles. Other maneuvers have been described and they all provoke a snap, with shifting of one row
onto the other and a zigzag deformity of the carpus. According to Johson et al., such maneuvers cause pain with dorsal translation of the
capitate. They used fluoroscopy to verify the dorsal subluxation of the capitate [36]. ]. According to these authors, there is a traumatic lesion
of the radiocapitate ligament. Louis placed the wrist in slight flexion while he exerted a traction. He then pressed on the distal tuberosity of
the scaphoid and checked the results by fluoroscopy [37]. Louis reported that the 11 cases in his series displayed a double dorsal
subluxation of the proximal carpus over the radius, and of the capitate over the lunate [37]. As only one patient was operated, the
ligamentous lesion responsible was not clear, but Louis attributed it to weakening of the dorsal lunocapitate ligaments [37]. A
triquetrohamate shear test has been described, in which the triquetrum is held between the thumb and index finger and pressure is applied
dorsally on the palmar surface of the hook of the hamate [13] (Figure 28)

Figure 28: The triquetrohamate shear test.

Scaphotrapezial instability :
This rare injury is due to hyperextension of the wrist and thumb column, sometimes with an associated rotational injury [38,40]. The pain,
located on the anterior side of the scaphotrapezial joint, is increased by forced extension of the thumb column. Pinch weakness, limited
rotational movements of the thumb, and fatigue during writing have been reported [40]. Schernberg reported one case in which the
mechanism of injury was a forced flexion with traction loading [15,39]. A ballottement test combined with compression can be applied to the
scaphotrapezial joint. The examiner holds the trapezium and trapezoid with one hand and moves the scaphoid with the other [13]. This test
is different, but close to, the scaphoid compression test described by S.C. Chen to detect scaphoid fractures (J.Hand.Surg.Br. 1989;14:323)
and its sensibility is probably rather low for Scaphotrapezial instability .

Distal radioulnar joint (DRUJ) instability :

As the ulna is fixed, the radius is the dislocated bone, but we have kept the usual convention which describes "dislocation of the ulna". A
traumatic movement in supination is responsible for anterior DRUJ instability, while posterior DRUJ instability follows a pronation injury.
Dorsal ulnar dislocation is responsible for loss of supination and protrusion of the ulnar head. In case of dorsal ulna subluxation, the
protrusion of the ulnar head may be clearly visible when viewed laterally, and unlike what occurs in the normal wrist, does not disappear if
the injured wrist is flexed [8]. Anterior ulnar dislocation makes the dorsal skin depress and limits pronation. In anterior subluxation, the
usual protrusion of the ulnar head is reduced or disappears [8]. Pain secondary to DRUJ instability is located on the ulnar side of the wrist
and is intensified by pronation or supination. In such cases the examiner stabilizes the patient's forearm with one hand while with the other
hand, he grasps the patient's hand as if for a vigorous handshake. When the patient resists forced passive rotation, or when there is active
rotation against resistance, pain usually is elicited. If the pain is caused by compressing the ulna against the radius, it is mostly suggestive
of chondromalacia [3,15]. Patients may also complain of a snap which occurs during pronation or supination and corresponds to either
dislocation of the ulnar head or to its reduction. Radioulnar instability is tested by the radioulnar ballottement test, in which the patient's
elbow is flexed, and the examiner uses his thumb and index finger to stabilize the radius radially and the ulnar head ulnarly (Figure 29).
Normally, there is no mobility in the anterior or posterior direction in maximum pronation or supination. Pain or mobility is very suggestive
of radioulnar instability. The ballottement test must not only be done during extreme motions of pronation and supination, but also in
various intermediate pronation and supination positions, because instability may only appear in some of these positions. It is sometimes
possible to mobilize the ulnar head medially with two fingers [13]. As for other tests it is useful to compare with the opposite wrist to help
detect normal variation.

TFCC lesions are usually of degenerative origin, but may also constitute the first stage of radioulnar instability. Pain is always ulnar and is
intensified by wrist movements but not necessarily by pronation or supination. It is usually aggravated by ulnar inclination or rotational
loads: thus, in the screwdriver test, the examiner holds the patient's hand while performing screwing and unscrewing movements [13,15].
Extensor carpi ulnaris tendon dislocation is not a ligamentous injury but occurs after combined hypersupination and ulnar inclination [15].
Passive pronation and supination are usually painful and may be accompanied by a visible and palpable snap which can be reproduced by
placing the wrist in flexion and supination [7] (Figure 30).

Figure 29: Figure 30:

The radioulnar ballottement Displacement of the
test. extensor carpi ulnaris
is more visible when
the wrist is placed in
flexion and

Radiocarpal instability:

This very rare condition usually seems to follow an injury associating hyperextension, ulnar deviation and a rotational torque [41]. The most
severe lesions are sequelae of radiocarpal dislocation. Patients present with a positive ulnar drawer test and/or a positive anterior drawer
test [42,43] (Figure 31). Schernberg reported radiocarpal instability following less severe injuries caused by stretching of the posterior
radiotriquetral ligament. Patients presented with an ulnar radiocarpal drawer test which was sometimes painful [8]. For the anterior drawer
test, the examiner's thumb pushes on the dorsal surface of the triquetrum in a volar direction and then, for the posterior drawer test, the
thumb pushes dorsally on the anterior surface of the pisiform. Counter-pressure is exerted by the pulp of the index and long fingers of the
same hand, which are placed on the anterior then the posterior surface of the radius. In patients with radiocarpal instability there is also an
increased ulnocarpal sag, which is reduced during the posterior drawer test, thus confirming its radiocarpal origin. A click may sometimes
occur during ulnar deviation but it disappears if a posterior drawer test is applied [8].

Figure 31: Figure 32:

Spontaneously reduced Pisiform grinding test. This
radiocarpal dislocation. maneuver is painful in
Major ulnar translation is pisiform instability but also
present. in pisotriquetral arthritis
which is more frequent.

Pisotriquetral instability:

Pisotriquetral pain is usually of degenerative origin. A fall with the wrist in extension and ulnar deviation may induce acute instability [44].
Chronic pisotriquetral instability is rare and is usually seen in racket players. Pain is increased by the pisiform ballottement test: With the
wrist in slight flexion and ulnar deviation to relax the flexor carpi ulnaris tendon, the pisiform is mobilized over the triquetrum [45,46]
(Figure 32).

Figure 35: The midcarpal shift test or Lichtman's test:

To perform this test on the right wrist, the forearm is stabilized with the
examiner's left hand and held in pronated position (figure 35a). With the
patient's wrist in about 15° of ulnar deviation, the examiner's right hand
grasps the patient's right hand and, with the thumb, exerts palmarly
directed pressure at the level of the distal capitate (figure 35b). The
distal row then palmarly translates and the proximal row assumes a VISI
position (figure 35c). The wrist is then simultaneously axially loaded
and ulnarly deviated resultng in sudden reduction of the midcarpal joint
(figure 35d). This test is only positive if a painful clunk occurs with
passive ulnar deviation that reproduces the patient's symptoms.

Carpometacarpal instability:

This usually follows a fall with the wrist in hyperflexion. Pain is usually located at the base of the metacarpals. It is revealed or increased if
the examiner squeezes the metacarpals [13] (Figure 33). Linscheid's test also increases pain [7]: in this test the examiner stabilizes the
second row of the carpus with one hand, while he pushes up and down on the metacarpals with the other [4] (Figure 34). For the mobile 4th
and 5th metacarpals, compression and/or translation testing can also be performed to reveal pain [13].

To Eric Lenoble, M.D. and Pr. Alain Charles Masquelet, M.D. who reviewed the first versions and to Angèle Travadel for her drawings.


1.Linscheid R.L., Dobyns J.H., Beabout J.W., Bryan R.S. Traumatic instability of the wrist: diagnosis, classification, and
pathomechanics: J Bone Joint Surg Am 1972; 54-A:1612-1632.

2.Solgaard S., Carlsen A., Kramhoft M., Petersen V.S. Reproducibility of goniometry of the wrist: Scand J Rehab Med 1986; 18:5-7.

3.Lister G. The Hand:Diagnosis and indications. 3ème Ed. Edinburgh: Churchill Livingstone, 1993.

4.Linscheid R.L., Dobyns J.H. Physical examination of the Wrist. In: Post M. ed. Physical examination of the musculoskeletal system.
Chicago: Year Book Medical publishers; 1987: 80-94.

5. Masquelet AC: [Clinical examination of the wrist]. Ann Chir Main. 1989; 8: 159-175.

6.Masquelet A.C. Examen clinique du poignet: Cahiers d’enseignement de la société Française de chirurgie de la main 1994; 6:101-121.

7.Beckenbaugh R.D. Accurate evaluation and management of the painful wrist following injury. An approach to carpal instability: Orthop
Clin North Am 1984; 15:289-306.

8.Schernberg F. Le poignet-anatomie radiologique et chirurgie. Paris: Masson, 1992.

9.Watson H.K., Black D.M. Instabilities of the wrist: Hand Clin 1987; 3:103-111.

10.Truong N.P., Mann F.A., Gilula L.A., Kang S.W. Wrist instability series: increased yield with clinical-radiologic screening criteria:
Radiology 1994; 192:481-484.
11.Manaster B.J., Mann R.J., Rubenstein S. Wrist pain: correlation of clinical and plain film findings with arthrographic results: J Hand
Surg Am 1989; 14:466-473.

12.Cantor R.M., Stern P.J., Wyrick J.D., Michaels S.E. The relevance of ligament tears or perforations in the diagnosis of wrist pain: an
arthrographic study: J Hand Surg Am 1994; 19:945-953.

13.Gilula LA, Yin Y. Imaging of the Wrist and Hand. Philadelphia: W.B. Saunders, 1996.

14. Green DP: The sore wrist without a fracture. Instr Course Lect. 1985; 34: 300-313.

15.Taleisnik J. Pain on the ulnar side of the wrist: Hand Clin 1987; 3:51-68.

16.Jeanneret B., Buchler U. Snapping palmar drawer phenomenon of the wrist: a physiologic or pathologic study finding? Handchir
Mikrochir Plast Chir 1988; 20:111-112.

17.Jones W.A. Beware the sprained wrist. The incidence and diagnosis of scapholunate instability: J Bone Joint Surg Br 1988; 70:293-297.

18. Dauphine RT, Linscheid RL Unrecognized sprain patterns of the wrist. J Bone Joint Surg Am. 1975; 57-A: 727.

19. Ruch DS, Siegel D, Chabon SJ, Koman LA, Poehling GG Arthroscopic categorization of intercarpal ligamentous injuries of the
wrist. Orthopedics. 1993; 16: 1051-1056.

20. Fischer M, Sennwald G [Value of arthroscopy in diagnosis of carpal instability]. Handchir Mikrochir Plast Chir. 1993; 25: 39-41.

21. Taleisnik J The Wrist. Churchill Livingstone, New York, 1985

22. Linscheid RL: Scapholunate ligamentous instabilities (dissociations, subluxations, dislocations). Annales de Chirurgie de la Main.
1984; 3: 323-330.

23. Tiel van Buul MM, Bos KE, Dijkstra PF, van Beek EJ, Broekhuizen AH: Carpal instability, the missed diagnosis in patients with
clinically suspected scaphoid fracture. Injury. 1993; 24: 257-262.

24. Watson HK, Ashmead D4, Makhlouf MV: Examination of the scaphoid. J Hand Surg Am. 1988; 13: 657-660.

25. Easterling KJ, Wolfe SW: Scaphoid shift in the uninjured wrist. J Hand Surg Am. 1994; 19: 604-606.
26. Lane LB: The scaphoid shift test. J Hand Surg Am. 1993; 18: 366-368.

27. Taleisnik J: Current concepts review. Carpal instability. J Bone Joint Surg Am. 1988; 70: 1262-1268.

28. Lichtman DM, Noble WH, Alexander CE: Dynamic triquetrolunate instability: case report. J Hand Surg Am. 1984; 9: 185-188.

29. Pin PG, Young VL, Gilula LA, Weeks PM: Management of chronic lunotriquetral ligament tears. J Hand Surg Am. 1989; 14: 77-83.

30.Reagan D.S., Linscheid R.L., Dobyns J.H. Lunotriquetral sprains: J Hand Surg Am 1984; 9:502-514.

31.Ambrose L., Posner M.A. Lunate-triquetral and midcarpal joint instability: Hand Clin 1992; 8:653-668.

32. Kleinman WB. Wrist reconstruction: pitfalls. 1993;1-13.(Abstract)

33. Pin P.G., Nowak M., Logan S.E., Young V.L., Gilula L.A., Weeks P.M. Coincident rupture of the scapholunate and lunotriquetral
ligaments without perilunate dislocation: pathomechanics and management: J Hand Surg Am 1990; 15:110-119.

34.Lichtman D.M., Bruckner J.D., Culp R.W., Alexander C.E. Palmar midcarpal instability: results of surgical reconstruction: J Hand
Surg Am 1993; 18:307-315.

35.Lichtman D.M., Schneider J.R., Swafford A.R., Mack G.R. Ulnar midcarpal instability-clinical and laboratory analysis: J Hand Surg
Am 1981; 6:515-523.

36.Johnson R.P., Carrera G.F. Chronic capitolunate instability: J Bone Joint Surg Am 1986; 68-A:1164-1176.

37.Louis D.S., Hankin F.M., Greene T.L., Braunstein E.M., White S.J. Central carpal instability-capitate lunate instability pattern:
Orthopedics 1984; 7:1693-1696.

38.Masquelet A.C., Strube F., Nordin J.Y. The isolated scapho-trapezio-trapezoid ligament injury. Diagnosis and surgical treatment in
four cases: J Hand Surg Br 1993; 18:730-735.

39. Schernberg F: [Mediocarpal instability]. [French]. Annales de Chirurgie de la Main. 1984; 3: 344-348.

40.Kuhlmann J.N., Mimoun M., Fahed I., Boabighi A., Baux S. [Severe and inveterate internal scapho-trapezial ligament sprain.
Apropos of an unusual case]: Ann Chir Main Memb Super 1992; 11:62-65.
41.Dumontier C., Lenoble E., Saffar P. Radiocarpal dislocations and fracture-dislocations. In: Saffar P., Cooney W.P.3. eds. Fractures of
the distal radius. London: Martin DUNITZ; 1995: 267-278.

42.Bellinghausen H.W., Gilula L.A., Young L.V., Weeks P.M. Post-traumatic palmar carpal subluxation. Report of two cases: J Bone
Joint Surg Am 1983; 65:998-1006.

43.Rayhack J.M., Linscheid R.L., Dobyns J.H., Smith J.H. Posttraumatic ulnar translation of the carpus: J Hand Surg Am 1987; 12:180-

44.McEwen Smith A. Sprain of the pisiform-triquetral joint: J Bone Joint Surg Br 1954; 36B:618-621.

45.Helal B. Racquet player’s pisiform: The Hand 1978; 10:87-90.

46.Nüesch B., Sennwald G., Segmüller G. Pisiformexstirpation: indikation und resultate: Handchir Mikrochir Plast Chir 1993; 25:42-45.