A literature search on the sternoclavicular joint rapidly makes the reader aware that there are a very limited number of publications on this joint and these predominantly cover medical and surgical concerns. Anatomical and biomechanical references are designed, most often, to support medical or surgical interventions.
From a Physical Therapy perspective this joint would appear to be the ‘poor cousin’ of the shoulder girdle in both interest and investigation.
However, whilst bemoaning this fact the reason almost certainly lies in the joint’s inherent strength and stability. These factors will be covered in the sections on anatomy and biomechanics.
What is most interesting to the writer, however, is this joint’s proposed ability to work in concert with the thoracic spine to facilitate the function of elevation through flexion/abduction without compromise to the neurovascular structures that supply the upper limb.
In view of our upper limbs we must accept the fact that they enable us to be primate ‘brachiates’ i.e., we are able to locomote with our upper limbs. Whilst, as we get older and heavier this seems an unlikely premise we need merely to view children at a playground or study gymnasts to realise this is at least one of the functions of the human upper limb and something we share in common with all other primates.
In all anatomy texts that I have read, the shoulder girdle ‘ends’ at the manubrium. Emphasis is therefore placed on how the clavicle moves at the sterno-manubrial articulation. The writer’s paper entitled ‘The functional shoulder girdle’ (21) inferred that during functional movements of the shoulder girdle there is indeed another biomechanical component to this joint that needs to be considered and that is manubrio-sternal motion i.e., that dictated by the thoracic spine.
Given the dearth of biomechanical research regarding this joint the writer must at least present a proposed biomechanical model, based on observation and palpation, that might lead to further investigation and research. Within this proposal will be a clinically reasoned explanation as to how thoracic and/or sternoclavicular dysfunction may directly affect glenohumeral joint function.
ANATOMY and HISTOLOGY
Descriptive anatomy of this joint is well covered in texts such as Gray’s Anatomy (11). The writer feels that the emphasis here should be on functional and comparative anatomy.
In embryologicaldevelopment the clavicle is present in almost all mammals. However, in quadrupeds the clavicle becomes a vestigial or rudimentary structure helping to provide muscle attachments that produce a muscular ‘sling’ to support the weight of the thorax, neck and head.
A fully developed, osseous clavicle that connects the scapula to the manubrium only exists in primates. This bony strut enables primates to enjoy a very large range of upper limb motion, especially away from the midline of the body.
Such motion gives primates the functional advantages of grasping, thrusting (throwing or punching) and brachiation (swinging).
Specialization of function within different primate groups appears to depend upon the position of the scapula (lateral or posterior to the thorax) (2) and the curvature(s) of the clavicle (24). The distinctive ‘S’ shape of the human clavicle has been likened to a ‘crank’. This enables our muscles to support a relatively heavier body weight during brachiation but also to increase the power and velocity of upper limb movements such as throwing. However, that same ‘S’ shaped clavicle does poorly with compressive loading, its weak spot being the junction between the medial convexity and its lateral concavity. This fact is underscored by this (midshaft) region being the commonest site for clavicular fractures during compressive loading such as a fall directly on the shoulder or on the outstretched hand (4). Most fractures of this region are uncomplicated but rarely may lead to brachial plexus involvement, pulmonary dysfunction or even death (17).
The sternoclavicular joint surfaces
The medial end of the clavicle presents as a large, bulbous head. The surface is concave horizontally and convex vertically making it appear saddle-shaped. Histological analysis of the clavicular head (7), at least developmentally, shows plates of cartilage within the bone. This is a direct comparison with the head of the mandible (25), both designed presumably to absorb extreme stresses and strains.
The corresponding surface of the sternum reciprocally has an obvious concave surface vertically and a slight convexity horizontally.
Since the articulating surface of the clavicular head is over twice that of the manubrial surface this apparent incongruence, whilst enabling a large amplitude of motion makes the joint potentially very unstable. It is the role of the joint’s ligamentous structures to maintain stability (15).
Ligaments of the sternoclavicular joint
Intra-articular (ligament) disc:
As described in the sub-title there is some disagreement as to whether this structure serves primarily as a ligament or intra-articular disc and this will be discussed later.
This dense fibrous structure has a strong peripheral capsular attachment that completely divides the joint into separate cavities (5), which in itself hints at a discreet function for each of the joint’s cavities. Occasionally, there may some central connection between the two joint cavities but this is believed to be secondary to wear and tear.
Inferiorly, the disc arises from the synchondrosis of the 1st rib cartilage and the manubrium. Superiorly, it attaches to the superior and medial aspects of the medial clavicle at the lateral joint margin but blends with the fibres of the capsular (superior) ligament.
This ligament, as the name suggests, blends with the same ligament of the opposite side. Also, it is attached to the superior part of the manubrium and blends with the ipsilateral capsular (superior) ligament.
Capsular (superior) ligament:
This ligament, perhaps the strongest of the sternoclavicular joint, really represents antero-superior and posterior reinforcements (? thickening) of the articular capsule, the antero-superior being the thickest.
Working in concert these three (above) ligaments afford both strength and static stability to the sternoclavicular joint with the shoulder girdles in a resting, weight dependent position. This has been referred to as ‘shoulder poise’ where the distal end of the clavicle is passively supported slightly higher than its medial end.
As a passive support mechanism they represent a significant saving in muscular energy expenditure to help carry objects on the shoulder girdles (e.g, a yoke, satchel and even a child) or carry objects by hand (hunted game, water containers and suitcases).
Also, passive shoulder poise is essential for efficiency of manual techniques that require minimal shoulder girdle excursion (e.g, moulding clay, cooking and ‘mousing’).
With regard to stability, the most important of the three ligaments appears to be the capsular (superior) ligament. Cadaveral experiments (1) have clearly demonstrated that static ‘clavicular poise’ is independent of myofascial support, or even support from the inter-clavicular and intra-articular disc ligaments.
Once the capsular ligament is torn minimal force is needed to tear the intra-articular ligament, leading to superior dislocation and disruption of the sternoclavicular joint. If the posterior part of the capsular ligament also fails then posterior dislocation is possible which may lead to more serious health or even life-threatening complications.
Also called the ‘rhomboid ligament’ because of the orientation of its fibres, there appears to be some disagreement as to its actual morphology (23).
Traditionally this ligament has been described as a ‘flattened’ cone. The best way to envisage this is to take a polystyrene paper cup and draw oblique parallel lines around its perimeter. Now squash it flat. The drawn lines would resemble a rhomboid if viewed from anterior and posterior. However the lateral and medial margins of the cup would appear to continue the spiralling lines originally drawn. As such, the fibres of the ligament would indeed be capable of resisting clavicular motion in all directions and planes, except for one and that is depression of the clavicle in neutral. The only argument within the literature is whether there is an interposing bursa (or space) between the anterior and posterior sets of fibres or do they form one solid mass.
Regardless, the orientation this ligament’s fibres are clearly designed to resist any motion of the clavicle away from its neutral ‘poise’.
The anterior fibres appear particularly vulnerable to excessive elevation and protraction of the shoulder girdle, which the writer feels, from cadaveral observation, is the close-packed position of the sternoclavicular joint.
During full elevation of the humerus (through flexion/abduction) the shoulder girdle (scapula and clavicle) moves into depression and retraction. The disagreement as to whether the clavicle elevates or depresses (19) is probably the result of different instructions to the observed models. If the model is asked to elevate their hand as high as they can, then elevation of the clavicle will result. However, the writer believes that functional elevation requires a stable, depressed clavicle.
During full elevation recruitment of lower trapezius coincides with activation of subclavius (18). This would make sense since the motion probably coincides with the greatest stress placed on the sternoclavicular joint for either throwing or brachiation. The ‘shunt’ action of subclavius would be most appropriate now for stability of the sternoclavicular joint.
In the writer’s experience most costo-clavicular ligament injuries (treatable by physical therapy) occur when there is forceful elevation thrust of the arm with a corresponding elevation and protraction of the shoulder girdle. This forced lateral displacement of the clavicle would not be resisted by the appropriate reflex shunt action of subclavius rendering the anterior fibres to damage.
Individuals who might perform such an action are limited to certain athletes (e.g, shot putters, javelin throwers, racquet players) but also those performing household tasks (e.g, cleaning a bathtub, painters).
At the sternoclavicular joint the clavicle is clearly capable of moving through at least two cardinal planes i.e, horizontal ( 35 degrees of combined protraction and retraction) and vertical (30-35 degrees of elevation) (15). The joint is therefore considered to possess two degrees of freedom which are both pure swings.
The largest displacement however is 45-50 degrees of rotation (15) (19) around the long axis of the clavicle (i.e, motion through a sagittal plane) but can this be considered a degree of freedom?
The joint is clearly divided into two separate anatomical compartments (which suggests two separate functions (c.f. the temporomandibular joint).
If the posterior edge of the lateral end of the clavicle is palpated during inspiration and expiration rotation of the bone is clearly felt. This is because the clavicle is ‘crank’ shaped, and as the manubrium rises with inspiration elevation of the medial end of the clavicle produces a posterior rotation around its longitudinal axis.
Since there is no other displacement of the clavicle (shoulder girdle) obvious it is assumed that the rotational motion at the sternoclavicular joint occurs within the medial component (disc/manubrium) of the joint. So it could well be argued that the sternoclavicular joint complex has indeed got three degrees of freedom of motion.
The large, superficial head of the clavicle is easily palpated during motion of the shoulder girdle on a relatively stationary manubrium. From full retraction towards protraction the most obvious motion initially appears as a posterior male glide but this glide only occurs through the first two thirds of the range (from full retraction to neutral poise). After that, as protraction continues an anterior rotation is apparent, representing a female glide.
A similar change in glides is apparent when the joint is palpated from full depression to full elevation of the shoulder girdle where there is initial inferior male glide followed by a superior anterior female glide.
Understanding that the male motion occurs at the disc/clavicle component and the female from the disc/manubrium component allows for a very simple palpatory assessment technique to discern which component is in dysfunction, or indeed whether the whole complex might be deranged.
Proposed interaction between clavicular motion and manubrial motion during elevation of the arm through flexion and abduction
As the arm is elevated through flexion/abduction the initial motion appears to occur at the glenohumeral and acromioclavicular joints on a relatively fixed clavicle. The inferior angle of the scapula displaces laterally and anteriorly to produce an upward rotation of the scapula (glenoid surface), the motion occurring at the acromioclavicular joint.
At about 150 degrees of elevation the inferior angle stops moving. Presumably fixated by an isometric contraction of lower serratus anterior the axis of shoulder girdle motion now shifts from the acromioclavicular joint to the sternoclavicular joint and the girdle is seen to depress and retract in the last 30-50 degrees of arm elevation.
It might be reasonably assumed that the clavicle should significantly rotate posteriorly. However, if the clavicle is palpated during this terminal range minimal, if any, rotation is sensed.
To solve this apparent conundrum one must now study what is occurring at the manubrium. As the arm is elevated beyond 150 degrees upper thoracic motion can also be both seen and felt. The upper thoracic spine extends and ipsilaterally rotates and side bends towards the moving arm.
The 1st thoracic vertebra, first rib and manubrium now all move in concert, dictated by thoracic spine motion. This is easily felt at the manubrium by palpating bilaterally just below the first rib cartilage. The manubrium also side bends and rotates towards the elevating arm. So the manubrium is now moving under the clavicle producing a relative anterior rotation of the sternoclavicular joint. This simultaneous motion of both the clavicle and the manubrium ensures that there is no resultant posterior rotation of the clavicle. The main question now is why would this be necessary?
The deep cervical fascia blends with the posterior and superior periosteum of the clavicle. If the clavicle were to rotate posteriorly up to 45 degrees, as has been suggested, the deep cervical fascia would undergo an extreme increase in tension potentially compromising the neuro-vascular tissue passing through it.
Although not strictly a sternoclavicular joint disruption or injury, an inability for the thoracic spine to move appropriately during the final stages of arm elevation would prevent the disc/manubrium component of the sternoclavicular joint from de-rotating the clavicle. Clinically, those who perform habitual or sustained arm elevation in their recreational or work environment would potentially complain of signs or symptoms of abnormal neural tension within the arm and/or damage to distal shoulder girdle structures from mechanical compensation.
For this reason assessment of thoracic and manubrial motion should be a routine part of shoulder girdle assessment.
PATHOLOGIES AND ASSESSMENT
For the physical therapist pathologies of the sternoclavicular joint might best be divided into two main sections i.e, those requiring a medical/surgical consult and those for whom physical therapy intervention is indicated.
Patient’s requiring a medical/surgical consult
The sternoclavicular joint is susceptible to any of the pathophysiologies that affect synovial joints (14) (15). Whilst not attempting an exact diagnosis the therapist needs to be able to identify patients suffering from serious traumatic injuries and non-traumatic or degenerative arthritic conditions.
Serious traumatic injuries
Dislocations, although uncommon, represent the greatest threat to articular function. They include anterior, superior and posterior.
Dislocations can occur from direct trauma to the clavicle or manubrium as may occur in motor vehicle accidents or sports. They also result from indirect trauma, especially to the postero-lateral shoulder (superior and posterior dislocations) and antero-lateral shoulder (anterior dislocations) (15).
The posterior dislocation is of greatest concern because of the threat to retro-sternal structures such as the trachea and major blood vessels (3) (22) (26). If these structures are involved the patient may well be observed as having breathing problems and changes in skin colour due to vascular or airway compromise.
Dislocations tend not to be subtle. The therapist may suspect them from a history of extreme trauma, a gross loss of motion of the upper limb and an obvious, observable and palpable change in the natural contours of the sternoclavicular joint.
During palpation of motion (described later) the therapist may detect gross disruption of the anticipated (male to female) motion sequence.
Although the clavicle is the first long bone to begin ossification it is the last to complete it. The epiphysis of the medial end of the clavicle ossifies in the 18th-20th year and fuses with the shaft between the 23rd-25th years. Direct and indirect trauma to the medial end of the clavicle may result in epiphyseal disruption, even fracture. Closely resembling the presentation of a dislocation, only medical examination can provide an accurate diagnosis (15).
Hyperostosis at the sternoclavicular joint (6) (20), felt initially by the therapist as an apparent bony hypertrophy of either the clavicular head or the manubrium could signify serious pathology, and a medical consult is certainly warranted (9). However, the writer has seen two cases of physeal trauma where fracture or disruption were ruled out but the trauma resulted in benign hyperostosis of the head of the clavicle. Apart from the distressing cosmetic appearance, joint function and stability appeared normal.
Non-traumatic arthritic conditions
This joint has been shown to suffer from almost all potential causes of non-traumatic arthritis, the more common including septic arthritis, rheumatoid arthritis, tuberculosis and ankylosing spondylitis. The writer has rarely seen gouty arthritis mentioned in the literature (16) but clearly it cannot be ruled out.
The presentation of a painful, hot swollen joint with no history of injury should immediately raise enough concern for the therapist to request a medical consult, regardless of its ultimate diagnosis.
Physical therapy intervention is indicated
These patients will, of course, include those with sprains and strains of the sternoclavicular joint. Acute traumatic arthritis may present with enough pain, swelling and dysfunction that a medical consult should be sought. They are most effectively treated with a resting sling and subsequent referral to physical therapy.
The reference of pain from sternoclavicular joint injury is commonly localised to the joint itself, but distal reference (13) e.g, neck, shoulder and arm is also possible.
With subacute and chronic arthritis the therapist’s main concern is whether joint motion has been lost, and if so assess which component of the joint is responsible.
Another concern would be the possibility of any ligamentous sprain. The writer is unaware of any confirmed discreet tests for the intra-articular disc ligament, capsular ligament or interclavicular ligament. However, if articular motion is normal but localised pain is reproduced by overpressure of shoulder girdle movements, a ligamentous injury is suspected. Accurate palpation followed by deep transverse friction massage (DTFM) and ultrasound would appear to be the treatment of choice.
The costoclavicular ligament can be stressed discreetly.
With the patient in contralateral side lying the therapist moves the (affected side) glenohumeral joint into extension and adduction. Applying pressure through the elbow the therapist then pushes the shoulder girdle into full elevation and protraction. Continued pressure through the patient’s elbow now provides a lateral distractive force to the sternoclavicular joint, maximally stressing the costoclavicular ligament.
The anterior fibres are proposed the most likely to be injured. They are accessible to DTFM if the shoulder girdle is positioned into depression and retraction (posterior rotation of the clavicle).
Osteoarthrosis is suspected when crepitations, or even ‘clunking’, are detected during motion palpation. One paper on cadaveral dissection of the sternoclavicular joints suggested that 80% of people over 50 may have osteoarthrosis of this joint (12). Since cadaveral dissection can rarely be correlated with symptoms it is unclear how much symptomatology this condition is responsible for.
In the writer’s experience minor, asymptomatic joint crepitations are common in the presence of normal joint function and should probably be ignored. However, if the crepitation or clunking is significant, or corresponds to the reproduction of the patient’s symptoms then a medical consult should be sought. The degenerative state of the joint may help in an eventual prognosis, but also in determining the appropriate magnitude of force used in rehabilitation procedures. It is worth remembering that all resisted forces on the upper limb must ultimately be transferred to the sternoclavicular joint.
Joint mobility tests
As was inferred earlier the size of the clavicular head, coupled with the fact it is so superficial, enables the therapist to easily palpate sternoclavicular joint motion.
Following the taking of a history and observation the therapist palpates the anterior surface of the head of the clavicle.
From a position of full retraction the patient is instructed to pull their shoulder girdles into protraction. In normal motion the therapist should be able to feel the head of the clavicle move initially posteriorly (male clavicular/disc motion). At the position of neutral ‘poise’ the motion should be felt to change to a female (disc/manubrial motion) anterior glide (roll).
From a position of full depression the patient is instructed to lift their shoulder girdles into elevation. In normal motion the therapist should be able to feel the head of the clavicle move initially inferiorly (male clavicular/disc motion). At mid-range this motion is felt to change to a female superior/anterior roll.
This simple test enables the therapist to decide which articular component is lacking.
Passive joint mobilizations
Female/menisco-sternal motion(e.g, left shoulder)
The patient is in right side lying facing the therapist. Therapist’s left middle and ring finger tips are tucked posterior and inferior to the lateral edge of the patient’s left clavicle. Therapist’s right hand grasps the inferior angle of the scapula
Passively draws the patient’s left shoulder girdle into elevation and protraction until anterior rotation is sensed to cease.
Therapist instructs the patient to take a short breath in followed by a long breath out. As the patient breaths out increased anterior rotation of the clavicle is taken up by passively increasing elevation and protraction and also by the therapist’s left hand pulling the posterior edge of the clavicle upwards and forwards.
This procedure is repeated until no further motion is detected.
The patient starting position is the same as described above. However, the therapist’s middle and ring fingers are over the superior aspect of the posterior edge of the lateral clavicle.
Therapist moves the shoulder girdle into depression and retraction until posterior rotation of the clavicle ceases. The patient is instructed to take a short breath out followed by a long breath in.
As increased posterior rotation of the clavicle is detected the therapist pushes the patient’s shoulder girdle into further depression and retraction with an accompanying push on the posterior edge of the clavicle inferiorly by the therapist’s left fingers.
This procedure is repeated until no further motion is detected.
Male /clavicular-meniscal motion (e.g, right shoulder)
The patient is in supine with therapist standing adjacent to the patient’s opposite shoulder girdle. Therapist’s left thumb pad or thenar eminence is placed over the superior aspect of the head of the patient’s right clavicle.
Therapist’s right hand draws the patient’s right shoulder girdle into elevation until the inferior glide of the clavicular head ceases. Therapist instructs the patient to resist an attempt to push the right shoulder girdle into depression. An inferior glide of the right clavicular head will be detected and this motion slack is taken up by pressure from the therapist’s left thumb. Any slack in right girdle elevation is now taken up by the therapist’s right hand.
This procedure is repeated until no further motion is perceived.
The patient is in supine with the therapist standing on the opposite side to the joint being treated. Therapist grasps the patient’s right shoulder with their left hand and instructs the patient to place their right hand on therapist’s left arm.
Therapist’s right thumb or thenar eminence is placed over the anterior surface of the head of the patient’s right clavicle.
The patient is instructed to resist therapist’s attempt to push patient’s right shoulder girdle into retraction. A posterior motion of the clavicular head will be noted and therapists right thumb takes up the slack. Any increased protraction is taken up by therapist’s left hand.
This procedure is repeated until no further motion is perceived.
Active exercises to maintain range of motion of the sternoclavicular joint gained by passive mobilizations should simply be instructed in functional sets i.e, emphasis on either elevation and protraction or depression and retraction.
With regard to normal sternoclavicular joint function however, the writer cannot over emphasise the need for normal thoracic joint motion. This may necessitate additional thoracic spine exercises to facilitate shoulder girdle depression and retraction (extension and ipsilateral side bending/rotation of the thoracic spine) or girdle elevation and protraction (flexion and contralateral side bending/rotation of the thoracic spine).
(1) Bearn JG. Direct observations on the function of the capsule of the sternovlavicular joint in the clavicular support. Anat. 1967;101:159-170.
(2) Chan LK. Scapular position in primates. Folia Primatol. 2007;7:19-35.
(3) Cooper GJ, Stubbs D, Walker DA, et al. Posterior sternoclavicular joint dislocation: a novel method of external fixation. Injury 1992;23:565-7.
(4) Denard PJ, Koval KJ, Cantu RV et al. Management of midshaft clavicle fractures in adults. Am J Orthop. 2005 Nov;34(11):527-36.
(5) DePalma AF. The role of the disks of the sternoclavicular and the acromioclavicular joints. Clin Orthop. 1959;13:222-233.
(6) Dihlmann W, Schnabel A, Gross WL. The acquired hyperostosis syndrome: a little known skeletal disorder with distinctive radiological and clinical features. Clin Investig. 1993 Dec;72(1):4-11.
(7) Ellis E 3rd, Carlson DS. Histological comparison of the costochondral, sternoclavicular and temporomandibular joints during growth in Macaca mulatta. J Oral Maxillofac Surg. 1986 April;44(4):312-21.
(8) Ferrara PC, Wheeling HM. Sternoclavicular joint injuries. Am J Emerg Med. 2000 Jan;18(1):58-61.
(9) Fritz P, Baldauf G, Whilke HJ et al. Hyperostosis: its progression and radiological features. Ann Rheum Dis. 1992 May;51(5):658-64.
(10) Frosi G, Sulli A, Testa M et al. The sternoclavicular joint: anatomy, biomechanics, clinical features and aspects of manual therapy. Rheumatismo. 2004 Apr-Jun;56(2):82-8.
(11) Gray,s Anatomy: The Anatomical Basis of Clinical Practice. 39th edn. Editor Standring S. Pub 2004 Churchill Livingstone.
(12) Hagemann R, Ruttner JR. Arthrosis of the sternoclavicular joint. Z Rheumatol. 1979 Jan-Feb;38(1-2):27-28.
(13) Hassett G, Barnsley L. Pain referral from the sternoclavicular joint: a study in normal volunteers. Rheumatology (Oxford). 2001 Aug;40(8):859-62.
(14) Higginbotham TO, Khun JE. Atraumatic disorders of the sternoclavicular joint. J Am Acad Orthop Surg. 2005 Mar-Apr;13(2):138-45.
(15) Iannotti and Williams. Disorders of the Shoulder. Pub Lippincott Williams and Wilkins. 1999. Ch 28. 765-808.
(16) Kearn A, Schunk A, Thelan M. Gout in the area of the cervical area and sternoclavicular joint. Rofo. 1999 May;170(5):515-7.
(17) Kendall KM, Burton JH, Cushing B. Fatal subclavian artery transaction from isolated clavicle fracture. T Trauma. 2000;42(2):316-18.
(18) Konstant W, Stern J, Fleagle J, et al. Function of the subclavius muscle in a non-human primate, the spider monkey. Folia Primatol. 1982;38:170-182.
(19) Ludewig P, Bahrens S, Spoden S et al. J Orthop Sports Phys Ther. 2004;34(3):140-149.
(20) Noble JS. Degenerative sternoclavcular arthritis and hyperostosis. Clin Sports Med. 2003 Apr;22(2):407-22.
(21) Pettman E. The functional shoulder girdle. 1984. International Federation of Orthopaedic Manipulative Therapists (IFOMT). Vancouver, BC, Canada.
(22) Rodrigues H. Case of dislocation, inwards, of the internal extremity of the clavicle. Lancet. 1843;1:309-310.
(23) Tubbs SR, Shah NA, Sullivan BP, et al. The costoclavicular ligament revisited: a functional and anatomical study. ROM J Morphol Embryol. 2009;50(3):475-9.
(24) Voisin JL. Clavicle, a neglected bone: morphology and relation to arm movements and shoulder architecture in primates. The Anatomical Record Part A 288A:944-53 (2006).
(25) Wolford LM, Cottrell DA, Henry C. Sternoclavicular grafts for temporomandibular reconstruction. J Oral Maxillofac Surg. 1994 Feb;52(2):119-28.
(26) Worman LW Laegus C. Intrathoracic injury following retrosternal dislocation of the clavicle. J Trauma. 1967;7:416-423.