A Context for Change
The pressure for health care reform has increased awareness that rehabilitation costs are too high. As a result, both payers and consumers are demanding high-quality, cost-effective rehabilitation protocols that are scientifically validated, which is not an easy task because most rehabilitation protocols are developed from empiric data and are individualized for each patient. As a consequence, protocols used for shoulder rehabilitation are as variable and as numerous as the clinicians who develop them. Scientifically validating rehabilitation protocols requires the establishment of reliable and valid functional outcome measures. As the first step in that direction, this chapter describes a functionally based neuromechanical approach to rehabilitation propitious to more precise documentation and provides modular guidelines for the rehabilitation of the most common disorders, including (1) adhesive capsulitis, (2) impingement syndrome, (3) rotator cuff tears, and (4) glenohumeral instability.
Limiting Factors for Rehabilitation Research
Many factors hamper the rehabilitation research needed to remedy the lack of adequate validation of treatment procedures: (1) There is no statewide or national plan for rehabilitation research that coordinates academic and clinical resources; (2) clinicians lack specialized training in rehabilitation research; (3) mechanisms to ensure quality and scientific relevance have not been developed; (4) there is no central source for the collection and dissemination of the results of rehabilitation; (5) inadequate funding and many bureaucratic barriers impede the growth and development of rehabilitation research; and (6) current rehabilitation research focuses primarily on the pathophysiology, rather than the functional outcomes, of rehabilitation research. Therefore validation of any rehabilitation protocol will require an expenditure of substantial effort and money.
A Direction for Research
Feinstein, a famous Yale scholar and epidemiologist, believes that all clinicians are researchers in their own right because they gain clinical experience by a trial-and-error research method. A diagnosis is made, an intervention is performed, and the results are assessed. In this overall medical model, the practitioner develops a keen sense of what a given patient needs by clustering signs and symptoms and then deriving an appropriate treatment strategy from his or her education, training, and experience. The core of the medical model is the pathologic diagnosis and a cause-and-effect relationship. In rehabilitation, the core that determines an appropriate rehabilitation strategy is not the pathologic diagnosis alone but rather the ability to cluster the signs and symptoms in ways that identify the behavioral impairments, physical impairments, and environmental constraints to normal functioning. The Feinstein model for clustering signs and symptoms can lead to the development of a patient classification system centering on identification or characterization of the movement dysfunction that the patient presents regardless of the injury or the related disorder. For example, practitioners have described characteristic patterns of posture, balance, movement, and behavior with back pain, antalgic and neurologic gait, reflex sympathetic dystrophy (RSD), overuse syndromes, injury, and disease.
Rose, Dellito and Synden-Mackler, Sahrmann, and Rothstein and Echterach modified the Feinstein model for rehabilitation and introduced a patient classification system for movement-related disorders. Such an approach enables the practitioner to determine accurately the behavioral and physical impairments and environmental constraints that prohibit any patient from functioning in his or her environment, regardless of the pathologic condition, and to design rehabilitation protocols that meet the patient’s functional needs rather than treat all patients with a given pathologic diagnosis under a common protocol.
The examination and rehabilitation of shoulder disabilities started from the premise that the pathologic diagnosis leads automatically to the development of an appropriate rehabilitation protocol. In fact, most practitioners would question the validity of this assumption. However important and accurate pathologic diagnosis may be, Dellito and Synden-Mackler and Sahrmann argue that efficient and effective rehabilitation protocols that truly meet the patient’s needs cannot be derived solely from the pathologic diagnosis. In practice, modern therapists design their rehabilitation protocols by identifying the movement restrictions on normal function that the patient presents rather than by deducing them from the identified pathologic condition. Therefore they emphasize identification of the movement behavior characterized by the presenting condition. By establishing a given movement-related diagnosis or disorder of the shoulder, each practitioner can efficiently identify all the impairments and constraints to normal function and can design an appropriate strategy to maximize or normalize movement. A movement-related diagnosis considers the patient’s age, injury (traumatic or arthritic), emotional and physiologic status, movement behaviors, and environmental factors inhibiting normal function. This diagnosis considers that the whole person is functionally driven.
For the fields of physical and occupational therapy, measuring patient progress in functional outcomes is the only prudent assessment of patient progression until reliable and valid measures of functional outcome have been established by scientific rigors. Therefore the challenge of the future will be the establishment of these functional measures.
A New Physical Therapy Practice Model
Rehabilitation protocols for given clinical or pathologic conditions usually are based on the use of exercises or specific functional tasks. These protocols usually are administered in a specified rehabilitation environment and, for the most part, require the use of equipment and varied positions for the exercises or functional tasks performed. Most recently, neurologic therapists have introduced a systems model for rehabilitation of neurologic dysfunction based on an understanding of movement science. The emergence of an integrated scientific core for describing normal functional movement and characterizing abnormal or pathologic movement patterns has introduced a new concept for rehabilitation of neurologic patients, especially those suffering from stroke.
In 1992, a group of multispecialized physical therapists expanded the neurologic views and developed a comprehensive and integrated physical therapy practice model. The purpose of this model was to design more effective ways to establish functional outcome measures for patient progression, clinical research, and reimbursement. This model emphasized the importance of integrating specialty practice, including cardiopulmonary, orthopedic, and neurologic practice, into the neuromechanical model for rehabilitation so that function could be assessed and the energy costs measured to compare with normative movement or functional data. This type of comparison has the potential to be used in research of patient-treatment effectiveness or outcomes to validate rehabilitation protocols.
Postural Control and Function of the Shoulder
The shoulder joint has a unique anatomic structure to allow great freedom of movement during an almost infinite variety of human tasks. Although the shoulder joint seems unstable anatomically, it is able to (1) resist gravitational pull with the arm hanging at the side for long periods; (2) lift, pull, and push heavy loads; (3) position the hands for both power (grasping) and delicate hand activities; (4) allow for throwing at speeds up to nearly 100 miles per hour; (5) assist with body- and trunk-supporting reactions; and (6) hold together during the application of an almost infinite variety of forces of differing magnitude, direction, duration, and abruptness. The shoulder participates in many actions that require complex neuromechanical postural and balance mechanisms, such as gesturing, feeding, grooming, supporting reactions, throwing, lifting, pushing, pulling, reaching, grasping, and swinging the arm during gait.
Kendall described good posture as that state of muscular and skeletal balance that protects the supporting structures of the body against injury or progressive deformity irrespective of the attitude (erect, lying, kneeling, squatting, stooping, bending, walking, running, or any functional activity) in which the structures are working or resting. Under these conditions, the muscles function most efficiently and the structures are aligned for the most efficient movements with minimal energy costs. Fig. 83-1 shows the theoretic position for the center of mass over the base of support. This implies that each joint has a balanced distribution of weight and a stable position or alignment. In the erect posture, this represents the ideal or theoretic position for standing, but the body must maintain a similar relationship during dynamic movement and must be able to maintain balance during varied functional activities.
Posture is controlled by an interaction between the neurologic and musculoskeletal systems. The neurologic system uses sensory processing (visual, vestibular, and proprioceptive from the limbs, neck, and trunk), motor planning, and programming to provide motor output for strength and endurance. The musculoskeletal system provides postural alignment of the skeleton and the flexibility of the spine, pelvis, and extremities needed to achieve appropriate balance during any functional activity. Balance control is preset for each activity and occurs when the intention for the function is first realized.
The intention to perform an action and the environment in which the action takes place stimulate the appropriate motor programming to accomplish the task. For example, ask someone to demonstrate drinking from a glass without actually using a glass. Note the movement pattern that he or she uses. Now take three glasses of water, one half full, one filled to the very top of the glass, and one with only a few drops of water. Ask the individual to drink from each of the glasses and note the differences in the movement patterns. You will see three different strategies for accomplishing the task of drinking, and each posture depends on the amount of water in the glass. In particular, note how adjustments must be made when drinking from the full glass and how the trunk flexes to bring the head and mouth to the glass rather than the glass being brought to mouth, as with the other two attempts. Also note the action of the cervical spine with both the half-full glass and the one with only a few drops of water. In both of these situations, the individual brings the water to the lips by tilting the glass and head at the same time. Drinking the last few drops requires a greater amount of tilting with cervical spine extension than the other tasks. The posturing during a simple task of drinking under varying conditions requires a complex series of postural adjustments to accomplish the task. These adjustments are environmentally dependent. Any alterations, limitations, or impairments in the neurologic and musculoskeletal systems caused by injury, aging, or environmental constraints, such as drinking from different-diameter glasses requiring altered grips, will change the overall movement pattern or strategy to accomplish the task. Overall balance control is necessary to drink under varying conditions without spilling the water. Varying the posture reduces the degree of freedom between body segments and allows preprogrammed activities to be executed in the most efficient manner.
Interrelationships between Body Segments
The shoulder never functions in isolation from other body segments. Many rehabilitation protocols describe techniques to improve shoulder motion and function. Functional movements of the shoulder require that all body segments from the feet to the head work in a coordinated manner to produce efficient movement. Fig. 83-2 represents the interrelationships between body segments required with any functional activity. These relationships are easiest to describe in the standing position. The movement of the head is controlled by sensory input from the auditory, vestibular, and visual systems. These systems account for the basic orienting systems for postural control. The intersegmental relationships of the spine from the cervical segments to the lumbosacral junction produce interplay between the left and right sides of the body, rotation around the midline, and the integration of righting and supporting reactions. The shoulder girdles have great mobility and allow the hands to be placed for grasping and manipulation of objects, and stability when the arm is providing support or when strength is needed with functional movements of the hands. The hands are free for manipulation, exploration, and various functional tasks. The stabilization of the pelvis, with associated extensor activity, is required to free the movement of the upper extremity during gait and reaching activities. The feet require the acceptance of the floor as a base of support. Proprioceptive feedback produces the perception of verticality, and pressure changes on the feet and changes in the angle of the ankles adjust the postural sway to maintain balance during functional activities.
Kinematic and Kinetic Linkages
The interrelationships between the head, spine, shoulder girdles, hands, pelvis, and feet must be controlled and adjusted with each functional activity. Winter advocated the development of linked segment models ( Fig. 83-3 ) for analyzing human motion during functional activities. With appropriate kinematic, anthropometric, and kinetic output data, investigators of human movement may be able to perform analyses that could improve our understanding of normal and dysfunctional movement patterns that may follow disease. Winter further describes the interrelationship between skeletal and muscle mechanics, neural control, and muscle metabolism ( Fig. 83-4 ). The force, moments, power, and work required during functional, elite, pathologic, or fatigue movements can be analyzed to determine the pattern of muscle synergy, biomechanics of joint and supporting structures, and energy costs. The goal of rehabilitation is and always has been to develop the most efficient patterns of movement with the least amount of energy expended. Rehabilitation research must be directed toward this end if we are to develop high-quality, cost-effective rehabilitation protocols.
To understand the concept of linked body segments in a practical sense, we begin with functional examples illustrating how changing the position of linked segments can affect the functions of the shoulder.
Sit in an armless chair with your feet flat on the floor or on a supporting surface. Sit near the front of the chair in an erect posture (not a slumped position). Raise both arms through the full arc of flexion and then lower the arms. Note the ease or effort required by the arm flexors to lift the hands overhead. Now sit back in the chair in a slumped position. With one hand, pull the sternum down to slightly increase the thoracic kyphosis, bring the head forward slightly, and extend the upper cervical spine to place the eyes on the horizontal. Now lift both arms and sense any difference in range of motion (ROM) and effort. The observer should note that the ROM is limited and the effort required to lift the arms has increased.
Repeat the two positions, but change the tasks. This time elevate your right arm to 90 degrees of flexion and hold it in that position. Have another person test the isometric strength of the shoulder flexors in both the erect and slumped postures. Note the weakness of the shoulder musculature in the slumped position.
Repeat the two positions while taking a deep breath. Note that rib cage expansion is significantly limited in the slumped position. The latter position would significantly increase the energy costs during any functional activity if it were maintained.
Repeat the two positions while reaching with the right hand toward the left (a turning or twisting movement). Note that reaching is severely limited in the slumped position. To add to the limitation, keep your weight on your right hind quarter. Note that reaching to the left is almost impossible.
In each of the examples, the body linkages were altered from the erect posture to the slumped posture to decrease the ROM, strength, and thoracic expansions. These same examples could be repeated in the standing position or during walking, with the same result. The slumped or forward-head posture shifts the center of gravity forward over the base of support. The change in postural position produces a significant flexion synergy and allows gravity to control the individual’s posture instead of the individual’s erect posture controlling gravity.
One of the most important functions to restore during rehabilitation is that of reaching. The ability to use the hand under visual guidance to interact with the environment represents a high level of human action. Palliard describes with clarity and eloquence the functional segmentation of reaching behavior ( Fig. 83-5 ). His work identifies three distinct stages of motor activity: (1) the eye-hand orientation that allows the foveation of the target and the acquisition of information necessary for processing the visual cues about the object’s features and its location in peripheral space; (2) the mobilization of the arm linkage for transporting the object toward the desired object, which involves mostly the movements of the shoulder and the elbow joints; and (3) the grasping movement itself. Palliard describes the functional segmentation ( Fig. 83-6 ) in reaching behavior in relation to two parallel visual channels subserving identification and location, respectively. Visually triggered programs for the hand grip and orienting arm movement may be assisted by visual guidance of grip positioning (wrist orientation) and visual steering of the arm trajectory in its initial and final stages of grasping (closed-loop condition). Tactile cues become the focus to assist manual grasping after contact has been made. Palliard provides an example of neural mapping for reaching behavior. Fig. 83-7 appears complicated, but observe how the special senses (vision, hearing, balance or vestibular components, the tactile sense, and proprioception) are interrelated with body segmentation (eyes, head and neck, trunk, forelimb, hindlimb, hands, and digits). The movements of the shoulder never can be segregated or isolated from other body segments during functional movement requiring eye-hand coordination. Therefore any functional approach to rehabilitation of the shoulder cannot be performed with isolated movements or exercises independent of linked segments.
The forward-flexed position biomechanically fixates linked body segments by displacing the instantaneous axes of joint rotation and reducing the degrees of joint freedom. The moments, forces, power, work, and effort of movement are significantly increased with higher energy costs during any and all body movements from the slumped position. No better example can be given than the altered and robotic movement patterns seen in the elderly and patients suffering from chronic pain.
Forward-Head Posturing: Clinical Implications
The slumped posture often is called the forward-head posture or the kyphosis-lordosis posture ( Fig. 83-8 ) and is the most common postural fault seen in clinical practice. Kendall describes this posture as having the head forward, hyperextension of the cervical spine, increased flexion of the thoracic spine (kyphosis) with an increased angle at the cervicothoracic junction, abduction of the scapulae, hyperextension of the lumbar spine (lordosis), anterior tilt of the pelvis, flexion at the hips, slight hyperextension of the knees, and slight plantar flexion of the ankle joints.
Darnell proposed a chronology of events leading to the development of the forward-head posture. Kendall’s concept was integrated into a flowchart ( Fig. 83-9 ) that shows the interrelationship and interdependence of the musculoskeletal structures in forward-head or slumped-posture position. This model looks at the relationship of this postural fault to altered breathing patterns, chewing and swallowing difficulties, and pathologic conditions at the shoulder.
The biomechanical imbalance produced by an anterior displacement of the center of gravity leads to adaptive changes in muscle firing patterns and in muscle length. The upper cervical spine is kept in extension to keep the eyes on the horizon. The suboccipital muscles adaptively shorten, and the atlantooccipital joints are maintained in extension, which severely restricts rotation at the atlantoaxial and C2 to C3 segments, where 50% of cervical rotation occurs. The remaining cervical segments are held in the hyperextended position, thereby also limiting the ROM in flexion, side-bending, and rotation. The deep cervical muscles weaken by reciprocal inhibition and are maintained in a lengthened state.
Because the head is extended and the body must establish its vertical relationship, the upper chest is depressed to bring the head forward toward the vertical. This mechanically reduces expansion of the thoracic cage during breathing and alters the position of the scapulae (abducted, slightly elevated, and tipped anteriorly, if the pectoralis minor muscle is short). Because the pelvis is tipped anteriorly, lumbar curve is increased slightly. The combination of thoracic depression and anterior tilting of the pelvis effectively shortens the rectus abdominus and limits the amount of functional rotation of the lumbar spine. The loss of trunk rotation effectively reduces the role of the internal and external obliques, especially with stabilization of the ribs when the diaphragm is being used in deep breathing. The loss of the synergistic relationships between the trunk rotators and the diaphragm allows the accessory muscles of inspiration (anterior and lateral neck musculature) to take on a primary inspiratory action with breathing. This results in a shallow and less expansive breath, which increases the energy costs for any movement or action. This can be observed easily with a hemiplegic patient and any patient with chronic back problems. It demonstrates the need to aggressively change the adaptive relationships among the head, spine, and pelvis to improve arm function and cardiopulmonary status.
The forward-head posture also has a dynamic effect on the movement of the upper extremity. This is produced by an altered positioning of the scapula. The scapulothoracic joint, controlled by 16 muscles, must have great mobility to accommodate the varied positions of the arm and hand, and must be able to stabilize when strength of the shoulder is required. As previously stated, in the forward-head position the scapula assumes an abducted, slightly elevated, and anteriorly tilted position, if the pectoralis minor muscle is short. The length-tension relationships for the rotator cuff muscle are altered in this position, decreasing the force-couple relationship between the deltoid and the rotator cuff, effectively reducing strength and stability at the shoulder. The arm adapts by internally rotating (round shoulders), the forearm is pronated, and the wrist and finger flexors take on a more dominant role with wrist and hand positioning.
The adaptive positioning of the entire body with the forward-head posture decreases shoulder elevation by decreasing the height of the subacromial interval and effectively reducing the amount of glenohumeral depression during shoulder elevation. This concept can be demonstrated easily with the following example:
Sit forward and erect in an armless chair. With your left hand, palpate the greater tuberosity of the right humerus. Slowly elevate the right shoulder through its full range and feel the humerus depress under the palpating finger. Note the extent of the depression as the arm moves from 70 to 110 degrees of elevation. Repeat the activity from the slumped sitting posture with the upper chest depressed and the head forward. Note that the amount of depression is significantly reduced, ROM is decreased, and muscular effort is increased.
Repeat the activity in the standing erect position and then the slumped position and note that the relationship is maintained. From a clinical standpoint, the effect of varied posture on glenohumeral depression is important because the forward-flexed posture is a primary cause of shoulder impingement resulting in the development of bursitis, tendinitis, and degeneration of the rotator cuff. The loss of glenohumeral depression combined with scapular abduction and internal rotation at the shoulder severely limits ROM in shoulder elevation and increases the likelihood of further impingement. The painful hemiplegic shoulder is a good example. A painful hemiplegic shoulder rarely is seen before 3 weeks after onset of stroke. The premorbid forward-head posture of the patient and the fact that passive ROM of the affected arm often is performed with the scapula in the abducted and elevated position increase the possibility of impingement. One may be able to assume that stroke patients who maintained their center of gravity before the stroke (a more erect posture), and whose hemiplegic arm was moved passively with appropriate care given by the therapist to reposition the scapula and trunk before the movement, may not develop a painful hemiplegic shoulder. This observation could explain why some hemiplegic shoulders become painful and others do not. Poor posture also could explain the development of adhesive capsulitis following a surgical procedure or a proximal humeral fracture.
The biomechanical adaptations to the forward-head posture have a profound effect on muscle action at the shoulder. When the shoulder is injured, surgically repaired, or diseased, pain and spasm protect the joint from movement. Attempts to move the hand with a painful shoulder result in the initiation of flexion or abduction by elevation of the scapula through the action of the upper trapezius. This is the typical shoulder-shrugging action seen with all painful shoulder conditions or with shoulder or scapula weakness. The synergistic patterning of muscle firing after injury or in the presence of disease is a common characteristic. Any injury produces a strong total body flexor response characteristic of that already described. This is an automatic mechanism and typically follows the same pattern as seen with the upper extremity flexor synergy with stroke.
In a stroke patient, the spastic hemiplegic arm assumes the characteristic positioning of shoulder elevation, scapular protraction, glenohumeral internal rotation, elbow flexion, forearm pronation, and wrist and finger flexion. This same pattern can be seen after fracture, after dislocation, with severe arthritis, after surgical repair, or after any painful episode. The synergistic patterns for both the spastic arm and the arm protected by pain appear to be the same. The patient with chronic pain of the shoulder best illustrates the commonality of the two patterns. If these patterns are identical, the pattern is programmed into the nervous system early in development. Voluntary actions then mask these primitive patterns as the child, adolescent, or adult learns a large variety of functional movements.
Some believe that flexion synergy seen with hemiplegia is a learned response in an attempt to move the arm in an environmentally induced action, rather then being produced by a central nervous system lesion alone. This assumption has led neurologic therapists to use early intervention strategies that introduce selective functional movements during the flaccid stage and during the stage between flaccidity and the development of a strong spastic pattern. If selective functional movement is not introduced early in the rehabilitation process, spastic synergy will dominate and reduce the possibility for functional gain.
Where the pattern is dominated by only a musculoskeletal injury and the neurologic system is intact, selective movements can be introduced early and produce a beneficial effect. Functional retraining of reaching, grasping, pushing, pulling, supporting, and gesturing is far more important than the introduction of exercise alone. The use of this concept will be key to the establishment of a functional approach to rehabilitation.
One of the major problems any therapist faces in rehabilitation of the shoulder is how to resolve the night pain that disturbs the patient’s sleep. There are two types of night pain: (1) positionally induced and (2) viscerally induced. Patients with painful shoulders almost always sleep on the uninvolved side. This places the arm in adduction. Ratburn and McNab demonstrated that the rotator cuff vasculature is severely compromised in this position. Night pain of this origin can be treated by having the patient sleep in the supine position with the arm abducted 45 degrees or in the side-lying position with sufficient support under the arm to bring the arm into neutral or slight abduction. Increasing the vascularity to the rotator cuff not only will reduce night pain but will enhance healing of all the subacromial tissues inflamed by narrowing of the subacromial space and subsequent impingement during shoulder elevation.
Questions to be Addressed
Lillegard and Rucker developed a diagnostic algorithm for acute ( Fig. 83-10 ) and chronic ( Fig. 83-11 ) shoulder pain. The differential diagnosis begins with the identification of the location of the patient’s pain. Travell has further delineated the pain patterns and trigger point locations for myofascial pain. One must know these pain patterns to develop an efficient evaluation strategy. Developing a functional rehabilitation protocol requires the therapist to answer some specific questions, which are addressed in the following sections.
Where is the pain, and can it be isolated to one or more tissues?
Is the pain continuous or reduced if the patient’s arm is kept still?
Is night pain and the lack of sleep a serious problem?
Does changing the position of the head, thorax, and/or pelvis reduce the perceived pain?
Does the pain intensify after activity?
Does the patient have cervical extension with weakened deep neck flexors?
Does the patient have tightness of the clavicular portion of the sternocleidomastoid, malaligning the clavicle?
Does the patient have tightness in the upper trapezius that malaligns the scapula on the chest wall, thus interfering with its mobility and proper stabilization?
Does the patient have tightness of the scalenes, elevating the upper ribs?
Does the patient have a significant midthoracic break that decreases thoracic cage mobility?
Does the patient have excessive activity in the upper trapezius, pectoralis major, teres major, rhomboids, and latissimus dorsi muscles?
Is the scapula excessively elevated with the glenoid rotated downward?
Does the patient have excessive recruitment of the upper trapezius with the inability to recruit the lower trapezius and serratus anterior?
Does the patient have decreased scapular mobility?
Does the patient have decreased stability of the scapula on the chest wall or decreased antigravity scapular control?
Do the latissimus dorsi and teres major substitute for the action of the lower trapezius?
Is the head of the humerus translated anteriorly compared with the uninvolved side?
What is the degree of spasm protecting the glenohumeral joint?
Can the movement of the glenohumeral joint be dissociated from the head and trunk, trunk and pelvis, scapula and thorax, and scapula and arm?
Does glenohumeral depression occur either actively or passively?
Does the patient have restricted active and passive movement that indicates the presence of adhesive capsulitis?
Is external rotation full or limited?
Is internal rotation full or as limited as external rotation?
How much selective functional movement does the arm have?
Does the patient recognize the arm functionally and attempt to use it, even in the disabled state?
Are any signs of shoulder-hand syndrome or RSD present?
Principles of Rehabilitation
Rehabilitation of the painful shoulder begins on the first day after onset. Preventing any patient from developing a painful, tight shoulder begins with good bed positioning and early intervention strategies to produce good head, trunk, scapula, and arm alignment in preparation for reaching and walking.
Goals of Rehabilitation
The key to rehabilitation is preventing, minimizing, or eliminating any postural malalignments, muscle imbalances, muscle and connective tissue shortening, or patient neglect that will affect recovery. The following sections present the primary goals of treatment of the painful shoulder.
Goal 1: Pain Modulation
The reduction of pain requires the patient and the family to understand the mechanisms producing the pain and how to minimize pain and improve function with movement and posturing. In addition, the therapist can use modalities (microcurrent, heat, or cold) to reduce pain. Antiinflammatory drugs and analgesics may be indicated for short-term use. In the case of RSD, oral steroids and/or stellate blocks may be needed. All modalities are secondary to the primary interventions of posturing, moving, exercising, and performing functional tasks when selective movement has been achieved.
Goal 2: Postural Orientation
Reestablishing the center of gravity over the base of support in sitting, standing, and walking and during functional tasks is important for minimizing associative movement reactions and to developing selective voluntary movements. The trunk must be elongated and extended to restore good trunk alignment. Rolling, moving from lying to sitting, and support sitting using both the involved and uninvolved extremities are indicated.
Flexing the upper cervical spine while extending the thoracic and lumbar spines will improve alignment in both sitting and standing. Reestablishing the center of gravity over the base of support will decrease tension on pain-sensitive (stretched) tissue and reduce the effect of pain on the development of abnormal synergistic movement reactions.
Goal 3: Realignment of the Scapula on the Chest Wall
The scapula must be realigned on the dorsal chest wall to allow the tight anterior structures to elongate. Two different therapeutic methods can be used to realign the scapula: (1) elongating the anterior structures by stretching or (2) stimulating the antagonist to the tight muscle and allowing reciprocal inhibition to allow lengthening through an active movement. The latter appears to be the easiest and most effective way to lengthen tight contractile structures. If the scapular muscles are too long and not synergistically active during functional movements, any stretching will only temporarily lengthen the tissues anteriorly, because as soon as the stretching ceases, the tight tissues will return to their original shortened position. This is the basis of the approach taken by Sahrmann.
The rotator cuff muscles, which control the axis of the glenohumeral joint, steer and glide the humeral head during functional movement. The length-tension relationships of the rotator cuff muscles are controlled by good alignment of the scapulothoracic joint. One cannot overestimate the rotator cuff’s role in providing glenohumeral depression to minimize impingement of the pain-sensitive structures. The strength of the deltoid muscle during functional movements is reduced 40% to 60% when the rotator cuff muscles are unable to control the glenohumeral axis within a 1-mm space. If the upper trapezius initiates elevation without glenohumeral depression, the deltoid will pull the head of the humerus superiorly. During the early stages of recovery, shoulder elevation must be stimulated by contraction of the rotator cuff muscles and of the deltoid without excessive action of the upper trapezius.
Goal 4: Development of Scapula Mobility and Stability
Relaxation of the scapulothoracic musculature is of prime importance to regaining scapula mobility and stability. The scapulothoracic joint is a prime area for disuse and neglect. Sixteen muscles attach to the scapula and control its movements. The rhomboids, middle trapezius, inferior trapezius, and serratus often are inhibited, allowing the scapula to tilt forward and assume the protracted position. This reactivation of these muscles is important to regaining the position of the scapula dorsally and for stabilization of the scapula when the arm is to be elevated. Proximal antigravity control of the scapula is necessary to gain control distally.
The scapula must be able to move freely on the dorsal chest wall to maintain the length-tension relationships of the rotator cuff muscles that provide mechanical advantage to the deltoid muscle when it is acting as a power muscle. The mobility index for the shoulder is so great and the position of the hand and the shoulder are so varied during functional activities that the scapula must move and stabilize in many different positions. Observers once thought that the ratio between glenohumeral motion and scapulothoracic motion was 2:1. For every 15 degrees of motion, 10 degrees of motion occurs at the glenohumeral joint and 5 degrees of motion occurs at the scapulothoracic joint. However, this ratio does not exist functionally because the scapula actually oscillates on the dorsal chest wall until the hand reaches its target, and then the scapula is stabilized at that point. The coordination of muscle contraction around the posterior chest wall during functional movement requires complex neuromechanical coordination among all body segments. The muscles attached to the scapula must undergo controlled synchronized firing coupled with antagonistic inhibition to position the scapula in the ideal position to be stabilized. The overactivity of the upper trapezius is a serious problem that malaligns the scapula and restricts its ROM. The inhibition or lack of selective control of the rhomboids, serratus anterior, and middle and lower trapezius maintains the malalignment.
Goal 5: Stretching of Tight Structures
Several soft tissue structures may be tight as a result of chronic shortening caused by positioning, including the anterior/inferior glenohumeral capsule, shoulder internal rotators, pectoralis major and minor, teres major, and latissimus dorsi. The stretch must be slow and passive to regain length. However, passive stretch alone will not be effective unless the antagonist of the tight muscles can fire functionally to produce reciprocal inhibition. Physiologic stretching by antagonistic action always will produce a more meaningful, lasting, and functional length. The latter is not accomplished easily without the reduction of associated reactions and development of selective movement of the involved joints.
Goal 6: Activation of Deltoid and Glenohumeral Depression without Action of the Upper Trapezius and Other Synergistic Patterns
The fine coordinated movement of the shoulder complex during functional activities of the cervical spine, trunk, scapula, arm, and hand requires the coordinated action of the rotator cuff. These cuff muscles are considered to be the intrinsic muscles of the shoulder because they guide and steer the humeral head, control the instantaneous axis of rotation, act as a force-couple mechanism to provide mechanical advantage to the deltoid muscle when it is used as a power muscle, seat the humeral head against the glenoid, and depress the humeral head so the greater tuberosity can clear the subacromial space. Because the upper trapezius is so active, during elevation of the dysfunctional shoulder, the glenohumeral depressors will become inactive. The combined action of the deltoid and upper trapezius without glenohumeral depression by the rotator cuff muscles must be corrected to achieve a full coordinated action of the shoulder. This represents the final and necessary tuning of selective movement before a truly functional ROM is restored. This step is the only way to prevent impingement of pain-sensitive structures. Synergistic patterns of muscle contraction that must be activated to restore functional movement include (1) deltoid and rotator cuff without upper trapezius, (2) serratus anterior with the shoulder flexors, and (3) lower trapezius with the shoulder abductors.
Goal 7: Restoration of Thoracic Mobility and Proper Breathing habits
The loss of thoracic mobility caused by a midthoracic break, anterior tilt of the pelvis, lack of stabilization of the thorax by the oblique abdominals, and chronic use of the scalenes and sternocleidomastoid as the primary muscles of inspiration leads to increased energy expenditure during functional movements. The practitioner must teach proper body alignment with activation of the diaphragm during normal breathing to maintain good alignment and cardiovascular function. The realignment of the central key points (center of gravity) over the base of support (good posturing) provides the thoracic mobility and stability necessary to reduce the energy expenditure for normal movement.
Goal 8: Neuromotor Training
The final step in neuromotor rehabilitation, after associated reactions have been reduced and selective movements have been restored, is to integrate all neuromechanical systems necessary for the coordinated motor activity important to the patient (i.e., activities of daily living [ADLs]). This will require task- and environment-specific activities that fine tune the rehabilitation process and require the therapist to analyze the demands of the task within the framework of the patient’s goals, abilities, and disability.
Patients with painful shoulders, regardless of the cause, lose their ability to function efficiently. Motor programs for simple functional tasks are masked by a strong synergistic response of the muscles to protect the involved extremity from painful movement. The central program for the functional task is either eliminated, as occurs with severe chronic pain (e.g., in RSD), or modified significantly to produce an awkward and inefficient movement pattern. These abnormal patterns are so strong that they cause the patient to lose awareness of how to use the arm functionally. Exercises do not necessarily produce a functional response and, in fact, in many cases may slow a functional return by introducing movements and patterns of muscle contraction that inhibit early functional return. This is why the practitioner always must introduce functional movements before assigning exercises. In this way, the therapist uses exercises to enhance those functions that have been introduced rather than asking exercises to produce a meaningful functional outcome.
Patients with chronic pain illustrate the lack of functional carryover from exercises in that these patients are assigned many exercises but continue to be unable to function in their environment in a meaningful way. They continue to use robotic movement patterns, poor posturing, and inappropriate muscle contractions in their daily activities. The learned patterns of dysfunction are learned behaviors that require very specific interventions to change the neuroprogramming and to allow more efficient functional movement patterns to emerge. If functional outcome measures are to determine the extent of the rehabilitation outcome, a functionally based rehabilitation program must be introduced before exercises are administered. This does not mean that exercises are not important, but rather it is more important to establish the functional movement patterns first so that patients can use their increase in ROM and strength to perform functions that are important to them. Too often in physical and occupational therapy, exercises and/or tasks are chosen that have little carryover to the home environment. Patients do their exercises several times during the day but rest the arm in the interim. Rehabilitation protocols should require patients to use the affected arm as much as possible during the day within their pain tolerance so they can take an active role in their rehabilitation. Functional retraining must use simple functions at first (i.e., arm swing, reaching, support reactions) and then proceed to more complicated tasks of pushing, pulling, lifting, carrying, dressing, or writing. The therapist must be able to break each of these tasks into its component strategies and identify which components are present and which components must be introduced to allow more efficient movement patterns to emerge with practice and repetition.
The Pace of the Rehabilitation Process
Establishing a predictable and realistic time course for any rehabilitation protocol is difficult because of the variability of human behavior. The risk factors that may slow rehabilitation progress in the earliest stages of rehabilitation must be identified. These risk factors may include (1) a significant period of stress, anxiety, or depression before the injury; (2) a sedentary lifestyle; (3) uncontrolled hypertension ; (4) Raynaud’s phenomenon or hypersensitivity to cold; (5) negative childhood behaviors (e.g., resulting from sexual, physical, and/or emotional abuse; abandonment by one or more of the primary care providers; and a family history of drug or alcohol abuse); and (6) negative adult behaviors (e.g., grief; unhappy marriage; need to provide care for an ill or dependent parent; family stress from problem siblings, children, parents, or relatives; workaholic personality; low self-esteem; negative outlook on life; and sleeplessness). When the patient has risk factors to a significant degree, they can produce a slow, painful, and possibly chronic rehabilitation course. These risk factors must be identified and possibly corrected to determine the most appropriate and timely rehabilitation course.
Regardless of etiology, the aforementioned negative behaviors can lead to the development of RSD or shoulder-hand syndrome, and they always must be considered before any rehabilitation protocol is initiated. RSD is one of the most painful and debilitating conditions that can afflict patients. Because its cause is not known, it is difficult to treat and presents a dilemma to the referring physician and the rehabilitation practitioner. Shoulder-hand syndrome progresses through stages, with each subsequent stage being more severe and debilitating. The following list presents the characteristics of the stages.
Severe burning pain
Stiffness and limited joint mobility
Hyperhidrosis (drying of the skin)
Increased hair and nail growth on the involved side
Osteopenia on x-ray examination
Increase of pain with manual contact and emotional stress
Pain becoming more severe and diffuse
Allodynia: all stimuli perceived as painful
Hyperalgesia: increased sensitivity to pain
Hyperpathia: increased threshold for pain, but after the threshold is exceeded, the pain sensation increases in intensity more rapidly than expected
Edema spreading and changing from soft to indurated and brawny
Hair becoming scant
Nails becoming brittle, cracked, and heavily grooved
Increased joint thickness
Hyperhidrotic cool skin with cyanosis
Osteoporosis and cystic and subchondral bone erosion on x-ray examination
Irreversible marked trophic changes
Intractable pain, often involving the whole limb
Atrophy of muscles
Weakness of interphalangeal joints of hand with decreased ROM, even ankylosis
Contraction of flexor tendons, occasionally producing subluxations
Significant and diffuse bone demineralization
Thin and shiny skin
Thickened palmar fascia
Shoulder-hand syndrome is an interesting condition in that the patient develops a form of sensory motor amnesia in which reflex inhibition of both sensory and motor neurons at the spinal-cord level blocks nerve conduction both proximally (sensory) and distally (motor). The altered sensation effectively produces a distal extremity that does not feel or act like a hand. The weak and uncoordinated voluntary movements during functional tasks lack appropriate muscle and proprioceptive synergism. This syndrome seems to have a pattern of neglect similar to that demonstrated by patients with a hemiplegic arm. The patient does not know how to use the extremity for functional tasks and keeps the arm still because of the severe and diffuse arm pain and associated lack of awareness. Swelling occurs because of the inability of the arm muscles to contract effectively, reducing the efficiency of the venous and lymphatic pumps and causing fluid to accumulate distally. Swelling of the hand, a difficult problem to treat, dramatically slows the rehabilitation process. Activation of muscle during functional activities is the key to rehabilitation, and contraction during close-chain movements seems to be more effective in activating muscle and proprioceptive synergism than open-chain movements or exercise. Therefore weight-bearing movements or exercises are more appropriate to stimulate functional use in the early stages than open-chain movements.
The key to timely and cost-effective rehabilitation is to minimize the effects of behavioral risk factors or at least prevent the debilitating effects of RSD. One must evaluate and treat the behavioral elements leading to RSD at any level early to attempt to halt or slow its clinical progression.
Adhesive capsulitis is probably the most common shoulder rehabilitation problem seen in the clinic. It is characterized by a limitation of both active and passive ROM. Three different movement patterns—capsular, diabetic, and RSD—limit both the active and passive ROM. Cyriax identified a capsular pattern of limitation of shoulder motion, in which external rotation is limited more than abduction or internal rotation. This is the typical pattern that we see in our patients with frozen shoulders. The capsular pattern of limitation of movement is predictable because of its developmental origin, and it seems to be genetically hard-wired into the nervous system. Any painful injury or disease, anxiety, or depression produces a strong total-body flexion response with a subsequent increase in tone of the upper trapezius, scapular protractors and depressors, glenohumeral internal rotators, elbow flexors, and wrist and finger flexors. An antalgic posturing of the extremity to prevent movement and reduce pain, and a primitive pattern for resting the shoulder after injury or disease cause adhesion formation and shortening of connective tissue and muscular structures, and produce abnormal muscle firing patterns during functional activities. The flexion synergy further displaces the center of gravity forward over the base of support, changing the patient’s balance strategy from an ankle strategy to a hip strategy, which is the least efficient mechanism for balancing in the standing position and which may predispose older citizens to falls.
The overall efficiency of walking, reaching, and breathing also is severely altered within this postural arrangement, which requires greater effort to walk and exacts higher energy costs. When watching patients with a frozen shoulder ambulate in an open space, the abnormal postures and walking patterns can be observed easily. Because such patients are limited by pain in external rotation of the humerus, they almost always turn their bodies away from the painful side and, when turning from supine to side-lying, almost always turn toward the uninvolved side. Even from the sit-to-stand position, one can observe patients leaning toward the uninvolved side and placing more weight on that leg and hand when rising to the standing position. Reaching to either side is limited by the depressed chest and the mechanical locking of the lumbar spine in flexion. This is a postural biomechanical disaster that leads to severe limitation of motion, altered muscle synergies, and loss of function. The capsular pattern of limitation is not functionally static but rather produces learned patterns of dysfunction that can be habituated and that slow functional return.
The diabetic pattern, first identified by Bowles and characterized by Tenhula et al., exhibits severe limitation of both internal and external rotation when compared with abduction. This pattern is seen with patients suffering from diabetes mellitus, alcoholism, and hyperthyroidism. The peripheral neuropathy associated with diabetes and alcoholism may play a primary role in the development of this pattern, but with hyperthyroidism the cause is not known. Regardless, the patient experiencing this pattern is difficult to rehabilitate by conservative means and may require manipulation after injection or after an interscalene brachial plexus block to gain a functional ROM. Of all frozen shoulders, 15% demonstrate this pattern, which is difficult to rehabilitate without breaking the adhesions restricting the movement. In addition to the pattern of limited movement, these patients also may demonstrate a significant loss of scapulothoracic movement (hypomobility) and a loss of scapulohumeral rhythm. The patient experiences pain deep within the axilla and not, as is more common, over the lateral arm. The patient also can exhibit significant atrophy of the posterior deltoid muscle. Posterior deltoid atrophy has been observed early in the disease process because the posterior deltoid has less contractile volume (bulk) than either the anterior or middle deltoid.
Reflex Sympathetic Dystrophy and Metastatic Pattern
The third pattern seen with frozen shoulder is RSD with metastatic disease, in which the arm is limited by severe and diffuse pain and swelling that extends below the elbow. The arm usually is fixed against the patient’s thorax and is severely limited in ROM and function. Accompanying this painful pattern is multimyotomal weakness of the arm and especially in the intrinsic muscle of the hand (i.e., abductor digiti and dorsal interossei). Because osteopenia is associated with RSD and altered bone status is associated with metastatic disease, radiographic examination is indicated because the manual techniques that are used to increase ROM or function may produce pathologic fractures that further complicate the rehabilitation process.
Principles of Rehabilitation
Because many patients do not develop a frozen shoulder after injury or surgery, we must ask why some patients develop this problem and others do not. We suspect that the premorbid activity level of the patient; the patient’s ability to control his or her center of gravity during walking and functional activities; the patient’s motivation to return to recreational, family, or work-related activities; the absence of major risk factors; and the use of early functionally based intervention strategies probably decrease the incidence in this population.
Many patients are referred for shoulder rehabilitation after the problem has become well established and more difficult to resolve. Involving the patient in his or her rehabilitation from the time of onset will reduce many complicating factors that can slow the rehabilitation process. Patients must be educated about their problems, ways to modulate pain and to improve functions at home, and their responsibilities in the rehabilitation process. Allowing the patient control, responsibility, and flexibility in his or her rehabilitation at home provides an environmental stimulus to use the arm and hand functionally during ADLs. The home program must allow the patient to be aggressively treated four to six times per day, and patients should be encouraged to normalize their ADLs and social contacts as soon as possible, even at a lower activity level. During the waking hours, the arm and hand should be used in functional positions in which pain is not a limiting factor. Retraining arm swing during gait and performing low-level reaching produce the greatest amount of movement of the shoulder and hand in the early stages of rehabilitation and take the painful arm out of the dependent rest position.
Meaningful and measurable goals for rehabilitation must be established. This is best achieved by asking the patient one simple question: “If I treat you, how will you know you are better?” The answer must be functionally based and not related to a decrease in pain. If pain reduction is the primary response of the patient, then follow with the question, “If the pain is reduced or eliminated, what will you be able to do tomorrow that you can’t do today?” Patients who have realistic functional goals that they can measure usually have the motivation to comply with the proposed rehabilitation program. Compliance is always easier if the patient has a reason to comply and can measure whether the program is making a difference in how he or she moves or lives. Patients who do not have functional goals usually are dominated by their illness behavior and show slow progress and noncompliance with their program. This is evident especially with patients with chronic pain who are only passively involved with their rehabilitation program and are dominated by, and dependent on, their pain experience. Such patients typically say that when the pain goes away, they will be able to function at a higher level—a rehabilitation formula for chronicity, dependency, and failure.
Reduction of Pain and Spasm
Patients with shoulder pain usually follow the axiom, “If it hurts, don’t move,” and therefore place the arm in the antalgic position, resting on the anterior chest wall, and tilt the head slightly away from the painful side. The scapula is abducted or protracted 1 to 3 inches beyond the normal resting position of 3 inches from the spinous processes, which increases the anterior angulation of the scapula 10 to 15 degrees beyond the normal of 30 degrees. In addition, the scapula is elevated in the coronal plane. The abnormal positioning of the scapula causes the humerus to be adducted and internally rotated to place the arm tightly against the chest wall. The upper chest is depressed, with an increase in the thoracic kyphosis. The lumbar spine and hips are flexed slightly, and there is anterior shift of the center of gravity. When the patient lies supine on the treatment table, the increased flexor tone is observable and can be felt with palpation. The increased tone is the body’s protective response to pain that splints the glenohumeral joint and prevents it from moving. The subsequent muscle guarding and spasm restrict the motion of the cervical spine, the thorax, the glenohumeral and scapulothoracic joints, and the pelvis.
Changing the Flexion Synergy
The increased body tone is reduced with a method called effort substitution. Patients are placed supine on the treatment table with both hands placed over the iliac crests. Towels are placed under the head ( Fig. 83-12 ) to position the cervical spine in neutral. Additional towels are placed under both scapulae, and the posterior arms and forearms, and small towel rolls are placed in the hands. Rolls are placed under both knees and outside the ankles to position both hips in a neutral position ( Fig. 83-13 ). The towels and rolls allow the patient to relax in a comfortable position. Supporting the body parts allows the muscles involved with the flexor synergy to relax, reducing pain and muscle spasm. The therapist then performs a series of gentle oscillatory movements to increase the relaxation, first with gentle cervical distraction ( Fig. 83-14 ); then with movements that oscillate the body totally from the feet ( Fig. 83-15 ); and finally with a diagonal push and pull on a single leg ( Fig. 83-16 ). The movements are performed at 15- to 20-second intervals at each location at an oscillatory rhythm of two to three cycles per second. Ounces of force are used to produce relaxation. As the patient relaxes, change can be seen in his or her breathing pattern as upper chest inspiration changes to diaphragmatic breathing. As the patient breathes quietly and gently, the systemic muscle tone continues to be reduced, and the patient lies comfortably in a pain-free state. The patient is not mesmerized nor hypnotized, but rather senses a calmness and relaxation away from the painful contraction of muscle. This altered metabolic state is essential to the reduction of pain and spasm and the activation of functional movements.
With the patient in this altered state, the passive ROM of the glenohumeral joint is tested and recorded in flexion and external rotation. The point in the ROM where the pain and spasm are initiated is noted and recorded. When lifting the arm to measure the ROM, the examiner must feel the weight of the patient’s arm as a single mass. If the patient painfully splints the glenohumeral joint away from movement, the arm is not relaxed, and further oscillations are needed. An average of 5 to 7 minutes of oscillations are needed to relax the patient and gain control of the arm.
After the patient is relaxed, specific glenohumeral mobilizations can be performed to further increase the ROM in elevation and external rotation. In the first phase of rehabilitation, the patient should gain 70 degrees of passive, pain-free elevation in one or two treatment sessions. The oscillatory mobilizations can be directed to specific areas of the glenohumeral capsule. These techniques do not stretch the capsule but rather provide proprioceptive awareness of the movement in specific directions. Graded oscillations of different amplitudes are used to produce the neuromechanical oscillations. Fig. 83-17 shows an inferior glide oscillation for improving flexion and abduction. The arm is kept in 15 degrees of abduction to keep open the circulation to the rotator cuff. The force is applied to the greater tuberosity through the thumbs, and the oscillations are produced at the therapist’s shoulders and move through the extended elbows to the thumbs and to the greater tuberosity. Fig. 83-18 is an anterior/posterior oscillation to improve flexion, extension, and internal and external rotation. The arm is placed in 15 degrees of abduction and in a neutral position for flexion, extension, and internal and external rotation. If the patient has a capsular irritability in external rotation, the therapist can place the glenohumeral joint in slight internal rotation to reduce the pain and protective spasm. When the therapist lifts the arm gently off the table, the weight of the patient’s arm can be felt. The therapist produces a gentle oscillatory motion by bouncing the elbow on the soft surface of the table or towel. As the pain and spasm decrease, the glenohumeral joint can be positioned in more flexion ( Fig. 83-19 ) and the oscillations are continued. Fig. 83-20 illustrates an alternate way of producing an anterior/posterior glide of the glenohumeral joint. If the therapist’s hands are placed more proximal with the thumbs under the spine of the scapula, the oscillation force is directed toward scapula protraction. A rotation mobilization ( Fig. 83-21 ) can be performed by placing the patient in the prone position with a towel support under the anterior shoulder. The glenohumeral joint should be placed in a neutral position. When the therapist picks up the distal arm with both hands, he or she should feel the weight of the arm in his or her hands and the glenohumeral joint should be free to rotate internally ( Fig. 83-22 ) and externally ( Fig. 83-23 ).