Manipulation and traction are methods of reducing pain and disability caused by internal derangement of intervertebral cartilaginous structures. This treatment undoubtedly stems from man’s experience with those examples of lumbago, low back pain and sciatica that were relieved by a sudden twist or a fall. Schiötz and Cyriax ²⁰ described several examples in which patients reported such a sudden relief of symptoms after an unintentional movement. One of our colleagues reported a patient who recovered from backache on falling down her basement cellar steps. Undoubtedly, such experiences have given rise to various, sometimes bizarre mechanical treatments in folk medicine, such as trampling on a patient’s back, wrestling the back, striking the back with the weight of a steel bar or standing back to back ‘weighing salt’ (two persons stand back to back, hooking their arms at the elbow; each person bends forwards in turn, lifting the other from the floor). These measures seem non-specific and possibly harmful but Schiötz commented: ‘It seems justified to infer that methods found effective by the natives of parts of the world as far apart as Norway, Mexico and the Pacific Islands over many, many centuries must be valid.’ These primitive methods are also the models for more modern techniques of manipulation and traction taught and practised by trained doctors and therapists all over the world.

Although different methods and different theoretical concepts developed, this seemed not to matter much as all claimed their successes. Cyriax ²¹ was convinced that, in nearly all cases, a similar mechanism is at work – a dislocated fragment of disc returns to position or a protrusion is ‘sucked back’. He also stated: ‘Spinal manipulative methods of orthopaedic medicine are perfectly straightforward and possess an explicable intention directed to a factual lesion. They are based on the discovery of disc lesions as the primary cause of degenerative change and pain of spinal origin.’

It is hard to believe that, even nowadays, many doctors still refuse to recommend manipulation of the back as the treatment of choice in nearly all cases of acute low back pain and sciatica. Avoiding discussion of the topic of spinal manipulation is also unwise, since a significant number of their patients will have received or will be considering this form of treatment: in the USA in 1980, 120 million surgery visits were made to chiropractors.²²

An increasing number of controlled clinical trials on well-delineated subgroups have been published, which compare the results of manipulation to different forms of placebo therapy as well as to other forms of conservative management.23–34 The positive effects of manipulation appear to occur either immediately after the manipulation session or within the first 4–6 weeks of treatment.^35–37

Definition of manipulation

In orthopaedic medicine, spinal manipulation is defined as a method of conservative treatment by passive movements, carried out with a single thrust or sustained pressure, in order to return a displacement to its proper position (see Ch. 5, p. 95). During the thrust, an audible click is often produced. This may accompany immediate relief of symptoms and signs, which supports the precision of the diagnosis and selected therapy.

There are different schools of thought in manipulation, which derive from different attitudes towards spinal disorders. The theories on which manipulation is based include the reduction of disc protrusions,38,39 the correction of posterior joint dysfunction,⁴⁰ the mobilization of blocked vertebral joints,^41–44 the reduction of nerve root compression,^45,46 the normalization of reflex activity^47–50 and the relaxation of muscles.⁵¹

The objective of the manipulative techniques discussed in this book is to alter the discodural or discoradicular interaction by moving a displaced cartilaginous rim away from sensitive structures. The method relies on two principles:

• The lumbar spine is positioned in such a way that the affected intervertebral joint opens, i.e. gives the loose fragment room to move.

• The manipulation is usually carried out during traction. This tautens the posterior longitudinal ligament and causes suction in the disc, so exerting a centripetal force (Fig. 40.3).

Fig 40.3 Effects of long-lever manipulations. Positioning of the spine ‘opens’ the intervertebral space. A combination of traction and rotation produces a torque with tautening of the posterior longitudinal ligament and helicoidal traction on the lateral part of the annulus.

Most of our techniques are non-specific, long-lever manipulations: the force is exerted on a part of the body some distance away from the area where it is expected to have its beneficial effect. Leverage enables the manipulator to apply more force at the affected level. The normal joints are moved as far as they can go. The posterior longitudinal ligament becomes taut. The moment that resistance of the blocked joints and the taut ligament is felt, a quick additional thrust is given, to act at the affected level. Manipulation of the lumbar spine either is quickly successful or fails. If, after one or two manœuvres performed in a certain direction, signs and symptoms remain unaltered, another direction or another technique is tried. If these also prove ineffective, manipulative treatment is abandoned. If, by contrast, manipulation has led to reduction, both local and referred discomfort cease. Previously painful movements immediately become normal. So it is the patient, rather than the manipulator, who judges the effect of treatment.

Most other manipulative schools claim to work more selectively, i.e. on the affected level only. They claim to have developed the clinical skills to localize, by palpation, the exact site of the ‘fixation’ or ‘locking’. Several studies have failed, however, to demonstrate the reliability of this approach.52–57 We support McKenzie’s⁵⁸ conclusion that demystification of spinal manipulative therapy is an urgent priority. Both chiropractice and osteopathy thrive by creating the impression that there is something complex and exclusive about the practice of passive end-range motion that only chiropractors or osteopaths can understand or have the skills to ‘feel’. They generate the belief that, in order to become skilled in the understanding and delivery of spinal manipulative therapy, it is necessary to undergo 3 or 4 years of training.⁵⁹ This suggestion is undermined by the fact that the majority of lay manipulators in Britain have never had any tuition at all and yet have amassed many satisfied clients and also very rarely figure in actions for damages.²⁰

The main advantages of the methods discussed in this book are, first, that they are much simpler but at least as effective as those advocated by chiropractors and osteopaths. Second, it takes only about 180 hours of tuition, provided that the student has completed medical or physiotherapy studies.

In orthopaedic medicine, the manœuvres are always intended to relieve the current cartilaginous displacements. This is in contradiction to other methods, where a protocol of regular or intermittent manipulation sessions is commonplace. The type of displacement, as well as the patient, has to be assessed before any kind of manipulative manœuvre is undertaken.

• The displacement should be cartilaginous, not too large and not placed too far laterally. Soft nuclear protrusions are seldom reduced by manipulation unless they are small and very recent, and the technique of manipulation is changed to sustained pressure. If the consistency of the displacement is not quite clear, with symptoms and signs pointing in opposite directions, it is worthwhile making one attempt at manipulation. During the first session, it is usually quickly apparent whether reduction by this means will prove feasible or not. If it fails, traction is substituted the next day. Reduction of cartilage displacements, together with full relief of symptoms and signs, has proved to be possible in two-thirds of all cases of backache and in one-third of all cases of sciatica.⁶⁰ Just about half of all lumbago cases are relieved in one treatment.^61,62

• The patient must be mentally stable and keen to get well. If this psychogenic aspect is neglected, on some occasions a patient may be treated who claims to have been made worse by a type of therapy that is regarded in retrospect as unacceptable. Hence it is important to avoid these active methods of treatment when the patient’s attitude appears to be more important than the minor mechanical disorder found on examination.

Indications for manipulation

History and clinical examination almost always supply sufficient information to select those cases suited to manipulation (see Box 40.3, below).

Acute annular lumbago

The attack is initiated by a click in the lower back, followed by sudden agonizing lumbar pain fixing the back in slight flexion or lateral deformity. The pain radiates in an extrasegmental way and there are marked dural signs and symptoms.

Reduction should always be attempted, except in hyperacute cases where the attempt proves impossible to bear. An epidural injection is then substituted and followed by manipulation the next day. If such an injection is refused, it is still possible for the patient to recover in about a fortnight with bed rest and the use of McKenzie’s extension mobilizations ⁵⁸ and anti-deviation techniques. The moment the process has ceased to be hyperacute, treatment by manipulation can be tried again.

Backache

Acute or recurrent backache that has started suddenly usually responds well to manipulative treatment. There are a number of symptoms and signs which indicate that manipulation is likely to be successful.

The description of a click and sudden pain in the back on bending forwards or on coming upright from a forward-bent or sitting position indicates displacement of a small cartilaginous fragment. Another ‘favourable symptom’ is the patient’s age because, over the age of 60, nuclear protrusions no longer occur and a hard and mobile fragment of disc material is very likely.

‘Favourable signs’ are:

• A partial articular pattern in which some movements are only painful at extreme range: for example, flexion, extension and side flexion away from the painful side

• The existence of a painful arc with or without momentary deviation

• Absence of gross deviation caused by muscle spasm on standing or during as much flexion as the patient can tolerate

• Absence of gross limitation on movement – gross deviation or limitation of movement always requires several sessions of manipulation (Box 40.1).

Box 40.1 Symptoms and signs favouring manipulative treatment of backache

However, some small protrusions do not respond well to manipulation. Patients under 60 years of age, in whom pain is greatest on pinching the lesion by side flexion towards the painful side, are usually ‘unfavourable’. If such a manœuvre causes pain in the lower limb instead of the lumbar region or upper buttock, manipulation nearly always fails. A better response will be achieved with traction.

Sciatica

Reduction proves possible in about one-third of all patients with sciatica. Again, several symptoms and signs indicate those patients who can be expected to respond well (Box 40.2).

Box 40.2 Symptoms and signs favouring manipulative treatment of sciatica

Patients with mixed protrusions

If neither the symptoms and signs nor the patient’s age establish the consistency of the displaced fragment, manipulation should be tried first; it quickly becomes clear whether this is effective or not. In contrast, traction usually requires a week before efficacy can be determined.

Patients over the age of 60

The rules that determine a nuclear or an annular lesion are no longer applicable over the age of 60 years. The nucleus has become hard and dry, and will react correspondingly. Bed rest is wholly ineffective because there is no great difference in intradiscal pressure on lying and standing, and as the tension of the posterior longitudinal ligament lessens, it becomes elongated and loses its ability to apply a correcting centripetal force. Hence, all discodural or discoradicular interactions in this age group are best treated by manipulation, as this is the only way to achieve reduction.

However, for elderly patients, manipulative treatment should also be adapted. This means that only one or two manipulations are performed during a session. The interval between two sessions is also extended: say, to once a week. The intensity of each technique should not be changed, however, otherwise the centripetal forces acting on the joint at the moment of manipulative thrust are too small to influence a displacement.

Lumbar manipulations are not contraindicated in osteoporosis. However, some techniques are not used for fear of fracturing a bone.

Contraindications

These may be divided into circumstances in which manipulation is absolutely contraindicated and those in which manipulation is of no use, although not harmful to the patient (Box 40.3). Appropriate selection of patients and choice of techniques can avoid such serious complications as have been reported.^63,64

Absolute contraindications

Danger to the fourth sacral roots 65,66

Although these roots lie in the centre of the spinal canal, well protected by the posterior longitudinal ligament, they may be threatened by a massive central lumbar disc prolapse which has caused considerable bulging and possibly partial rupture of the posterior longitudinal ligament. Manipulation may rupture this ligament completely, causing extrusion of the entire disc.

The important symptoms of an S4 syndrome are rapid progression of bilateral sciatica and neurological symptoms in both legs. Pain and paraesthesia in the perineum, rectum, genitals or anus are other symptoms suggesting this menacing lesion. Finally, bladder weakness, causing frequency of micturition without a strong urge, loss of rectal tone and faecal incontinence result.

Acute lumbago and bilateral sciatica with compression of the nerve roots at the same level are examples of a large central protrusion in which bulging of the posterior longitudinal ligament is to be expected. Such a protrusion may also cause spinal claudication. These patients have symptoms during walking, immediately relieved by lying down.⁶⁷

In all these conditions, a high-force rotational technique in the side-lying position could rupture the last protecting fibres of the posterior longitudinal ligament with massive extrusion of the entire disc.

Anticoagulant medication

Manipulation of patients on anticoagulant therapy may lead to an intraspinal haematoma.⁶⁸ A patient who has a clotting abnormality should also not be subjected to forceful manipulations.

Aortic graft

Although there have been no reports of damage at the junction of a graft in the aorta by manipulation, its presence is reasonably regarded as a contraindication to any forceful manipulation – rotations or extension thrusts.

Final month of pregnancy

During the final month of pregnancy, lumbar manipulation is impracticable. Bed rest or epidural local anaesthesia is an alternative possibility. There is no bar to manipulation during the first 4 months; after this, extension techniques which necessitate a prone-lying position are omitted.

Manipulation should not be employed at any time if there is any predisposition to miscarriage.

Weakened bone structure

This includes unstable fractures, severe osteoporosis, osteomyelitis, multiple myeloma and bony tumours.

Muscle spasm

Forced movements should never be used when there is obvious muscle spasm, which may indicate some kind of joint irritability, fracture or metastasis. Therefore, any sign of muscle spasm at the moment the lumbar spine is positioned and the manipulator starts to stress the affected intervertebral joint should be taken as a sign to stop the procedure immediately. Of course, such a warning sign should already have been detected by an adequate clinical examination.

Mental state

Patients with serious mental disturbance should not be treated by manipulation, even if there is an underlying disc protrusion which seems suitable for manipulative treatment. The symptoms largely exist in the patient’s mind and indicate certain needs – compensation or protection – and it can be expected that manipulation will fail or even aggravate the disorder. In addition, an obsessional neurosis about vertebral displacement may result. This not only may make manipulation troublesome to undertake but also may cause the patient to blame the treatment for making matters worse.

Cases in which manipulation is not useful

Too much pain

In hyperacute lumbago, manipulation is impossible because of unbearable pain on any attempted movement. These patients are best treated by the immediate induction of epidural anaesthesia and manipulation is postponed until dural irritation has mostly ceased.

Too large a protrusion

There are two categories.

Protrusions causing impaired root conduction

All such protrusions are impossible to reduce by either manipulation or traction. If clinical signs of muscle weakness, cutaneous analgesia or reflex disturbances are present, the protrusion is too large (and located too far laterally) to be replaced.38,69–73 Epidural local anaesthesia is the treatment of choice.

Sciatica with an excessive flexion or side flexion deformity

In such cases, an attempt to move in the contrary direction causes pain to shoot down the leg. Surgery is indicated.

Too soft a protrusion

Nuclear protrusions causing backache and sciatica do not respond to manipulation (except in small and very recent cases, and provided that the manipulation technique is changed to sustained pressure). The consistency of the protrusion is too soft to be influenced by a quick thrust. Traction is the treatment of choice. The history is rather typical and usually identifies this type of disc lesion.

Acute nuclear lumbago is also an example of a protrusion that is too soft to manipulate with a thrust. The history is of pain that began gradually, after much stooping and lifting, and became slowly worse over the next few hours. The following morning, the patient wakes unable to get out of bed because of severe lumbar pain. The patient is always under 60 years old and, although manipulation is indicated, it must be exerted by sustained pressure. If this makes the patient better, techniques should follow in the supine, side-lying and standing positions to correct a persisting lateral deviated position of the trunk (see pp. 554–556).

Alternatively, an epidural injection can be tried, again followed the next day by manœuvres to correct a lateral deformity. However, if these measures all fail, constant pelvic traction, in a supine position and continued for some days, is called for, slowly changed to periodic half an hour daily traction. Extension mobilizations, as recommended by McKenzie,⁵⁸ have also been found to be effective; this treatment is based on McKenzie’s hypothesis that flow or displacement of fluid, nucleus or sequestrum can occur within the intact annulus of the intervertebral disc as a result of prolonged or repetitive loading. This most commonly occurs with flexion loading. He recommends well-defined extension forces in order to reverse the direction of flow or displacement.

Too long a duration of root pain

In patients under 60 years, 6 months of root pain have been shown to be the limit for manipulative reduction. However, if backache persists together with root pain, manipulation should be tried because these patients often do not recover spontaneously.

Compression phenomena

Central stenosis, lateral recess stenosis and the ‘self-reducing’ disc protrusion do not respond to manipulative treatment. In stenosis, the underlying condition is the reason that attempted manipulation or traction fails.71,74,75 The self-reducing disc protrusion, with symptoms at the end of the day only, may be reduced by manipulation but will prove to be transient anyway.³⁸

After laminectomy

New protrusions, at the same level, are seldom successfully reduced by manipulation. Traction is often more effective.

Unfavourable articular signs in a patient under 60 years

In backache, side bending towards the painful side is an unfavourable sign for manipulation but this does not apply in lumbago.

If any movement other than flexion hurts in the lower limb instead of the back, manipulation is almost certain to fail.

If an attempted extension manipulation gives rise to root pain, the manœuvre should be discontinued, as the protrusion is pressed harder against the nerve root.

Primary posterolateral protrusions

These protrusions all consist of nuclear material and therefore cannot be reduced by manipulation (see p. 458).

Dangers of manipulation

Lumbar manipulation is quite safe.63,64 The most frequently reported serious complication is further prolapse of a herniated disc, resulting in a cauda equina syndrome. However, the risk of spinal manipulation causing a cauda equina syndrome is estimated to be less than 1 per 3.7 million treatments.^65,76 Most of the incidents were described in patients undergoing manipulation under anaesthesia or chiropractic adjustments.⁶⁶ Long-lever, high-force rotation techniques in the side-lying position are regarded as responsible. This is only partly true: the underlying cause is the lack of adequate examination to rule out unsuitable disorders. If, in contrast, manipulative procedures are instituted after a thorough examination, those described in this book have never led to severe accidents. The main advantages of these manipulations are:

• A great deal of traction is used to exert a strong centripetal force on the intervertebral joint.

• Movements towards flexion, which can intensify potentially harmful centrifugal forces, are excluded.

• Each manœuvre is followed by a fresh assessment of dural, root and articular signs, which affords a clear pointer to what has happened inside the joint and what the next step should be.

Other complications, such as sprains of the costovertebral and costochondral junctions or fractures of a transverse process, are less serious and either the result of poor technique or inappropriate indications. Should they arise, spontaneous recovery is to be expected after a short period of, say, 4–8 weeks.

Side effects, remarks and precautions

If it becomes clear, after the history and clinical examination, that the orthopaedic problem is less important than the psychological one, the patient is best left untreated. Even if such a patient can be helped, the improvement will not persist. The moment the patient realizes that a cessation of symptoms may have adverse consequences, a postmanipulative mental crisis is to be expected, blaming the treatment.

Manipulation should cause only minor discomfort, which is due to stretching effects on soft tissue structures. In acute lumbago a more gentle start is often necessary, to assess patients’ reaction and to gain their confidence.

Sometimes the patient leaves a session pain-free but, for the following 2 days, a rather strong reaction follows. However, examination during the next consultation shows that symptoms and signs have decreased or even disappeared. Therefore patients should be warned of some after-pain, which is due to muscular and/or capsular–ligamentous reactions. It disappears within 2–3 days and is unrelated to the lesion.

Elderly patients can be manipulated safely. However, the number of manœuvres during one session should be confined to, say, two or three. In these patients high-force, long-lever techniques should also be omitted.

Results

Acute lumbago

If there is no deviation on standing or in the maximal forward-bent position and the lumbago is of recent onset, 50% of patients will get well with one treatment.61,62 Potter⁷³ noted that 93% of such patients either fully recovered or were much improved following manipulation.

Backache

Manipulation either works fast and leads to a swift disappearance of symptoms and signs, or has no effect at all. If there is no clear improvement after one or two sessions, this treatment form should be abandoned. Uncomplicated low back pain of recent onset seems to be significantly more responsive to manipulation than are chronic cases.29,73,74,77,78 The beneficial effect of manipulation is particularly significant in low back pain with dural signs.⁷⁹

Sciatica

The success rate of manipulation in an uncomplicated discoradicular conflict is about 1 in 3 (J. Cyriax: personal communication, 1982). This result was confirmed by one report that claimed complete recovery in 18 out of 50 chronic cases with unilateral sciatica after one manipulative session.⁶⁰ Others found that 25–75% of their patients with uncomplicated sciatica recovered or improved considerably after spinal manipulations.^{39,62,70,73,79}

Manipulation techniques

The manipulative techniques used in orthopaedic medicine can be divided into three groups:

• Rotation

• Extension

• Antideviation.

After a detailed description of each technique, the reader will find a ‘practitioner’s checklist’ regarding choice of technique, assessment of progress, repetition of techniques and the course of a manipulative session.

Rotation techniques

Rotation strains have been shown to be very effective in reducing displacements at a low lumbar level. A session of manipulation therefore always starts with these manœuvres. First, a ‘stretch’ is performed, being the smallest rotation strain. The patient lies with the painful side uppermost in order to bring the joint surfaces apart on the side of the displacement. Then, if necessary, this technique is followed by stronger rotations, using the femur as a lever. However, the latter techniques are impracticable in patients with arthritis of the hip or in elderly patients in whom osteoporosis is suspected. If the displacement lies centrally, straight leg raising may indicate which side should be treated first. However, in the absence of any symptom or sign to indicate the side, either side can be treated and the manipulator proceeds by trial and error.

Five different rotation techniques are described, all of which are used frequently.

Stretch

The couch should be stable and adjustable to about 30 cm in height. Using a high couch makes it impossible to aid the distraction of body weight, which in turn decreases the effect of manipulation. The patient lies on the painless side. The upper thigh is flexed to a right angle with the underneath leg extended.

The manipulator stands behind the patient, level with the patient’s waist. One hand is placed in front of the shoulder and rotates the thorax backwards and upwards as far as it will go. At the same time, the heel of the other hand, placed against the greater trochanter, rotates the pelvis forwards and downwards to the same extent. This brings the joint surfaces apart on the side of the displacement.

By using the body weight and leaning well over the patient, the manipulator obtains considerable distraction at the lumbar joints. At the moment the limit of tissue tension is felt, the manipulator’s body is pushed forwards on the vertically outstretched arms to apply overpressure (Fig. 40.4). At that moment a ‘click’ or ‘snap’ is nearly always heard and felt, after which the result of the manipulation is assessed.

Fig 40.4 Stretch.

Practitioner’s checklist

• Angles of rotation should be equal

• Hands positioned with fingers pointing in the direction of movement

• Elbows extended

• Equal amount of pressure on both hands

• Body weight over the patient but assess depending on build

• Manipulation is a stretch, not a rotation

Leg crossed over

The curved arrow symbolizes a ‘leg crossed over’. The direction of the arrow, to the left, indicates that the patient’s trunk is rotated to the left posteriorly. The letter R indicates that the patient lies on the right-hand side.

The couch should be stable and adjusted to about 60 cm in height. The patient lies supine about 20 cm or a hand’s breadth from the edge of the couch.

Practitioner’s checklist

• This manipulation creates maximum rotation but little distraction

• Use of leverage and gravity means that only minimal force on the knee is needed

• In stiff patients, it is impossible to hold the far shoulder on the couch; with the pelvis rotated as described, the far shoulder is more or less off the couch and held in that position

• The degree to which the leg is flexed depends on posture in the standing position: in patients who show considerable forward and/or lateral deviation of the lumbar spine, the hip must be well flexed, up to 120° if necessary. Thus end-feel is reached more effectively, making the manipulation more likely to be successful. Patients without deviation are often best treated initially with the thigh brought up to rather less than 90°

The manipulator stands on the painless side, level with the patient’s waist, facing the feet. With both hands flexing the thigh on the far side up to 90° and drawn forwards, the pelvis and lower back are rotated towards the operator. In this way, hip adduction is avoided. The ipsilateral knee of the manipulator is applied to the pelvis, if necessary, to prevent the patient from falling from the couch. Next, the contralateral forearm is turned into supination and the palm of the hand applied to the outer side of the knee. The other hand pushes the patient’s far shoulder flat on the couch (Fig. 40.5). Then rotation of the pelvis is continued until tissue tension is felt to be maximal. At that moment, rotation is forcibly increased by pressing the patient’s knee strongly and with high velocity towards the floor, using the thigh as a lever. At the same moment, the other hand maintains the position of the patient’s far shoulder (if possible) flat on the couch.

Fig 40.5 Leg crossed over: (a) with the leg forwards to rotate the pelvis; (b) shortly before manipulation.

Leg crossed over with side flexion

This manœuvre is a variation on the previous ‘leg crossed over’ technique and also achieves side flexion. The patient lies supine, both legs flexed and crossed, the leg on the painful side underneath. The manipulator stands on the painless side, level with the patient’s waist. Holding the patient’s knees in the hands, the manipulator moves the patient’s hips into 90° of flexion. Then both legs are twisted, in order to tilt the pelvis laterally and open up the lumbar spine on the painful side. This position is maintained at full range. The hand that has been on the patient’s uppermost knee is now freed to fix the far shoulder on the couch. The side-bent position of the lumbar spine is ensured by the manipulator’s thorax and abdomen, which are used to engage the knee from the side. Next, rotation is stepped up slowly: under the influence of gravity, the legs turn in the direction of the floor until the limit of tissue tension is felt at the end of range. At that moment, the manipulator’s thigh, engaging the uppermost knee from the side, has taken over to secure the side-bent position of the lumbar spine. Lastly, the hand at the knee is supinated to increase the manipulative force. Manipulation is performed by pressing the knee quickly downwards (Fig. 40.6). At the same moment, the other hand is used to maintain the position of the patient’s far shoulder flat on the couch, if possible. Rotation is thus forced during side flexion.

Fig 40.6 Leg crossed over with side flexion: (a) flexion of the hips and knees and side flexion of the pelvis; (b) rotation of the pelvis; (c) manipulation.

Practitioner’s checklist

• This manœuvre must not be performed in arthritis of the hip and in the elderly, where osteoporosis is to be expected

• Hip of the leg underneath should be flexed to 90°

• Lateral flexion should be maintained up to and including the final thrust

• Manipulation creates a quick but short downward thrust, without releasing the patient’s far shoulder

Stretch Reverse stretch

Again, the couch is adjusted to about 30 cm in height. The patient lies on the pain-free side, close to the edge of the couch where the manipulator stands. The patient’s upper hip is extended, the lower flexed to about 45° in order to stabilize this position. The upper arm hangs off the couch, the lower lies behind the back. The manipulator stands behind the patient, distal to the pelvis and facing the patient’s head. The ipsilateral hand takes hold at the anterior iliac spine and twists the pelvis backwards as far as it will go. In this position, the manipulator’s arm is fully pronated, with the hand placed against the anterior aspect of the anterior iliac spine, pushing the pelvis downwards and backwards. The other hand is placed against the scapula and pushes the thorax upwards and forwards (Fig. 40.7). Next, as the manipulator leans well over the patient, the joints are distracted by moving both hands in opposite directions, until tissue tension is felt to be maximal. Manipulation is performed by jerking the body downwards over the rigid arms. It is best to apply this overpressure at the moment of expiration.

Fig 40.7 Reverse stretch: (a) the ipsilateral hand twists the pelvis backwards; (b) manipulation.

Practitioner’s checklist

• Pelvis must be rotated well backwards; otherwise the manipulation compresses the pelvis instead of stretching the lumbar joints

• Angles of rotation should be equal

• Same amount of pressure is exerted by both hands

• Elbows are extended at the moment of the final thrust

• Body weight is brought well over the patient to reinforce the stretch but adapted according to the patient’s constitution

• Manipulation is a stretch, not a rotation

Reverse rotation with thigh

The couch is adjusted to about 60 cm in height. The patient lies on the pain-free side, the upper leg extended and the lower hip flexed to 60°, with the lower arm behind the back. The manipulator stands behind the patient, level with the lumbar spine. The ipsilateral hand grasps the upper thigh at the knee and flexes the hip to 90°, abducting the thigh horizontally. As a result, the pelvis is twisted as far as it will go. The other hand is placed against the scapula and pushes the upper thorax to the couch (Fig. 40.8). While pressure is maintained on the thorax, the patient’s upper thigh is now brought to 60° of flexion and full abduction. In some cases, it is also necessary to place a knee against the patient’s lower buttock, to prevent the pelvis from slipping backwards. The moment that the manipulator feels the limit of tissue tension, manipulation is performed by a short, sharp rotation of the manipulator’s body. This forces the arm at the thorax down, at the same time as it jerks the thigh backwards. Strong rotation and extension occur in the lumbar joints.

Fig 40.8 Reverse rotation with thigh.

It is obvious that this manipulation must not be performed in the elderly or in patients with arthritis of the hip or osteoporosis.

Practitioner’s checklist

• Suitable technique for lightly built manipulators

• Contraindicated in hip lesions, elderly patients and osteoporosis

• Upper thigh is in not less than 60° of flexion

Extension techniques

These techniques are very effective in small cartilaginous displacements that cause backache, especially in elderly patients and in those with persistent minor protrusions following incomplete reduction by a stretch in a rotated position. The techniques are milder than those performed with rotation strains and may substitute for the latter in osteoporosis. However, they affect one segment only.

During the manœuvre it is thought that interspinous pressure moves two adjacent vertebrae apart, so tightening the posterior longitudinal ligament and causing suction in the disc – a centripetal force which may reverse a displacement. The shape of the facets of L5–S1, which are more in a frontal plane, contributes to the better results achieved at this level than at the other lumbar segments. The impulse separates L5 from S1 in a cranial direction, instead of compressing the dorsal parts of the joint.

In acute lumbago and when extension pressure causes pain to shoot down the limb, extension techniques are contraindicated.

Deviation, as an expression of a large displacement, indicates that these techniques will almost certainly fail.

If a heavily built manipulator is dealing with a light patient, leaning on the patient’s back using the whole weight of the body may give rise to strong resistance and therefore the amount of weight applied should be reduced.

Central pressure

The patient lies prone on a firm couch adjusted to about 30 cm height. The manipulator stands level with the lumbar spine, facing the patient, with the knees against the edge of the couch. One hand is placed with its ulnar border at the interspace of two adjacent spinous processes (normally between S1 and L5). The other reinforces it with the heel pressing on the radial and the thumb pressing on the dorsal and ulnar sides of the lower hand (Fig. 40.9). To prevent any contact with the iliac bones, it is useful to use the right hand, standing at the patient’s left-hand side, and to turn this hand through about 45°. With the upper limbs extended and kept rigid, the manipulator leans well on to the patient’s back and extends the knees, one after the other. From this moment the body weight presses fully on the patient’s back and results in maximum tissue tension.

Fig 40.9 (a, b) Central pressure; (c) the ulnar border of the hand is placed between two adjacent spinal processes.

Practitioner’s checklist

• Weight applied should be adapted when a heavily built manipulator deals with a light patient: one or even both knees remain in a flexed position and lean against the edge of the couch

• Lean well over the patient

• Arms extended and kept rigid

• If pressure of body weight shoots a pain down the limb, stop the manipulation

At the moment the patient relaxes and some extension has been achieved, the final thrust is given by bending the head and thorax abruptly forward. Usually a thud is felt or a click is heard if the manipulation is successful.

Unilateral pressure

If repeated central pressure has neither fully relieved the patient nor made the problem worse, this technique is used immediately after central pressure.

The manipulator stands on the patient’s painful side, although, if the pain is central, there will be no indication whether to start on the right or on the left. The wrist of the ipsilateral hand is extended and the prominent pisiform bone is used to exert localized and unilateral pressure at the base of the spinous process of L5 or L4. It is necessary to lean well over the patient, in order to press in a slightly oblique direction (Fig. 40.10). The other hand reinforces the pressure, using the heel to press on the manipulating hand. In order for the manipulator to stay well balanced, both legs are moved slowly backwards at the moment the body moves forwards. The knees or thighs should stay in contact with the edge of the couch. Manipulation is all but identical to the previous technique, except that the thrust is now directed medially as well as downwards, which opens the joint on the painful side, at the same time also exerting some rotational stress and strong extension.

Fig 40.10 (a) Unilateral pressure; (b) the prominent pisiform bone is placed at the base.

Practitioner’s checklist

• Arms kept rigid

• The centre of gravity moves beyond the centre of the patient’s body

• Knees or thighs stay in contact with the couch

• Final thrust at the moment the patient relaxes and all the slack in the tissue has been taken up

• If pressure of body weight shoots a pain down the limb, stop the manipulation

Unilateral pressure with thigh I

This is a much stronger technique, which follows the previous extension manœuvres but is undertaken only if partial reduction has been achieved and repetition affords no further improvement. In the absence of any benefit, it is unwise to continue with this technique or the next one.

The patient lies prone and near to the edge of a low couch. The manipulator stands on the pain-free side, level with the pelvis. With the ipsilateral hand, the front of the knee is grasped at the painful side around its lateral aspect. The ulnar border of the other hand is placed just above the posterior spine of the ilium. Then the hip is extended and strongly adducted by leaning heavily towards the patient’s head (Fig. 40.11). This opens the joint on the side where the displacement lies. Manipulation is performed by a quick rotation of the manipulator’s trunk towards the patient’s head. In this way, the unilateral downward pressure of the lumbar hand and the upward pull of the hand on the knee are considerably intensified. This results in a combined movement of hyperextension, side flexion and rotation at the lower lumbar joints.

Fig 40.11 Unilateral pressure with thigh I.

Practitioner’s checklist

• Only use if the previous extension manœuvres have already achieved some reduction

• Not possible in a lesion of the hip joint

• Size of the patient should be comparable to that of the manipulator

Unilateral pressure with thigh II

With a heavily built patient, added force can be exerted by employing the knee; however, this technique should not be adopted if the previous extension strains have led nowhere.

The patient lies in the same position as in the previous technique. The manipulator stands on the painful side. With the contralateral hand, the front of the patient’s knee is grasped around its medial aspect and the thigh is extended and adducted until the pelvis rises just off the couch. The palm of the other hand is placed on the sacrospinalis muscle covering the fourth and fifth lumbar levels on the painful side, with the forearm fully supinated (Fig. 40.12). The manipulative thrust is performed by pressing the ipsilateral knee with the hand at the same time as the patient’s thigh is forced into full extension and adduction. A forced extension at the lower lumbar joints results.

Fig 40.12 Unilateral pressure with thigh II.

Practitioner’s checklist

• Use only if central and unilateral pressure techniques have achieved some reduction and the patient is heavily built

• Not possible in a lesion of the hip joint

Unilateral distraction

This technique is indicated if the previous manœuvres towards extension have helped but reduction has still not been fully achieved. It may also serve to remove a generalized ache that results from any manipulative manœuvre.

The patient lies prone and side-flexes the body to open the joint on the painful side as far as possible. The manipulator stands on the concave side, facing the patient, with the arms crossed and the elbows bent almost to a right angle. The heel of one hand is placed against the iliac crest, just lateral to the sacrospinalis muscle. The heel of the other hand is placed just under the lowest ribs (Fig. 40.13). To prevent the skin from being strained at the moment of manipulation, it is first pulled upwards with the lower hand, while the upper hand does the same downwards. Manipulation is now performed by repeated (10–20 times) forward movements of the trunk, keeping the elbows rigid. This forces the hands apart and imparts rhythmic further distraction, together with some extension at the lumbar level.

Fig 40.13 Unilateral distraction.

Antideviation techniques

These techniques are applied in backache and lumbago with an adapted posture, caused by posterocentral disc protrusions. The previous rotation and/or extension techniques will have already eased the pain, but when the patient stands for a few moments, the tilt of the trunk to one side quickly returns, as a result of persistent one-sided muscle spasm. On examination, side flexion towards the contralateral convex side, and sometimes extension, is still limited.

Three techniques can be used:

• Side bending in the supine position

• Rotation–distraction in side lying

• Side gliding, standing.

Side bending

The patient lies supine with both legs flexed and crossed, the leg on the concave side of the lumbar spine underneath. The manipulator stands on the convex side, level with the pelvis. With one hand the upper knee is pushed away, while the other is used to pull the lower knee towards the manipulator (Fig. 40.14). This simultaneous action tilts the pelvis and achieves full side flexion at the lumbar spine in the direction that was previously blocked. It is quickly repeated a number of times, whereafter the pressure is maintained for a few seconds. When there has been a previous nuclear protrusion, the extreme of range is better maintained for a minute or so. This position is consolidated either with the assistance of the manipulator’s ipsilateral knee, pushing from a distal position against the patient’s ischial tuberosity, or by using the contralateral knee to push from a lateral position against the patient’s pelvis. The manipulation is repeated until the patient can keep the trunk in a neutral position on standing.

Fig 40.14 Side bending.

Rotation–distraction

The patient lies on the side of the lumbar convexity with the upper thigh flexed to about 60°, thereby rotating the pelvis to just over 90°. The manipulator stands in front of the patient, distal to the pelvis and facing the patient’s head. The thigh of the uppermost lower limb is clasped between the manipulator’s knees, just proximal to the patient’s knee, to secure the position of the pelvis. Both hands are placed to one side of the upper thorax. Correction of the lateral tilt is achieved by pushing against the patient’s thorax in an upward and backward direction (Fig. 40.15). This correcting force should be sustained for as long as the patient can endure it. The manipulator must stand well balanced to prevent the entire body weight from pressing on the patient. After some repetitions the patient is re-examined in the standing position.

Fig 40.15 Rotation–distraction.

The manœuvre is repeated until correction has been achieved or until repetition affords no further benefit.

Side gliding

The patient stands upright, the feet about 20 cm apart to provide a stable base, with the elbow held against the lower rib cage on the side of the lumbar concavity. The manipulator stands on the same side and presses the thorax against the patient’s elbow, with the hands placed on the far side of the patient’s pelvis.

Correction and even slight overcorrection is achieved slowly, by pressing the thorax against the patient’s elbow, simultaneously pulling the pelvis from the far side towards the manipulator (Fig. 40.16). This pressure should be maintained for a couple of minutes and is repeated several times. It is essential that the movement is side-gliding rather than side-bending.

Fig 40.16 Slide gliding: (a,b) correction; (c) extension.

Once the spine is upright, an attempt is made to restore lordosis. To this end, the patient is brought into the corrected position again and asked to let the hips move forwards at the same time as the trunk bends backwards. In this way, the body stays well balanced all the time. This movement is repeated, until the range of extension is restored.

It will take at least 3–4 consecutive daily sessions to produce a lasting result. In addition, it is essential to instruct the patient in self-correction (Fig. 40.17).

Fig 40.17 Self-correction.

Standing in front of a full-length mirror, one hand is placed against the lower lateral rib cage at the concave side of the lumbar tilt. The other hand is placed on the opposite lateral iliac crest. Then the patient performs the side-gliding movement of the pelvis in the restricted direction so as to correct the deformity. Once this has been achieved, a controlled extension movement is performed: the patient supports the trunk by placing the hands at the lower back and slowly bends backwards as far as is comfortable. These exercises should be repeated every hour.

Practitioner’s checklist

• Correction involves side gliding, not side bending

The course of a manipulative session is summarized in Figure 40.18.

Fig 40.18 Course of a manipulative session (see text for explanation).

Manipulation procedure

Choice of technique

• Stretch or reversed stretch in the side-lying position is usually the first technique to be tried, especially if the pain is unilateral.

• Extension techniques are chosen first in case of central pain and in elderly patients.

• Acute lumbago is unsuitable for extension techniques; rotation manœuvres usually give good results.

• Usually, L3–L4 protrusions respond better to rotation techniques.

• L5 protrusions may respond better to extension techniques, especially in elderly patients.

• In elderly patients it is also better to avoid long-lever techniques for fear of fracturing weakened bones.

• If rotation–stretch techniques lead to incomplete reduction in minor protrusions, one can move on to extension techniques.

• After the use of each technique, the result is assessed and a decision taken whether to continue with the same technique or to change. If one manœuvre has helped, it should be repeated until symptoms and signs no longer alter. Then another is tried. Experience, the result of each particular manœuvre, end-feel during exertion, the patient’s age and estimation of tolerance all affect the types of manœuvre employed.

Assessment of progress (Box 40.4)

The physical signs are reassessed after each manœuvre. If there were dural signs, these are tested first: limitation of straight leg raising, before the manipulation, is tested afresh after the manœuvre. If there was pain on coughing, the patient is asked to cough once more and to describe how it feels.

It is only after the physical signs in the lying position have disappeared that lumbar movements are examined on standing to assess articular signs. In contrast, if it was only lumbar movements on standing that caused pain, the patient is asked to stand and bend backwards and sideways to judge whether there is any change in the degree of pain or in the amplitude of trunk movement. Because forward flexion movement is likely to increase any displacement, flexion of the trunk should not be tested before the manipulator can be sure that full reduction has been secured, i.e. all other tests have become negative.

Another important sign is ‘centralization’ of pain: a shift, after manipulation, to a more central position is regarded as an improvement.58,80,81

Assessment of outcome after each manœuvre assures the manipulator that:

• The displacement has moved.

• It has moved into the right direction.

Repetition of techniques

One manipulative session continues until symptoms and signs have been modified to the greatest extent possible. If a ‘thud’ is felt on performing an extension thrust and considerable improvement is noted, no more should be done until the next visit. If there is only a slight improvement, the same technique should be repeated until no further change occurs. In young patients, another technique can still be tried. However, 6–8 manœuvres are about the maximum a patient can tolerate per session. In elderly patients, it is better to stop after two or three and to continue on the next visit.

Sometimes a manœuvre may make a patient’s symptoms worse. If this is the result of an extension technique, a rotation manœuvre can still be tried. If rotation makes it worse, the direction of rotation must be changed (reversed stretch instead of stretch, or vice versa). If these measures are unsuccessful, the question arises as to whether the patient is in fact suitable for further manipulative treatment. Manipulation is by no means free from danger if continued in spite of warning signals. Starting gently, paying attention to the patient’s statements while being manipulated and re-examination after each manœuvre will avoid mistakes.

Traction

Although there is still a great deal of controversy about the effectiveness of traction,⁸² we still consider passive sustained stretching of the low back as the treatment of choice for nuclear, reducible disc protrusions causing backache and/or sciatica, unless there are specific contraindications.

Despite the poor design of most of the studies,⁸³ traction has been shown to be more effective than corsets, bed rest, hot packs and massages.^79,84–86

Historical note

The Ancient Egyptians utilized the beneficial effect of axial traction.⁸⁷ An illustration of traction employed by the Spanish–Arabian physician Abu’L Qasim (1013–1106) of Cordoba is reproduced by Schiötz and Cyriax in their book on manipulation past and present.²⁰ In the same book, illustrations show the way in which traction was used by Hippocrates (400 bc) and Galen (ad 131–202). A 14th-century method of manipulation during traction is illustrated in Figure 40.19.

Fig 40.19 Manipulation during traction in Albucasis’s Surgery. Latin translation. Vienna, 14th century. (By courtesy of Wellcome Trustees.)

Nowadays, two methods of performing traction are practised. The sustained manner, as described in this book and first suggested by Cyriax in 1950,⁸⁸ and several types of intermittent traction. Intermittent traction can be done either electrically, manually (by a therapist) or by the patient (autotraction). However, nearly all reported work has shown all types of intermittent traction to be ineffective.^89,90

Effects of sustained traction

Several studies have investigated axial traction. It has been established that during sustained traction at least three effects result (Fig. 40.20).

Fig 40.20 Effects of sustained traction. (a) Before traction, nuclear protrusion posteriorly displaced. (b) During traction: the intervertebral space enlarges; the posterior longitudinal ligament is tautened; a negative intradiscal pressure is created. (c) After traction, reduction is maintained (after Mathews ⁹¹).

The space between the vertebral bodies enlarges

This is an important precondition for a displacement to recede.

In young men, sustained traction of 60 kg, applied for 1 hour, results in an increased body length of 10–30 mm, which is thereafter lost at the rate of 4 mm/h.⁹² In an excised lumbar spine, sustained traction of 10–30 kg increases each joint space by 1.5 mm.⁹³ Vertebral separation is greatest in those subjects with wide disc spaces and least where there is evidence of disc degeneration.⁹⁴

The effect of lumbar sustained traction on stature has also been studied in 10 healthy young subjects; the investigators confirmed the significant increase in stature but also that this increase was over and above that known to occur when the load is taken off the spine by lying down.⁹⁵ The findings suggest that most of the vertebral separation takes place in the first 30 minutes. It has also been established that the enlargement between two consecutive lumbar endplates during normal traction is between 1.0 and 1.5 mm, which is 10–15% of the thickness of the disc.^96–98 Other studies demonstrate a widening of the lumbar intervertebral space of between 3 and 8 mm measured on radiographs of patients undergoing gravitational traction.^99,100

The heavy lumbar paravertebral musculature normally exerts significant resistance to distraction. At least 30–35 kg of traction, not dissipated by friction, is required to influence the lumbar spine.¹⁰¹ Other work has demonstrated that a traction force of at least 25% of the body weight is necessary to achieve distraction of the lumbar vertebrae against the inertia of muscular resistance of the body.¹⁰² This supports an earlier study¹⁰³ in which any traction power less than 25% of body weight was regarded as a placebo.

The posterior longitudinal ligament is tautened, exerting a centripetal force at the back of the joint

The increasing tension in this ligament is certainly of great therapeutic value, particularly if the protrusion is located anterior to, and remains in close contact with the ligament. Traction will therefore be less effective if the protrusion is laterally placed – a conclusion confirmed by computed tomography (CT) investigation of the effect of static horizontal traction on lumbar disc herniations: ‘The clinical responses of the herniation to conservative treatment and the location of herniated nuclear material seem to be related. Traction is more effective on median and posterolateral herniation cases, and clinical improvement is evident in these cases, but traction is not very effective on lateral herniations.’¹⁰⁴ Also, re-entry of ruptured or sequestered disc material into the intervertebral disc is not possible (Fig. 40.21).

Fig 40.21 (a) Contained disc herniation. Examples of displaced nuclear material that remains contained by annulus fibres and/or an intact posterior longitudinal ligament. Traction is indicated. (b) Non-contained disc herniations. Nuclear material has been extruded into the spinal canal. Traction cannot reduce the protruded material.

Suction draws the protrusion towards the centre of the joint

It is believed, on the basis of biomechanical calculations, that significant intradiscal negative pressure may be produced during sustained traction.¹⁰⁵ A traction load of 30 kg caused a lowering of the intradiscal pressure from 30 to 10 kp in the L3 intervertebral disc.¹⁰⁶ In another study, intradiscal pressure demonstrated an inverse relationship to the tension applied. Tension in the upper range was observed to decompress the nucleus pulposus significantly, to below 100 mm Hg.¹⁰⁷ Discography has established that the decrease in intradiscal pressure causes a suction effect with centripetal forces on the contents.⁹³ An interesting Chinese study investigated the changes in intradiscal pressure and intervertebral disc height on 31 prolapsed discs under traction. It was demonstrated that the intradiscal pressure decreased as the intervertebral distance increased in most cases under traction.^108,98

Repair of the disc lesion

It has also been suggested 109,110 that, during episodes of disc decompression, nutrition is improved, reparative collagen is deposited and natural healing of annulus tears and fissures is promoted.

Sustained traction has the same effect on the intradiscal pressure as prolonged bed rest, but much more strongly. A few hours’ traction achieves as much or more than bed rest for weeks. Although the latter can also bring about slow reduction of a nuclear protrusion, traction has the advantage of speed. Instead of simply avoiding the compression produced by the upright posture, it mechanically distracts the joint. In addition, the patient remains ambulant, which is far preferable to bed rest for several weeks – the latter not only is bad for morale but also increases the cost in lost working days and the payment of sickness benefit.

It is important to emphasize that this effect can only be achieved if traction brings about more in the way of reduction in half an hour than can be reversed during the rest of the day by new loading effects. To this end, traction should be sustained and must be given daily and as energetically as the patient can bear. Otherwise, it is merely a placebo.

There is increased motor activity of the sacrospinalis muscles on an electromyogram during traction, until the mechanoreceptors in the tendons are stimulated.¹¹¹ From that moment, motor activity is inhibited, the intervertebral joint takes the strain and reduction of the pulpy mass starts slowly. Electromyographic silence is reached after 3 minutes. This suggests that traction must be sustained. A study that measured the intradiscal pressure during 30 seconds of passive traction performed by two therapists and during 2 minutes’ autotraction with 50 kg weight¹¹² has established that intradiscal pressure did not alter much in passive traction, whereas autotraction increased the pressure considerably. These findings strongly contrast with those of sustained traction, which makes it obvious that only the latter is able to diminish a nuclear protrusion in volume and return it to its normal position.

Indications for traction

Nuclear disc protrusions

Pulpy nuclear protrusions which remain contained and in contact with the posterior longitudinal ligament (see Fig. 40.21) are more effectively treated by traction just as hard annular protrusions are more readily treated by manipulation. Cyriax always said: ‘You can hit a nail with a hammer, but treacle must be sucked.’

It is important to emphasize that signs of irreducibility, such as neurological deficit or gross lumbar deformity maintained by root pain, should be absent.

The typical clinical pattern of a lumbar nuclear protrusion is as follows. The patient is under 60 years of age. On bending forwards for a while, some aching in the back is initially felt, which gets slowly worse later in the day. The next morning, it is impossible to rise out of bed because of severe low back pain. Clinical examination shows a partial articular pattern and movements pinching the lesion, i.e. side flexion towards the painful side or extension, are often most painful.

This is quite different from the patient with an annular protrusion, who describes a sudden onset of lumbago on bending forwards, perhaps lifting a heavy object. This may happen at any age. Again, a partial articular pattern is present. A painful arc, with or without momentary deviation, suggests a small mobile annular fragment and is a further encouraging sign for manipulation.

The distinction between these different types of disc protrusion is not always as clear as in these examples. Nevertheless, for therapeutic reasons, it is important to differentiate between these mechanisms. The summary given in Table 40.1 may be helpful.

Table 40.1

Differences between nuclear and annular protrusions

	Nuclear protrusion	Annular protrusion
History
Age (years)	< 60	All ages
Onset	After a lot of stooping and lifting Pain increasingly evoked by sitting in a kyphotic posture	During bending forwards and coming up again, abrupt displacement with a click, initiating acute lumbago
	Backache after exertion	Backache as soon as exertion starts
Clinical examination	Partial articular pattern	Partial articular pattern
	Pain on pinching the lesion in backache (not in lumbago or over the age of 60)	Pain on side bending away from the painful side

Indeterminate protrusions

If signs and symptoms do not indicate the consistency of the protrusion very precisely, manipulation is tried first because it may be effective much more quickly. If it fails or succeeds only partly, sustained traction is indicated. However, traction should not be given immediately after the manipulation but should be started the next day. There are two reasons for this:

• Painful twinges may occur after release if traction follows manipulation too quickly.

• It is not possible to judge the effect of the traction if it is combined with other forms of treatment at the same session.

First and second lumbar disc lesions

At this level, traction is regularly successful, whereas manipulation always fails. A disc lesion secondary to lower lumbar arthrodesis constitutes an exception: the fixation of the low lumbar segment alters the biomechanical relationship in such a way that manipulation can often succeed.

Recurrence after laminectomy

Manipulation can be tried first but seldom proves successful if the protrusion is situated at the level of the operation. Traction is more often effective, although the prognosis remains less favourable than for those patients who have not had surgery.

Primary posterolateral protrusions

These protrusions may constitute a difficult therapeutic problem. Because they all consist of nuclear material, manipulation has no effect. The protrusion usually responds to daily traction but the tendency to relapse after reduction is considerable. Epidural local anaesthesia is a good alternative but only works when the protrusion is at its maximum size.

Considering the therapeutic approach in a primary posterolateral protrusion, Cyriax ²¹ (his p. 317) gives the following advice: ‘[S]uch a protrusion of a month or two’s standing should be reduced by daily traction.’ However, if relapse occurs after a successful reduction, or the protrusion has persisted for 3 or 4 months, it is better to leave it where it is, especially in a young patient with slight pain only, which is the common situation. Cyriax again: ‘Spontaneous recovery usually takes nine months from the onset of root pain, and the strong tendency to recurrence is largely obviated by allowing the patient to get well of himself. [However] he should be kept under observation until the protrusion is stable at its maximum size, i.e. the range of straight-leg raising has stopped decreasing and is found unaltered at two examinations a fortnight apart. This is the moment for one or two inductions of epidural local anaesthesia which usually abolish the root pain in a few weeks.’

Backache together with bilateral long-standing limitation of straight leg raising

Those affected are mainly young adults with months or even years of backache associated with marked bilateral limitation of straight leg raising. Daily traction has proved to be successful but only if it is continued for about 3 months. It may take at least 1 month before any improvement is noted.

Pain with fourth sacral reference

Symptoms and especially signs of impaired conduction of the S4 root probably call for immediate surgical intervention. Nevertheless, if pain is the only symptom and paraesthesia, referred to the coccyx or the genital area, is absent, traction may first be attempted with caution. However, this measure is not entirely safe and should only be performed after discussion with the physician in charge.

Contraindications

Acute lumbago

Because lumbago with twinges is made so much worse for several days even by just one session of traction, traction is absolutely contraindicated. Even in patients with lumbago who have had no twinges recently, the first session of traction must be undertaken very cautiously. It is not during traction, but rather at the moment when tension is diminished, that the patient gets agonizing twinges, making it impossible to release the traction at that point. It may take the patient as much as 3 or 4 hours to rise from the couch. Cyriax advised a very slow diminution of the traction force. Then, once released, it can take a good 15 minutes for neck flexion to stop hurting and the range of straight leg raising to begin to increase. If twinges still persist on initiating movements, repeated rotation manipulations (leg crossed over) should be carried out, gently at first. If these fail, epidural local anaesthesia must be induced without delay.

Abdominal surgery or hiatus hernia

Patients who have had previous abdominal surgery or hiatus hernia may not tolerate abdominal compression.

Respiratory or cardiac insufficiency

The patient may not be able to lie down at all, let alone tolerate the harness.¹¹³

Respiratory irritation

In discodural interactions, pain may be evoked by coughing or sneezing as a result of the sudden increased expansion of the dura pressed against the protrusion. During traction, a powerful cough or sneeze is often painful too, making it impossible for the patient to relax. For this reason, respiratory problems should be dealt with first.

Painful reactions

If traction increases pain in the back or in the leg, the patient is unsuited to it. This often happens in chronic low back pain and sciatica, where the lumbar spine is fixed in flexion. Pain that increases as soon as the traction starts may indicate a worsening of the condition and is therefore reason enough to stop the treatment.¹¹⁴

A large protrusion

In gross lumbar deviation, any attempt to move in the contrary direction is aborted by sharp pain; traction is therefore not applicable. Discodural and discoradicular interactions with a high degree of inflammation (indicated by continuing pain or pain at night) are also best treated by epidural injections.

Mental state

Some patients cannot tolerate restrictions such as the harness used. The moment the thoracic harness is put on, they become very anxious and start to hyperventilate. In this case, it is better not to continue the treatment.

Cases in which traction is not useful

Although not really contraindicated, traction has no effect and should therefore not be applied in the following circumstances.

In patients with neurological deficit or gross lumbar deviation, the protrusion is larger than the pathway through which it emerged. Such material extruded through the outer rim of the annulus certainly cannot be reduced by traction or manipulation.

In the presence of a free protrusion or sequestered fragment, disc herniation traction will always fail.

In both these instances, an epidural injection may relieve the symptoms and, given adequate clinical and neurological justification, surgery may be indicated.

Long-standing clinical features

A primary posterolateral protrusion lasting longer than 3 months is better not treated by traction, especially in a young patient with only slight pain. There is a strong tendency to relapse after reduction, whereas spontaneous recovery takes place within 9 months of the onset of root pain.

A secondary posterolateral protrusion in a patient under 60 years old, causing root pain for more than 6 months, has passed the time limit for traction. The treatment of choice is epidural local anaesthesia. Spontaneous recovery is the rule and is to be expected within a few months.

Indications for and contraindications to traction, and cases in which it is not useful are summarized in Box 40.5.

Traction procedure

Traction apparatus

This consists of a couch and an electrically or manually operated machine, assembled at the foot of the couch.

In the non-electrical form, a spring balance, attached to the traction rope, is necessary. It measures the amount of traction and ensures an even degree of traction. Whenever the pelvic belt slips slightly over the bench, the buffer mechanism of the spring balance intervenes to take up the slack and subsequently optimum force must be restored. The attendance of a therapist is consequently required, especially during the first 15 minutes of application, when traction force is regularly lost. In the electrical form, an electric device obviates this, although the therapist should remain within hearing distance because automatic machines are not ‘fail-safe’.

There is mechanical equipment that has the advantages of the electrical appliance: constant traction force during the entire session (Fig. 40.22).

Fig 40.22 (a) Mechanical traction apparatus and harnesses for continuous lumbar traction. (b) Patient in position for continuous lumbar traction.

The couch should have an opening through which the patient can breathe when treated in a prone position. Frictional resistance in the system can be overcome by using a split-table model. However, the same results can be attained with an ordinary couch, although frictional resistance must be compensated for by a greater degree of traction force.

Thoracic harness

The thoracic harness is tightened over a layer of thick foam rubber or a folded towel, placed outside the patient’s clothing. This harness is placed as low down the thorax as possible, to confine the traction to the lumbar spine. Sometimes, though, patients cannot bear this position and fixation at a higher level is then required.

The harness anchors the patient at the head of the couch. The straps should clear the patient’s trunk when applied above (not shown). For this reason, the upright, at the head of the couch, should be at least 30 cm high.

Pelvic harness

This is also placed over a folded towel or layer of foam rubber, and encircles the entire pelvis. This belt takes its purchase from the iliac crest, and connects to the traction rope.

For the same reason, in order to clear the patient’s body, when applied above, the height of the traction rope should be adjusted to 30 cm.

Patient’s posture

The possibilities are prone or supine. Cyriax suggests the position that the patient finds most comfortable in bed (provided that this is either prone or supine) as the best guide for the first treatment. If this is of little help, the degree of pain and/or limitation during flexion and extension should be assessed:

• If both flexion and extension are painful (or neither), the patient may lie supine or prone. One strap is situated posteriorly, and the other anteriorly. In this way, the articular surfaces stay parallel during distraction (Fig. 40.23a–d).

Fig 40.23 Eight positions for lumbar traction (arrows indicate the position of the straps and direction of the pull).

• If flexion hurts and extension is pain-free, traction is carried out with the lumbar spine in slight lordosis. The patient is positioned supine with a small pillow supporting the lumbar spine, or lies prone. In both positions, the straps of the harnesses are situated anteriorly (Fig. 40.23e,f).

• If extension hurts and flexion is pain-free, traction is carried out with the lumbar spine in slight kyphosis. The patient is positioned supine with the lower legs supported on a small bench and the knees bent upwards, or lies prone. In both positions, the straps of the harnesses are situated posteriorly (Fig. 40.23g,h).

Thus, eight different positions may be used. In practice, however, the positions regularly used are:

• Supine (seldom prone), in kyphosis, lordosis or neutral position

• ‘Psoas position’, with the hip and knee joints flexed to 90°. This position is particularly indicated in L3 root lesions.

Instructions to the patient

Before the session, eating a large meal should be avoided. The patient is informed about the underlying disorder and the therapeutic effects that can be obtained with sustained traction, and is also encouraged to stay as relaxed and as still as possible during the stretch. Coughing and sneezing should be avoided, because either may cause a twinge and after-pain. The patient is further assured that traction should never be painful while it is being applied; nor should it increase pain that already exists. Any discomfort should be reported at once. This explanation will allay any fears engendered by the similarity of the apparatus to a medieval torture rack.

Force and duration

Theoretically, the pull is as strong and lasts as long as the patient can tolerate. Usually, this will be 40–60 kg, applied for at least half an hour to be effective. In practice, the amount of traction is measured in relation to body weight. Traction with 60% of the body weight is considered to have the best ratio of effectiveness to tolerance.

Interval between treatments

Traction must be given daily, as the intention is to achieve greater reduction in half an hour than the patient can reverse by weight bearing on the joint for the rest of the day. It is remarkable that this should be possible. Furthermore, traction is normally still effective with an interval over the weekend, provided that the patient relaxes on these days – lying down at regular intervals and not sitting for a longer time than is absolutely necessary. If these rules are neglected, the protrusion is likely to have returned to the same size by the time of the next visit. In an urgent case, traction can be performed more than once a day. Alternatively, one long stretch of up to an hour has proved effective in these circumstances. Passive back extension and prone lying can also be useful supplements.⁵⁸ If traction has not begun to have an effect after a few days, the question is whether to use a different position on the couch or a different position for the straps. If these alterations are made and the patient has not begun to improve after 2 weeks, traction should be abandoned. The physiotherapist should not despair too soon, however, as many patients only begin to improve during the second week. Obviously, if a patient is much better but not completely well after a fortnight, a third week’s daily treatment is justified. In young adults with backache and long-standing bilateral limitation of straight leg raising, it may even take a month before any improvement is noted.

Procedure (Box 40.6)

(Re)examination

The patient must be (re)examined before each session. Symptoms and some physical signs – for example, lumbar movements and straight leg raising – are noted afresh and any alteration given due consideration. However, there is no point in examining a patient immediately after a session of traction: the transient changes, for better or for worse, detected at this point mean very little.

Box 40.6 Traction procedure

1. Inform the patient about the technique, the mechanism and the possible side effects.

2. Examine the patient each time before traction is applied.

3. Look for the most comfortable position.

4. Apply the straps firmly but not too tight.

5. Start the traction slowly, reaching 30–35 kg after 1 minute:

First occasion: traction force may remain low.

Further occasions: increase to 60% of the body weight.

6. Keep the applied traction force steady for about 30 minutes.

7. Release the traction force slowly – over 2–3 minutes.

8. Allow the patient a few minutes’ rest before getting up.

9. Perform gentle active and passive movements of pelvis and hips before getting off the bench.

10. Patient’s back should be kept straight on getting off the bench.

11. Instruct the patient about activities of daily living and allow him/her to walk a short distance before getting into a car.

12. Give traction on a daily basis for 2–3 weeks. Treatment must be abandoned if no improvement is experienced after seven or eight sessions.

Traction

The patient then lies down on the couch in the position that is considered to be most effective and the harnesses are applied evenly and firmly but not in such a way that the patient tenses up. Finally, traction is started slowly, reaching 30–35 kg within a minute. As the patient becomes accustomed to the pull, traction can be increased to reach the maximum that is tolerable. However, on the first occasion, the traction force should be low – say, 30 kg for a person of moderate body weight – and applied for no longer than 15–20 minutes. Care should be taken, particularly in acute lumbago, when twinges have only recently ceased. The therapist should always be within earshot and from time to time must directly observe the patient. This helps the patient to feel that the situation is being continuously monitored. It is still useful to ask the patient regularly about any discomfort because, although instructions will have been given about reporting discomfort, some patients still accept painful reactions. Anything more serious than the minor discomfort caused by the harnesses should not be disregarded.

Release of traction

When traction has been completed, particular care is exercised while the tension is being released and the harnesses loosened, for this is the time when twinges are most likely to occur. To prevent this unpleasant experience, release is slow – over, say, 2–3 minutes. The harnesses are then loosened carefully, the pelvic one first because this exerts least influence on the lumbar segments. The patient is also reminded of the need to remain quite still and to resist the temptation to take a very deep breath as release is completed. Five minutes’ further rest on the couch before getting up is also advisable in that it gives the patient a chance to ‘regain normal length’ before compressing the joint by standing. Next, the low back is moved a little, first by flexing/extending each leg slowly in turn, then by tilting the pelvis. If this proves possible without twinges, the patient can get off the bench; this should be done keeping the back straight by rolling on to one side, putting the feet over the edge of the couch and rising sideways to a sitting and then a standing position. Any pain indicates that the patient spend another 5 minutes on the bench before trying to stand up again.

After-care

The patient is then shown how to put shoes on and how to sit in a car (see p. 583). If the lower back still feels stiff on leaving the room, it is better for the patient to take a short walk before getting into a vehicle. A lumbar support – for example, a rolled towel – is often useful to prevent the patient from sagging. In this way, the intradiscal pressure is kept as low as possible. However, during the whole period of treatment, the patient should be instructed to avoid sitting and stooping as much as possible.

Results

In most patients with discodural low back pain and sciatica, in whom manipulations have been insufficient, traction proves effective. The nuclear protrusions recede over the course of 2 or 3 weeks, while the patient is ambulant. First of all, the pain eases. Gradually, straight leg raising returns to full-range. Lastly, trunk movements cease to hurt. When the patient is almost better, a painful arc often appears on straight leg raising.

One study measured the effects of lumbar traction with three different amounts of force (10%, 30% and 60% of body weight) on pain-free straight leg raise test of 10 patients suffering from sciatica caused by discoradicular interactions. The straight leg raise measurements were found to be significantly greater immediately following 30% and 60% of body-weight traction, as compared to 10% of body-weight traction.¹¹⁵

The effect of sustained traction on herniated nuclear material has also been investigated with CT. One study investigated the regression of displaced nuclear tissue during and after traction.¹⁰⁴ Patients were between 20 and 40 years of age. The duration of traction application was 40 minutes, with an applied load of 45 kg. Evaluation of the results showed a regression of herniated nuclear material in 78.5% of median herniations, 66.6% of posterolateral herniations and 57.1% of lateral herniations. Low back and leg pain decreased in all cases, except when the disc was fragmented or calcified. A recent, prospective, randomized, controlled study investigated the effects of continuous lumbar traction in patients with lumbar disc herniation on clinical findings and size of the herniated disc measured by CT.¹¹⁶ They found a significant improvement in symptoms and signs, together with a decrease in size of the herniated disc material as measured by CT.