R.W. Porter. Royal College of Surgeons Edinburgh, Nicolson Street, Edinburgh EH8 9DW.

More than 20 00 spinal operations are carried out in the UK each year. The decision when and whether to operate requires mature judgement. Spinal surgery is technically difficult, demanding a high level of surgical skill. It is learnt only by lengthy apprenticeship. The after-care is equally important. The personal supervision of the surgeon who leads a coordinated team of clinicians, nurses and physiotherapists will ensure the best results. It is inevitable and unfortunate that mistakes will occasionally be made and only careful attention to detail in the pre-operative assessment, meticulous surgical care and supervised post-operative management will ensure consistently good results. The spinal surgeon needs to remain up-to-date, be disciplined with a systematic and careful approach and lead a coordinated team to maintain the highest standards.

Pre-operative management.

What is a patient entitled to expect at the present time and what are the standards that are recognized by the general body of spinal surgeons today?

Avoid unreasonable haste.

Spinal surgery offers the patient a life-threatening procedure for a non-life threatening condition. There are always potential surgical complications, and although the benefits of surgery can be considerable, all reasonable conservative approaches should have been attempted first. When to operate is a difficult decision demanding fine judgement. There is a spectrum of opinion amongst surgeons for most spinal conditions. Some prefer a radical approach offering early surgery in order to hasten recovery of normal function. Others adopt a more conservative attitude. In the long term, most spinal conditions have a good natural history but disability can be protracted with conservative management. For example, there is no absolute indication for a surgical discectomy unless the patient has bladder symptoms. Even with severe sciatica and marked root tension signs, surgical treatment is not mandatory. With time symptoms will resolve, perhaps not completely, but at least to a manageable level. If symptoms have been present for only two weeks, it is not possible to predict the natural resolution. Without surgery, many of these patients will recover over a 3-month period with complete relief of symptoms and a return to work. If there has been no improvement over a 6-week period, however, the prognosis is less optimistic. Most spine surgeons would argue therefore that disc surgery for a patient who has had less than 6-weeks of symptoms is not good practice. They would say that unless there were bladder symptoms it was operating with unreasonable haste to offer surgery to a patient with less than 2-weeks of symptoms and this would not be a practice adopted by any reasonable spine surgeon.

Avoid unacceptable delay.

Patients are sometimes distressed because they believe that there has been unnecessary delay in referral to a spinal unit. They may have suffered for many years with chronic back pain until finally they find an expert who is able to offer them successful spine surgery. They may remember years of pain and suffering, and at some time having been told the problem is psychological, and they might seek redress. Such a claim, however is unlikely to succeed because the surgery for chronic back pain is very unpredictable. There is always a body of reputable clinicians who would not recommend surgery for chronic low back pain. It is reasonable to operate but it is also reasonable to encourage patients to avoid surgery.

A more frequent claim for potential medical negligence is an unacceptable delay in a patient who has a bladder problem. Spinal pathology which compromises the sacral nerves supplying the bladder generally requires emergency treatment. Every surgeon knows that a patient who is unable to pass urine or alternatively is incontinent of urine has a potentially serious condition. Involvement of the L5 nerve may leave a patient with a foot drop and some numbness of the foot. It is a disability, but not serious even if it does not recover. However, involvement of the sacral nerves which supply the bladder is a major problem. When the symptoms have been there for more than a few hours the problem tends to be permanent, with loss of normal bladder function for a lifetime. The same nerves supply the bowel and sexual function, which are also permanently affected.

Pathology which affects the conus can cause these problems. The conus is the lower end of the spinal cord at the level of the second lumbar vertebra and this can be compressed by fractures at this site. The conus may be compressed mechanically or its blood supply may be affected by the fracture. The sacral nerves can also be affected in the mid-or lower-lumbar spine. These nerves form the lower posterior part of the caudia equina. They lie fairly close to the lamina and they can be damaged by decompressive surgery at this level. A massive disc herniation can so compress the cauda equina that the sacral roots are pressed tightly against the lamina (cauda equina lesion).

It is therefore an emergency situation when a patient presents with back pain and bladder symptoms. These patients require rapid and comprehensive neurological assessment. The sacral nerves also supply the skin around the anus and perineal region. They are responsible for the tone of the anal sphincter muscles. Clinical assessment of this region confirms that there is significant problem with the sacral roots and early surgery is imperative.

The patient with the massive disc protrusion and bladder problems requires urgent imaging with MRI and rapid removal of the herniated disc, decompressing the cauda equina. The patient who has had spine surgery and in the post-operative period has evidence of sacral nerve dysfunction also requires urgent investigation and re-exploration for a treatable condition. The patient with an upper lumbar fracture and conus lesion similarly requires urgent decompression after appropriate imaging.

It is agreed that the longer the sacral nerve roots are compressed the less likely is their recovery and therefore, for the best result early surgery is required. It takes only a few minutes of complete loss of blood supply to the nerve roots for permanent damage to occur. Thus if compression is complete, in practice early surgery is generally too late. However the majority of patients have incomplete compression, when surgery within the first few hours can be fairly effective. The longer the delay, the less chance there is of recovery.

Adequate pre-operative investigation.

The diagnosis of back pain syndromes is made from the history and the examination. Investigations are supplementary, usually to identify the level at which surgery will be required. A plain radiograph is helpful in a negative sense. It can exclude advanced ankylosing spondylitis where the sacroiliac joints are sclerosed. It can demonstrate a spondylolisthesis which may be a hidden cause for back pain. It can also show up spinal metastasis from primary tumors elsewhere when these have progressed to an advanced stage. It will also show the presence of an osteoporotic fracture. However, for the majority of patients with low back pain syndromes the plain radiograph is negative. There is no good correlation between degenerative change in the lumbar spine and back pain.

The image of choice is the MRI scan. This will identify a disc protrusion, degenerative discs, spinal stenosis, infection and neurological tumors. It is so sensitive that pathology is often demonstrated which is asymptomatic. It is only of positive value when the images correlate with the clinical features in the history and examination.

When MRI is unavailable, a CT scan is the next best imaging modality, followed by myelography. It is unacceptable to operate on a patient without one of these supplementary investigations. Spinal surgery should be a once-in-a-lifetime experience. It is not an exploratory procedure in order to make a diagnosis; rather, there should be as much information available to the surgeon as possible prior to the operation.

There are no reliable procedures which tell us where the pain is coming from. Some clinicians will carry out facet joint injections or a provocative discography in the hope that this will identify the pain source by reproducing the pain. However none of these are totally reliable.

Depending on the patient's clinical problem, other investigations may be necessary. For example, a patient with osteoporosis requires blood investigations to exclude other sources of demineralization. They need blood tests for osteomalacia and hyperparathyroidism, tests of the liver function and renal studies. However, for most of the patients having spine surgery, the history and examination often followed by MRI scan is the main pre-operative assessment. It may sometimes be supplemented by other studies.

Operate in the patient's best interest.

It may seem obvious that surgery should be offered only when it is in the patient's best interest. However, it is unusual to have an absolute indication for surgery on the spine. The natural history of spinal disorders without surgery is generally good although disability can be protracted. It is necessary therefore for the surgeon to have a complete understanding of the patient's past history and social history to understand their lifestyle and occupational requirements before recommending surgery. It is important to understand the patient's expectations and whether these are realistic, and to ask whether surgery should be carried out at all. Simply because it is possible to change pathology by a surgical approach, does not mean that it is indicated. It is justified, however, in the patient who requires a rapid return to normal function because of the family or occupational situation.

Operative management and competent surgery.

Competent surgery.

At the present time, there are few full-time spine surgeons in the UK. Most of the spine surgery is carried out by orthopaedic surgeons or neurosurgeons who besides having a general practice take an interest in the spine. In previous years, most orthopaedic and neurosurgeons would do an occasional spine operation but it is now becoming an acceptable practice for spine surgery to be the preserve of those who have a special interest. This means that the majority of spinal surgery for degenerative conditions is carried out by a surgeon doing at least 20 spine operations per year. There is no agreed minimum limit but competence is maintained by continued surgical practice.

There are occasions when a surgeon can predict technical difficulties. For example, obesity will add to the operative difficulties and repeat spinal surgery is more complicated than the first procedure. It can be anticipated that a grossly displaced spondylolisthesis will be difficult to fuse, and extensive spinal stenosis with gross degenerative change is not easy to decompress. With such problems a good surgeon will refer to one of his more experienced colleagues. Failure to do so, and particularly failure to inform the patient of potential hazards, is poor practice.

Correct level

One of the most common causes for medical negligence is operating at the wrong level. Most patients have five lumbar vertebrae followed by the solid segments of the sacrum. About 5% of the population will have six lumbar vertebrae where one of the sacral segments is lumbarized or four lumbar vertebrae where the lowest lumbar segment is sacralized. This can confuse the surgeon. The radiographs may show the disc protrusion say at L4/5 level. If the fifth lumbar vertebra is sacralizred and fixed to the sacrum, the surgeon may be confused and operate at the L3/4 disc by mistake. It is imperative therefore that the surgeon adopts a safe practice to identify the operative level. There are two methods. One is to expose the sacrum in the operative field. It is possible to identify the solid sacrum by vision and by palpation and the work up the spine identifying each of the lower levels. When using a minimal exposure, the correct level can be identified radiographically on the operating table by introducing a needle down to the appropriate segment and checking the level on the radiograph. Sometime, methelene blue dye can be injected through the needle and this mark recognized when the area is exposed. There really is no defence for operating at the wrong level.

Correct side.

Human error is responsible for operating on the wrong side. It is not defensible, but the surgeon can mistake the laterality when the patient is lying prone. Good practice requires a skin marker on the back identifying the side required for surgery.

Sufficient surgery.

There has been long debate how much disc material should be removed in a patient who has a disc protrusion. A consensus is developing that only the loose fragmented material needs to be removed, along with any other loose fragments within the disc space. The surgeon is operating through a deep hole and cannot visualize the centre of the disc space. Loose fragments are extracted by rongeurs and sometimes by a blunt curette. Previously, surgeons would remove large amounts of disc material from within the disc space to avoid a recurrence, but this is now considered unnecessary. It is a balance of clinical judgement as to how much or how little material should be removed and there are as yet no absolute guide lines.

When carrying out a discectomy the symptomatic lesion is usually at one level. Not uncommonly, imaging will show protrusion at perhaps two levels, and there is then a dilemma about which one is symptomatic. The surgeon usually operates at the level which is compatible with the clinical features of the nerve root involved. It is good practice to limit the surgery to as little as possible compatible with the clinical features.

Spinal decompression for spinal stenosis requires removal of the tight bony lamina to allow more room for the underlying nerves of the cauda equina. In previous years surgeons were very radical, sometimes removing the lamina of all five lumbar vertebrae. In the last few years, surgery has become much more conservative, frequently recommending only partial laminectomy at a selective level where stenosis is most significant. If decompressive surgery is too extensive, it runs the risk of the development of post-operative scar tissue causing further stenosis and also the risk of instability and post-operative back pain. However, if the decompression is too limited, the nerve roots may not be adequately decompressed. In addition bony ridges can develop post-operatively, tightening up the canal again.

Damage to nerves.

Disc surgery is surgery for the nerve root, and decompessive surgery removes bony structures where spinal stenosis is causing nerve problems. These nerves are already vulnerable, having been partially compressed, and they need to be handled with great care. Even gentle retraction of the nerves in order to expose a disc or to remove tight bone can further affect nerve root function. Even when these nerves are satisfactorily decompressed the patient may be left with some abnormal nerve function. This can occur even in the best hands and is unavoidable. There are, however, occasions where nerves are crushed or bruised by surgical instruments and sometimes a nerve can be severed. This is the result of poor surgical technique and is generally not to be expected from a competent spine surgeon. There are occasions, however, when an experienced surgeon is attempting to decompress a tight spinal canal, the nerves cannot be fully visualized and unavoidable contusion can occur. The surgeon may or may not be aware of this injury at the time.

The most popular method of performing a spinal fusion is to support the bone graft by using pedicle screws. These screws are inserted posteriorly through the pedicle of the vertebrae into the vertebral body. This is done blindly with an understanding of the direction of the pedicles. Studies have shown that even in expert hands, 20% of the pedicle screws transgress the pedicle to some degree. Frequently this is not clinically important but from time to time there may be a serious transgression of the pedicle and the screw can damage one of the nerve roots. This occurs when the anatomy of the spine is slightly abnormal and the surgeon will not usually be aware that the screw is not within the pedicle. It is only when the patient recovers from the anaesthetic complaining of leg pain that the surgeon is aware of this complication.

Radiographs do not usually help to confirm whether or not the screw has transgressed the pedicle because of so many overlapping shadows. CT scans and MRI are similarly unhelpful because the metal scatters images and impairs good definition. Nerve root damage as a result of a pedicle screw is a clinical diagnosis which can sometimes be difficult. Nerve root pain can be present post-operatively because of mechanical disturbance to a nerve root during an associated decompression. The spinal mechanics may have been slightly altered, compressing the nerve root, and it is sometimes difficult to be confident that a pedicle screw is responsible. If a screw has completely transfixed a root then surgical removal of the screw is not likely to significantly affect the symptoms. If, however, the nerve root is being irritated or contused by the screw, the removal will be beneficial. The surgeon tends to re-operate because of a high level of suspicion and sometimes is rewarded with a relief of the symptoms. On other occasions permanent damage will have resulted from this misfortune. Although the surgeon can not reasonably be blamed for this problem, the patient should be warned pre-operatively that there is a slight risk of nerve damage when pedicle screws are to be inserted.

Damaged Dura.

The dural membrane surrounds the cauda equina and the nerve roots bathed in cerebrospinal fluid. There are two layers of the dura. If both layers are cut or torn cerebrospinal fluid will leak into the wound. The extradural veins are usually compressed by the tight dura and when the cerebrospinal fluid leaks, the dural pressure falls and the veins become congested. The operation then becomes difficult because of cerebrospinal fluid and venous blood filling the wound. Damage to the dura occurs in about 5% of lumbar spine operation and it is considered an accepted complication that is sometimes difficult to avoid. Some surgeons then recommend a wider exposure followed by suturing the dura and others recommend that the dura not be repaired. These surgeons rely rather on a secure muscular repair to avoid post-operative leak.

If the dura is contused and torn, the nerve roots which lie posteriorly in the cauda equina are the sacral roots which supply the bladder and bowel and sexual functions. The surgeon is therefore particularly cautious when decompressing at the back of the spinal canal. Rough handling of the tissues at this level is of course poor practice, but the surgery can be difficult in the obese patient with marked degenerative change.

A few patients who have a dural leak at the time of operation continue to have a discharging sinus of cerebrospinal fluid. In the majority of these the sinus will become dry over a few days, while on other occasions it will produce a chronic leaking sinus which requires further surgery. Occasionally a pseudo-menigocele will form with a large cyst of cerebrospinal fluid, which requires surgical closure.

Damage to blood vessels.

Most patients have minimal blood loss during spine surgery and do not require a blood transfusion. Extensive decompression and a spinal fusion using a bone graft from the pelvis will, however, frequently cause sufficient blood loss to require a blood transfusion.

It is only when a major vessel is damaged that serious complications occur. The superior gluteal artery leaves the pelvis into the buttock through the lower part of the pelvis (the greater sciatic foramen). When taking a bone graft from the back of the pelvis it is possible to damage this vessel and the surgeon will therefore avoid this particular region. Bleeding from the superior gluteal artery is not the result of poor practice, but failure to ligate the vessel is not acceptable. Sometimes the help of a vascular surgeon is required to identify and secure the bleeding vessel.

It is possible to damage the aorta or the inferior vena cava when carrying out a discectomy, if the disc extractor penetrates through the anterior annulus of the disc. If the surgeon relies on removing only the loose fragment this complication is not likely to occur. However, if there is a radical excision of the disc space it is possible to penetrate through the anterior annulus. This may already be torn pathologically.

The surgeon is therefore particularly cautious when operating towards the front of the disc space. It is difficult to defend this injury. The surgeon is aware that there has been some damage to a blood vessel anterior to the disc space when there is a small amount of blood on the instruments. The disc space is avascular and the instruments should be dry. The anaesthetist may note that there is a drop in blood pressure and an increase in the pulse rate. There should be no delay in turning the patient into the supine position and with the help of a vascular surgeon exposing the damaged vessels. Tragedies occur when the surgeon procrastinates and hopes the injury is minimal. It is then frequently too late to avoid a fatal outcome.

Infection.

Infection can occur from time to time. This can be an airborne infection or bacteria transmitted from instrument. The surgeon's gloves can perforate with infection being transmitted from the surgeon's hands. Sometimes bacteria in the patients own blood (bacteraemia) can settle in the wound, producing the infection. These are occasional hazards which are difficult to avoid. Some surgeons recommend perioperative prophylactic antibiotics, but this is not routine practice.

Post-operative management.

The surgeon and the surgical team need to be vigilant in the post-operative period. As soon as the patient is awake from the anaesthetic it is important to carry out a neurological assessment, particularly to confirm that there is not abnormal neurology which was not present before surgery. If there is an area of anaesthesia in the lower leg and foot or some weakness of the lower leg, this suggests some nerve root damage by surgery. If the patient has severe pain in the root distribution which was not present before surgery this is again evidence of some nerve root damage. A complete transgression of the nerve root is not usually painful but will give some motor weakness and sensory loss. Root pain, however, suggests some degree of nerve compression or irritation.

Sacral anaesthesia and some loss of anal tone suggests sacral nerve root damage and a possible bladder disfunction. This is a serious sign which usually requires urgent repeat surgery. If there is sacral nerve root damage there may be a haematoma pressing on the nerve roots which can be relieved by decompression. However, if the sacral roots have been contused by the surgery, further exploration will not help.

In the next few days, if the patient continues with severe root pain or root pain which was not previously present, it is worth re-exploring the spine after suitable imaging. It is possible that the surgery has been carried out at the wrong level or the wrong side or that the decompression, although at the correct site, has not been adequate. Further surgery is likely to be helpful. If, however, nerve root symptoms develop some days after surgery, after a pain-free interval, this suggests that there has been a fragment of disc material previously missed which has now extruded and is pressing on a nerve root. It is no fault of the surgeon. Further surgery for this fragment after suitable imaging can be helpful. However, depending on the severity of the symptoms, the pain may be left to resolve naturally.

If there is a leak of cerebrospinal fluid after surgery the patient should remain in bed receiving appropriate antibiotics and routine dressings until the wound becomes dry. If after a few days the wound is continuing to discharge cerebrospinal fluid, exploration may be appropriate.

Post-operative infection can be the result of a chest infection or urinary infection. If these can be excluded then the infection of the wound is the most likely source. There should be no delay in identifying the source and giving appropriate antibiotics. If there is a high temperature a blood culture is indicated. Post-operative discitis is an infection in the disc space. It is associated with severe low back pain and usually spasm of the spinal muscles. The patient is in severe pain when attempting to stand. The radiographs are normal for a few weeks but the MRI scan is very sensitive to discitis and is the image of choice. A blood culture or needle biopsy will identify the organism and its sensitivity.

Repeat surgery.

It is a difficult decision to know whether or not to operate again in the post-operative period. The complications of the second operation are greater than the first, but provided the indications are correct there can be a considerable bonus in performing this procedure. The inexperienced spine surgeon should seek a second opinion before embarking upon repeat surgery.

Informed consent.

One of the most common causes of patient dissatisfaction is failure to receive sufficient information about surgery and its risks. At the present time, spine surgeons would agree that every patient should be given information about risks and benefits. In broad terms patients should be informed of the changes of improvement by surgery. For example, when performing a discectomy for nerve root pain there is something like a 90% chance of relief of leg pain by surgery. When decompressing the spine for neurogenic claudication there is a 60% chance of reducing the symptoms and perhaps a 60% chance of relieving chronic back pain by a spinal fusion.

The patient should also be told in general terms, the risk of not being improved by surgery. For example, in discectomy there is a 10% chance that some leg pain will persist in the short term. The nerve has been bruised for a long time and even though the fragment of disc has been removed the nerve can remain sensitive and painful. When offering fusion for low back pain, 40% may continue with their pain.

There is always a remote risk that the patient could be worse. They should be told of this possibility. In disc surgery 2 or 3% can be worse as a result of damage to a nerve root, a leak from the dura or some post-operative infection, and there is always the chance of some anaesthetic complication. Patients are not usually told about these remote risks in great detail, or why they could be worse by surgery. If, however, they specifically ask how they could be worse, then it is the surgeon's responsibility to explain some of these problems. When carrying out decompression for neurogenic claudication or a fusion for chronic back pain, patients are told that not only may their symptoms not be relieved but they could have more pain and over a period of time things could get steadily worse. There is about a 5% chance of having repeat surgery after discectomy and 10 to 20% chance of having further surgery after decompression for neurogenic claudication. There is a similar risk of repeat fusion for chronic low-back pain. The broad concept of improvement, failure to relieve symptoms and the change of being worse should be explained to every patient and also recorded.

In spite of these many pitfalls most patients do well. Fortunately those patients who have careful pre-operative selection, competent surgery and good post-operative management are significantly helped by their operation, and provide they are aware of potential risks, patients usually accept that the surgeon and the surgical team have done their best in a very difficult area of medical care.

Taken from the Journal of the Royal College of Surgeons Edinburgh. 1997; 42: 376-380.