Low back pain is a widespread and costly complaint that affects virtually all Americans at some point in their lives. After the common cold, it is the second most frequent cause of work absence in the United States.
Low back pain is a widespread and costly complaint that affects virtually all Americans at some point in their lives. After the common cold, it is the second most frequent cause of work absence in the United States. Table 1 provides a useful classification system for low back pain.
Although physicians often consider low back pain to be a frustrating problem, I enjoy the challenge it offers and I see many patients with this complaint. In this article, I address 12 questions I am frequently asked by colleagues.
1. When should I worry that back pain is caused by an ominous process, such as infection or tumor?
Always worry! Otherwise, you will never think of these important diagnoses. Fortunately, tumors and infections are rare and evidence-based guidelines for identifying "red flags" are available (Table 2).1,2 Save your workups for patients who appear to be at greater risk for an ominous process.
The risk factors for spinal malignancy include known malignancy elsewhere (eg, breast or lung cancer), age greater than 50 years, intractable pain and difficulty in finding a comfortable position, and elevated erythrocyte sedimentation rate (ESR). The presence of one or more of these factors should raise your suspicion of cancer and lower your hesitancy about obtaining radiographs and imaging tests.
The risk factors for spinal infection (eg, osteomyelitis) include known infection elsewhere (particularly bacteremia), injection drug use, systemic signs and symptoms (including fever), and laboratory evidence of infection (including elevated white blood cell [WBC] count and elevated ESR).3
2. When should I obtain an x-ray or imaging study?
Plain radiographs and imaging studies are unnecessary during the initial evaluation of acute low back pain, unless you have noted red flags that suggest a more ominous source of pain.4 If the patient's condition fails to improve after a couple of months of conservative therapy, it is reasonable to obtain a plain radiograph to rule out structural abnormalities, such as spondylolysis or spondylolisthesis. If radiographs have already been obtained for the patient, it is generally not necessary to repeat them unless the pain differs substantially from that of previous episodes or intervening trauma has occurred.
In the absence of red flags, studies such as CT or MRI should be ordered only to provide a road map for anatomic interventions, such as surgery or epidural corticosteroid injections. If both you and your patient are uncomfortable with your diagnosis, it is generally more fruitful (and far less expensive) to make a referral to a spine specialist than it is to obtain an imaging study.
The problem is that imaging studies are nonspecific; disk bulges are seen in more than half of asymptomatic persons, and disk protrusion in more than one third.5 Spinal stenosis is seen in more than 20% of older patients.6 Hence, imaging studies should be used to confirm--but not to fish for--a diagnosis.
3. When an imaging study is appropriate, should I order a CT or MRI scan?
Your choice depends on clinical features as well as institutional factors (eg, cost and access). Both of these modalities are generally sensitive, although somewhat nonspecific. MRI provides better visualization of the disk, while CT provides better visualization of bony details. Furthermore, CT is considerably less expensive. My first choice of a spinal imaging test is usually MRI, particularly if there is any suspicion of tumor, infection, or disk pathology. I use CT for patients with contraindications to MRI, such as claustrophobia or metal implant devices, or in cases in which bony detail is critical.
4. Should patients with low back pain stay in bed? If so, for how long? When should they return to work, and should their activity be restricted?
Results of controlled trials have shown that bed rest does not improve clinical outcomes but may prolong the period of disability.7 I discourage patients from bed rest.
The issue of returning to work is challenging.8 Prolonged work absence is associated with low subsequent rates of return to work9 as well as social role disruption and, often, depression. This suggests that patients should return to usual activities as quickly as possible.10,11 Be careful not to "prescribe disability" by arbitrarily limiting the activities of patients with low back pain.
It is useful if employers can provide flexible hours and limited duty to reduce mechanical strain on the patient's back during the recovery phase. When indicated, modification of workplace furniture and equipment (eg, adjusting the height of desks, chairs, or file cabinets) is also helpful.12
I work with patients to maintain their social roles. I generally suggest that patients return to work quickly, and I contact their employers to suggest limited duty (eg, lighter tasks or fewer hours) for days to weeks to reduce the risk of reinjury. These decisions, of course, must be individualized.
5. My patient wants to see a chiropractor. What should I say?
This depends on the specific diagnosis. For patients with acute, nonspecific low back pain, several rigorously performed, randomized controlled trials support the use of chiropractic manipulation.13 However, the prognosis of low back pain in these patients is excellent. The role of chiropractic in chronic low back pain, herniated disk, and spinal stenosis has not been evaluated rigorously.
6. What signs and symptoms point to a herniated nucleus pulposus?
Because imaging studies for the evaluation of potential herniated disk are nonspecific, you must make a clinical diagnosis based on the history and examination (Table 3).14
The most important historical clue is the presence of sciatica; the pain radiates from the back into the buttock and thigh and typically into the lower leg, along the L5 or S1 dermatomes. Provocation of pain with the Valsalva maneuver is noted frequently. The straight leg raising test on the ipsilateral side generally elicits pain that reaches the knee (a positive sign) in 80% of cases, with a sensitivity of 80%. Specificity is modest (about 40%).
The cross-straight leg raising test has a much higher specificity (greater than 80%) but is insensitive. Strength, reflex, or sensibility deficits occur in about 50% of patients; the specificity is about 80%.14
7. What are the clinical features of degenerative lumbar spinal stenosis?
More than 20% of asymptomatic persons older than 60 years have radiographic evidence of lumbar spinal stenosis.5 For this reason, the diagnosis must be based on the history and physical examination.
In lumbar spinal stenosis, lumbar flexion increases the cross-sectional area of the spine, while extension decreases it. Thus, characteristic historical findings include pain with prolonged standing or walking and relief with sitting.15,16 Patients often report that they tend to lean on the shopping cart and avoid social gatherings at which they have to stand. They can, however, remain comfortably seated while watching movies. I am aware of a neurosurgeon with spinal stenosis who bicycles 10 km to work but can no longer perform operations.
Approximately 50% of patients with lumbar spinal stenosis have deficits in vibratory sensibility, temperature sensitivity, or muscle strength. The specificity of each of these findings is about 80%. The nerve root involved is most commonly L5, followed by S1 and L4. Ankle reflexes are frequently abnormal in older patients and therefore are of little value.16
About 70% of patients with spinal stenosis complain of balance disturbance. Half have a wide-based gait or exhibit Romberg sign (swaying or falling when they stand with feet together and eyes closed), the specificity of which is about 90%. These patients have normal cerebellar signs; the pseudocerebellar presentation presumably results from the involvement of proprioceptive fibers. Another useful physical sign in this population is pain that is perceived in the back, buttock, and thigh and elicited by prolonged lumbar extension.16
Thus, spinal stenosis is characterized by 3 clusters of signs and symptoms:
8. When should I refer patients with back pain to a physical therapist?
Patients who are having a bout of acute low back pain do not benefit from exercise.17-20 They may benefit from passive measures, such as heat, but they can apply these at home without seeing a physical therapist. Exercise, supervised initially by a physical therapist, helps prevent subsequent episodes of acute low back pain by improving strength, flexibility, and conditioning.
Patients with back pain that lasts more than 1 month typically begin to experience limited motion and weakness. In this setting, a referral to a physical therapist can be extremely valuable. Exercise is crucial for patients with chronic back pain, who are usually deconditioned.17-20 A physical therapist can plan and supervise an exercise program.
9. What is the role of epidural corticosteroids in the management of spinal disorders?
The evidence is limited. A few randomized studies were done in persons with herniated disks, but these investigations had methodologic limitations. No randomized study has been conducted expressly in patients with spinal stenosis. Trial results suggest that injections may help in the first 6 to 12 months after surgery.21
Observational studies suggest that epidural corticosteroid injections are useful for patients with pain related to nerve root impingement by herniated disks or spinal stenosis.22,23 The benefit is generally transient-lasting weeks to months-and complications are rare. Leg pain is much more responsive to epidural corticosteroid injections than is back pain.
Adequate randomized controlled trials are urgently needed in this area. In the meantime, I present epidural corticosteroid injection to patients as a reasonable option that can buy a few months of relief.
10. What are the indications for and outcomes of surgery for herniated disk?
The only urgent indications for surgery are cauda equina syndrome and a rapidly progressive neurologic deficit, both of which are rare. For elective surgery, the risks of the operation must be balanced against the benefits of more rapid improvement. In more than 90% of those patients who have herniated disk syndromes, symptoms diminish over the course of several months without surgery. Surgical decisions should be made with this favorable natural history in mind.
When there is congruence among symptom location, neurologic deficit, and disk herniation location on imaging studies, along with a positive straight leg raising test result, surgery generally has an excellent outcome. Patients who are receiving workers' compensation, especially if litigation is pending, have less favorable outcomes.24 Among patients who have all of the positive prognostic factors and none of the negative ones, the likelihood of successful surgery exceeds 90%.25,26
Disk surgery has complications, although they are uncommon.27 Operative mortality is approximately 0.15%. Discectomy is associated with nerve injury in about 1% of patients, deep venous thrombosis in 1% to 2%, and pulmonary embolism in fewer than 1%. Wound infections occur in about 1% to 2% of patients. About 10% require repeated operations, a high rate compared with other surgeries, hence the expression "failed back surgery syndrome."
Although surgery relieves symptoms more effectively than does nonoperative therapy, 4 years after treatment (whether operative or nonoperative) all patients have the same level of symptoms. This finding suggests gradual resolution of disk-related symptoms among patients who did not have surgery and recurrence among those who did.28
The efficacy of discectomy and decompression for spinal stenosis will be clarified considerably over the next several years as the results of the Spine Patient Outcome Research Trial (SPORT) become available. SPORT consists of 3 studies of the role of surgery in disk protrusion, spinal stenosis, and degenerative spondylolisthesis.29
11. What are the indications for and outcomes of surgery in spinal stenosis?
The natural history of spinal stenosis is considerably worse than that of sciatica caused by a herniated disk: only about 20% of patients experience substantial improvement over time.30 This would provide less rationale for temporizing. On the other hand, patients with spinal stenosis are older and more often have comorbid conditions that increase the likelihood of complications. Surgical outcomes are more favorable when leg pain, rather than back pain, is predominant, and less favorable when the patient has worse functional status and a greater number of comorbid conditions.31,32
Observational studies suggest that surgery for spinal stenosis is successful in about 65% to 75% of patients.33 The rate of reoperation ranges from 10% to more than 20% in various series.
Because spinal stenosis has a poor natural history, laminectomy should be considered carefully for patients who find their symptoms unacceptable and have functional limitation. There has been considerable debate about whether laminectomy should be supplemented with lumbar fusion, a topic that is beyond the scope of this review. Nevertheless, it is important to recognize that fusion increases the likelihood of complications and the need for transfusion, and it lengthens the period of rehabilitation.
12. How do I help patients make the best decision regarding surgery for low back pain?
Our goal is to help patients make decisions that are fully informed and consonant with their preferences. By "fully informed," we mean that patients understand the risks and benefits of the choices they face. By "consonant with preferences," we mean that patients' decisions are consistent with their values. For example, patients who are more averse to taking risks will tend to opt for less risky treatments. Those who place great emphasis on functioning at a high level may be willing to consider a riskier treatment that can improve function.
Decision-making tools have been developed to make this process easier. In one randomized controlled trial, patients considering disk surgery who received the decision-aid intervention were less likely to opt for surgery than those who received a control intervention. In contrast, among patients considering surgery for spinal stenosis, those who received the decision aid were somewhat more likely to opt for surgery.34 These findings suggest that informed, preference-consonant decisions may differ substantively from usual clinical decisions.