Print Share

Library

APS Bulletin • Volume 18, Number 2, 2007

Pain Clinic Perspectives

Steven Stanos, DO, Department Editor

New Clinical Practice Guidelines on Low Back Pain

Jane Martinsons, Staff Writer

In the world of pain medicine, where the mere mention of low back pain (LBP) treatments can spur controversy, publishing guidelines on how to evaluate and manage LBP is all but guaranteed to ignite heated debate. Indeed, controversy was sparked last October after American Pain Society (APS) guidelines on evidence-based medicine evaluation and noninterventional management of LBP were published in the Annals of Internal Medicine—and it will undoubtedly happen again later this year with the release of clinical practice guidelines on the use of interventional therapies and surgery for LBP.

“We can appreciate the goals of evidence-based medicine, although we don’t always practice it. How do we find the balance? What is evidence-based medicine?” —Steve Stanos, DO Steve Stanos, DO

“Whenever you mention guidelines, an air of controversy quickly comes up,” said Steven Stanos, DO, who last May moderated a session on these APS guidelines at APS’s 27th Annual Scientific Meeting in Tampa, FL. “We can appreciate the goals of evidence-based medicine, although we don’t always practice it. How do we find the balance? What is evidencebased medicine?”

Stanos said that quality, validity, and size of trials must be evaluated to determine “good evidence.” Fortunately—remarkably, say some of the researchers—a 25-member multidisciplinary team was able to reach consensus on the recommendations. The following is based on the symposium by presenters Roger Chou, MD; Richard Rosenquist, MD, who addressed evidence of interventional procedures; and John Loeser, MD, who covered evidence for surgical interventions. The American College of Physicians collaborated with APS on the guidelines.

Challenges for Physicians

Physicians face several challenges when treating LBP, according to Chou, the first of which is unexplained practice variations in surgery. Rates of surgery vary widely worldwide (the rate in the United States is five times higher than in the United Kingdom), as well as nationwide. In Washington State, for example, rates of surgery vary 10- to 15-fold among counties.

“More surgery is not necessarily bad if people are doing better with it, but we haven’t been able to show that that is the case,” Chou said. “These unexplained practice variations mean there is uncertainty about the best thing to do. Clinicians may not know best practices.” There are a multitude of treatment options available—from pharmacologic and nonpharmacologic therapies to interventional therapies and invasive therapies such as intradiscal electrothermal therapy (IDET), radiofrequency denervation, steroid injections targeting certain areas of the spine, and surgery. “There are at least 40 or 50 interventions, each with its own set of proponents,” Chou noted. “Clinicians and patients are bombarded by advertisements saying how great all these things are, so it’s very confusing for people.”

Chou stressed that many people with acute LBP pain tend to get better over time—a point emphasized by the other presenters as well. “If you don’t have carefully controlled, well-done studies, you actually can be misled pretty easily.”

Second, it is hard to find a precise anatomic diagnosis for LBP. “This is frustrating for patients and clinicians. [Physicians] want to be able to tell people what is causing their back pain, and patients want you to fix it,” which often leads to diagnostic tests and interventions, Chou said. “We don’t have a good way of distinguishing symptomatic from asymptomatic abnormalities on spine films. A lot of people without back pain have degenerated discs and arthritis.”

Third, the effects of interventions are small to moderate—an average 10- to 20-point improvement with most interventions for chronic LBP. “Again, this is very frustrating for patients,” Chou said. “For acute pain, we expect people to get 50% to 90% better; for chronic LBP, you are talking 10% to, at best, 30% average improvements. I think this leads people to use things that are either unproven or not well-proven because they are always looking for something better.”

Methods

To achieve a complete and unbiased review, these researchers used prespecified methods for ranking the evidence, evaluating the balance of benefits and harms, and grading the strength of recommendations. “The evidence review tells you the benefits, harms, costs, and burdens of different interventions, but not necessarily what to do. Somebody needs to make those decisions,” Chou said. “Everything comes with a cost and is associated with a burden. There is no perfect treatment; some harm is always involved. [We were] comparing across 20 or 30 different interventions.”

Extensive peer review involved 40 external experts for peer review, followed by approval by the APS executive committee and the American College of Physicians’ Board of Regents.

Chou stressed that most LBP interventions are not particularly effective. “We defined moderate as 10- to 20-point improvement on a 100-point scale. To some, this is pretty small, even trivial, but our problem was that when we looked at all the interventions, 10 to 20 points was about the best we’d get.” Therefore, most interventions in the guideline received a “B” recommendation, with a ranking of “fair” for quality evidence and “moderate” for net benefit. “Good evidence” required at least two, but preferably three or more, well-done trials with consistent results.

Chou further noted that

Recommendations

  1. Clinicians should conduct a focused history and physical examination to help place patients with LBP into one of three broad categories: nonspecific LBP, back pain potentially associated with radiculopathy or spinal stenosis, or back pain associated with another specific cause. The history should include assessment of psychosocial risk factors, which predict risk for chronic disabling back pain.

    Over 85% of patients who present to primary care have LBP that can’t be attributed to a specific disease or spinal pathology, Chou said. There is no evidence that labeling a specific diagnosis, such as identifying a degenerated disc and informing the patient of it, improves most patients’ outcomes.

    Chou provided a few examples of strong versus weak predictors. For cancer, the strongest predictors are a previous history of cancer or an elevated ESR, whereas the weakest predictors include unexplained weight loss that fails to improve in a month’s time, and being older than age 50. For a herniated disc, a typical history and a positive straight leg raise are predictive. For spinal stenosis, a wide-based gait and a lack of pain when seated are predictive; however, age and neurogenic claudication are weak predictors.

  2. Clinicians should not routinely obtain imaging or other diagnostic tests in patients with nonspecific LBP.

    There is no evidence that routine plain radiography improves patient outcomes, Chou said. “We actually have three randomized trials comparing routine X rays with no routine X rays, and [they] don’t improve pain, function, quality of life, and patient satisfaction. Lumbar X rays actually result in significant radiation.” Routine magnetic resonance imaging (MRI) also hasn’t been shown to improve patient outcomes, but has been linked to increased rates of surgery.

  3. Clinicians should perform diagnostic imaging and testing for patients with LBP when severe or progressive neurologic deficits are present or when serious underlying conditions are suspected.

    This happens when there is a high risk for vertebral compression fracture, such as in an older person with osteoporosis or when there is a suspected infection, cauda equina syndrome, or a progressive or severe neurological deficit. “Cancer is a little bit tricky,” Chou noted. “If age is the only risk factor or if patients have another weaker risk factor, a time-limited trial of therapy may actually be warranted. According to the old Agency for Health Care Policy and Research (AHCPR, 1994) guidelines, you should immediately get an X ray in patients over the age of 50 or with any risk factor for cancer. But in those with only weak predictors, it may be reasonable to do a time-limited trial of therapy before doing an X ray, as long as [patients] don’t have other signs of cancer.”

  4. Clinicians should evaluate patients with persistent LBP and signs or symptoms of radiculopathy, spinal stenosis with MRI, or computed tomography only if they are potential candidates for surgery or epidural steroid injection. Patients with sciatica or spinal stenosis without severe or progressive neurologic deficits do not necessarily need to be imaged, Chou said, emphasizing again that people tend to get better over time or with nonspecific therapies. “Consider what the purpose of imaging is going to be. If somebody is not a candidate for surgery or is not interested in getting other procedures done, in general doing the imaging is not going to be helpful,” Chou said.

    5. “Everything comes with a cost and is associated with a burden. There is no perfect treatment; some harm is always involved.” —Roger Chou, MD Roger Chou, MD

  5. Clinicians should provide patients with LBP evidence-based information about their expected course, advise patients to remain active, and provide information about effective self-care options.

    Chou noted that physicians should explain to patients the importance of self-care and staying active, consider interventions that are shown to be effective by higher-quality evidence, remember that all medications are associated with adverse events and that little evidence currently exists on long-term benefits or harms, and always consider patient preferences, the cost of therapy, whether insurance covers a specific therapy, and the burden of therapy.

  6. Clinicians should consider the use of medications with proven benefits in conjunction with back information and self-care. Clinicians should assess the severity of baseline pain and functional deficits, potential benefits, risks, and relative lack of long-term efficacy and safety data before initiating therapy. For most patients, first-line medical options are acetaminophen or NSAIDs.

    • Acetaminophen or NSAIDs: “These old standbys remain first-line medication options,” Chou said. “They are not the strongest analgesics, but they are safe for appropriately selected patients.”
    • Skeletal muscle relaxants: There is some evidence for acute LBP, though the downside is that they cause a lot of fatigue or somnolence.
    • Tricyclics: “We rated tricyclics small to moderate; it’s not a first-line agent.” If a patient shows signs of depression, a tricyclic antidepressant may be considered.
    • Opioids and benzodiazepines: There is not a great deal of evidence specifically for patients with LBP, and the net benefits appear to be small or moderate, Chou said. “Antiepileptic medication, specifically gabapentin, seems to be beneficial for radiculopathy, but that is based on only a couple of small short-term trials.” Several studies show no benefits to using systemic steroids for LBP.

  7. Nonpharmacologic therapies: For patients who do not improve with self-care options, clinicians should consider the addition of nonpharmacologic therapy with proven benefits for LBP. They also recommend spinal manipulation for acute LBP, and for chronic or subacute LBP, options include intensive interdisciplinary rehabilitation, exercise therapy, acupuncture, massage therapy, spinal manipulation, yoga, cognitive-behavioral therapy, or progressive relaxation.

    This recommendation is weak, Chou said. “The burdens and costs of these therapies, in general, are a lot higher than those associated with self-care and pharmacologic therapy. We felt that this requires more of a discussion with patients about what works best for them. It doesn’t mean we don’t think these things work—we still think clinicians should consider them—but that is the reason for the grade.”

    For acute LBP, the net benefit of spinal manipulation was rated small to moderate and there was no benefit found with exercise therapy used within 4 weeks of onset of symptoms.

    For chronic or subacute LBP, several interventions were rated “moderate net benefit/good level of evidence”: behavioral therapy, or specifically, cognitive behavioral therapy; exercise therapy; interdisciplinary rehab; and spinal manipulation. Acupuncture, massage, and yoga were graded “fair.” Meanwhile, the evidence for many physical modalities was rated poor or inconsistent. These included interferential therapy, lumbar support, shortwave diathermy, transcutaneous electrical nerve stimulation therapy (TENS), and ultrasound. Meanwhile, low level laser therapy, spa therapy, and acupressure received no recommendations due to lack of research in the United States. Chou also noted that current studies on targeting therapies for individual patients are not readily usable in primary care.

Evidence for Interventional Procedures

The use of interventional procedures to treat LBP is clearly on the rise although trials are generally small in size and generate inconsistent results, Rosenquist told the APS Annual Meeting attendees. Concern with the number of procedures being done often follows the money, he said. “The question that comes up is: Are we truly looking at making a difference in outcome or simply doing things because there is revenue attached to them?”

That same question comes into play when physicians conduct an invasive diagnostic test that serves primarily to justify the performance of another invasive therapeutic procedure. Rosenquist points to the steady growth in the rate of lumbar fusion surgery (Weinstein, 2006). “We get better hardware and we do more lumbar fusion surgery,” he stated. Likewise, a study published last year in Spine shows that since 1994, epidural injections increased about 270% and sacroiliac (SI) joint injections rose about 230% (Friedly, Chan, & Deyo, 2007).

Recommendations

Invasive Tests

In patients with chronic nonspecific LBP, provocative discography is not recommended as a procedure for diagnosing discogenic LBP. There is insufficient evidence to evaluate the validity or the utility of diagnostic selective nerve root block, intra-articular facet joint block, medial branch block, or sacroiliac joint block as diagnostic procedures for LBP with or without radiculopathy.

“There is no gold standard for distinguishing between symptomatic and asymptomatic anatomic abnormalities. You look at an image and say, ‘I’m going to block that,’ but you may be blocking an anatomic abnormality that is not the source of the pain.” —Richard Rosenquist, MD Richard Rosenquist, MD

With this recommendation, “we’re touching on holy ground for a whole lot of physicians doing interventional pain medicine. What’s the rationale for doing the procedure, what is the evidence, and what is the outcome? Many of these invasive diagnostic tests are used to select patients with chronic LBP for invasive procedures targeting specific anatomic sources in the back. [But] we don’t always have specific anatomic locations identified,”

Rosenquist said. “There is no gold standard for distinguishing between symptomatic and asymptomatic anatomic abnormalities. You look at an image and say, ‘I’m going to block that,’ but you may be blocking an anatomic abnormality that is not the source of the pain. In many cases, these diagnostic tests have not been shown to improve clinical outcomes compared to clinical assessment and noninvasive testing.”

Problems also arise when physicians begin to look critically at tests such as provocative discography, Rosenquist said, noting that studies have demonstrated that “you can produce a positive result in somebody who has no other anatomic abnormality.”

Interventional Therapies for Nonradicular LBP

In patients with persistent LBP, facet joint corticosteroid injection, prolotherapy, and intradiscal corticosteroid injection are not recommended. There is insufficient evidence to adequately evaluate the benefits of epidural steroid injection, IDET, therapeutic medial branch block, radiofrequency denervation, sacroiliac joint steroid injection, or intrathecal therapy with opioids or other medications for nonspecific LBP or for nonradicular LBP with common degenerative changes.

According to Rosenquist, many interventional therapies have not proven to be effective in placebo-controlled randomized trials. However, certain nonrandomized studies examining the same procedures, such as studies on IDET and facet joint steroid injections, have shown very promising outcomes. Still, he adds, it’s unclear if the interventions are ineffective or if the patients are not accurately selected for procedures targeting specific spinal structures.

Nonspinal injections for nonspecific LBP:

Steroid injection for nonradicular LBP:

Interventional therapies for nonradicular LBP:

Interventional Therapies for Radicular LBP

In patients with persistent radiculopathy due to herniated intervertebral disc, clinicians should discuss risks and benefits of epidural steroid injections as an option. Shared decision making regarding epidural steroid injections should include a specific discussion about inconsistent evidence showing moderate short-term benefits and a lack of long-term benefits.

Radiculopathy and Prolapsed Disc

“Again, if you give [acute LBP] time, things will typically get better. The [problem] is when people don’t want to give it time,” Rosenquist said. “What can you do in the short term to try and get people back to work or to get them to have less pain? If you do an intervention, you may produce some reduction in their symptoms or allow them to go back to work, but these benefits don’t change the long-term outcome.” However, a reduction in pain or an earlier return to normal activity may be a good outcome in the mind of the patient.

Rosenquist suggests that physicians discuss with patients the potential benefits of interventional therapies, making clear that doing so “may buy them time or shorten the time during which we’re letting the natural history of the disease take its course.”

Evidence for Surgical Interventions

Only a very few trials compare surgery to some alternative management, Loeser told attendees of the annual meeting. Study participants included mostly patients who had already failed nonsurgical management and excluded those at higher risk for poor outcomes. “So,” he said, “the trials are stacked.”

“There is absolutely no evidence that increasing use of hardware improves the outcome for the patient.” —John Loeser, MD John Loeser, MD

“In general, the more technically difficult and, therefore, more costly procedures have never been shown to be more effective than less difficult or less costly procedures. For example, fusions now cost around $35,000 apiece. If you put in hardware, it’s going to go up another $6,000 to $10,000. If you go to the cost per quality-of-life year improvement, it’s like $3 million for an instrumented versus noninstrumented fusion. That is clearly much too high a price to pay.”

Recommendations

Surgery for LBP with Degenerative Disc Disease

For patients with nonradicular LBP, common degenerative spinal changes, and persistent and disabling symptoms, clinicians should discuss risks and benefits of surgery as an option. Shared decision-making regarding surgery for nonspecific LBP should include a specific discussion about the small- to moderate-average benefit from surgery versus nonsurgical therapies even in highly selected patients, and the fact that the majority of such patients who undergo surgery do not experience an optimal outcome (defined as minimum or no pain; discontinuation of pain medications; and return of high-level function).

Even with structural pathology, the cause may be unknown, Loeser said. Still, physicians should consider discussing the option of surgery for patients with nonradicular LBP who have common degenerative spinal changes and persistent and disabling symptoms. The result: a weak recommendation with high-quality evidence.

In regard to shared decision-making regarding surgery for nonspecific LBP, “the fact is that the majority of patients who undergo surgery do not experience what we would call ‘an optimal outcome’—minimal or no pain, stopping pain medicines, returning to a high level of function.”

Meanwhile, fusion appears superior to continued standard nonsurgical therapy, but not better than intensive interdisciplinary rehabilitation, Loeser stated. “The benefits of surgery versus nonsurgical therapy are less than 15 points on a 100-point scale,” Loeser said. “Even in highly selected patients, fewer than half experienced an optimal outcome. There is absolutely no evidence that increasing use of hardware improves the outcome for the patient.”

Artificial Disk Replacement

In patients with nonradicular LBP, common degenerative spinal changes, and persistent and disabling symptoms, there is insufficient evidence to adequately evaluate long-term benefits and harms of vertebral disc replacement.

“Studies [on these patients] have only looked at people with isolated one-level disease, which is not a common finding. We have very little information about the use of disc replacement in people with multilevel disease.”

Surgery for Herniated Disc and Spinal Stenosis

In patients with persistent and disabling radiculopathy due to a herniated lumbar disc or persistent and disabling leg pain due to spinal stenosis, clinicians should discuss the risks and benefits of surgery as an option. Shared decision making regarding surgery should include a specific discussion about moderate average benefits in patients who undergo surgery; these benefits do appear to decrease over time.

Surgery for Herniated Disc with Radiculopathy

Discectomy is clearly associated with a more rapid improvement in symptoms than nonsurgical therapy, Loeser said. “But we should not lose sight of the fact that symptoms do abate in patients who don’t have therapy. You don’t have to have an operation to have your radicular symptoms go away, but they will go away faster. Patients improve either with or without surgery. There are very few examples of progressive neurologic deficit in patients who do not have surgery. After a couple of years, the outcomes from surgery and nonsurgical therapy are pretty [similar].”

Most trials are on open discectomy or microdiscectomy, Loeser noted, and there is little evidence on some newer and innovative ways of removing a disc, including percutaneous strategies, laser-assisted strategies, and minimally invasive approaches.

“Regarding surgery for spinal stenosis with or without degenerative spondylolisthesis, decompressive laminectomy is associated with superior outcomes versus nonsurgical therapy. The improvement is relatively small and the benefits tend to fade over time,” Loeser said. He noted that interspinous spacer is too new for comment.

Surgery for Spinal Cord Stimulation for Failed Back Surgery Syndrome

In patients with persistent and disabling radicular pain following discectomy who do not have evidence of a persistently compressed nerve root, clinicians should discuss risks and benefits of spinal cord stimulation as an option. Shared decision making regarding spinal cord stimulation should include a specific discussion about the high rate of complications following spinal cord stimulator placement.

There are no trials for spinal cord stimulation, except for failed back syndrome with persistent radiculopathy. However, it “has been shown to be superior to a repeated surgery trial and to conventional medical management, but the cohort in that [one] study were people who had all failed conventional management. There are a lot of device-related complications; fortunately, very few of them represent significant patient morbidity but not costs and time.”

To address gaps in research, Loeser calls for

References

Agency for Health Care Policy and Research, U.S. Department of Health and Human Services. (1994). Clinical practice guideline: Acute low back problems in adults. Rockville, MD: Author.

Chou, R., Qaseem, A., Snow, V., Casey, D., Cross, J. T., Shekelle, P., et al. (2007). Diagnosis and treatment of low back pain: A joint clinical practice guideline from the American College of Physicians and the American Pain Society. Annals of Internal Medicine, 147(7), 478-491.

Friedly, J., Chan, L., & Deyo, R. (2007). Increases in lumbosacral injections in the Medicare population: 1994 to 2001. Spine, 32(16), 1754–1760.

Weinstein, J. N. (2006). An altruistic approach to clinical trials: The national clinical trials consortium (NCTC). Spine, 31(1), 2707–2714.

Issue Index