Treatment
There is little consensus in practice about how to manage persistent disabling low back pain for which the only structural findings are nonspecific. Some clinicians have focused on the identification and treatment of an occult local "pain generator," assuming there is specific pathology in the spine that accounts for the magnitude of symptoms. However, since the same findings on imaging studies in severely symptomatic patients are commonly seen in minimally symptomatic persons, it has been suggested that psychosocial factors and factors affecting pain tolerance influence the degree of illness in patients with persistent disabling low back pain. In this approach, treatment and prevention are directed at restoring function and supporting adaptive techniques, as opposed to medically or surgically treating the common spinal changes.
Pharmacologic Therapy
Pharmacologic treatment of chronic low back pain usually includes analgesics, antiinflammatory drugs, and muscle relaxants, but the evidence for their efficacy is not compelling. In randomized trials, the differences in pain after a patient has taken nonsteroidal antiinflammatory agents as compared with placebo have generally been in the minimally detectable range.25 For example, in a four-week trial involving patients with a flare of chronic back pain,25,26 pain scores (on a 100-point scale) decreased from 75 to 35 with valdecoxib, and to 45 with placebo. These marginal improvements do not warrant the long-term use of cyclooxygenase-2 inhibitors for patients with chronic back pain, particularly given the new data about increased cardiovascular risk associated with their use.27,28,29 Another short-term trial, with 30 patients, showed that diflunisal (Dolobid, 500 mg twice daily) was more effective in reducing chronic back pain than was acetaminophen (1000 mg four times daily), but interpretation is limited by the small sample and the recognition that there is often spontaneous variation in levels of back pain.25
Muscle relaxants may also alleviate pain only moderately. In a pooled analysis of two randomized trials involving 222 subjects, treatment with tetrazepam (50 mg three times daily for 14 days) resulted in a statistically significant but clinically marginal reduction in pain intensity as compared with placebo.30 Similar results are reported with other classes of muscle relaxants (such as cyclobenzaprine), and no particular class has proved superior. Long-term treatment with narcotics or sedatives is generally discouraged, given the associated risks of tolerance and side effects.25,30
Antidepressant drugs, specifically tricyclic and tetracyclic drugs, have demonstrated small but consistent benefits in pain reduction in randomized trials in patients with chronic low back pain without clinical depression (a 20 to 40 percent greater reduction in pain than with placebo, during a period of four to eight weeks).31 However, there were no consistent or substantial functional improvements, and side effects occurred in more than 20 percent of subjects (Table 1).47 Selective serotonin-reuptake inhibitors and trazodone have not been more effective than placebo in patients with chronic low back pain.31
Nonpharmacologic Therapy
Exercise seems to increase the rate of return to normal activities in patients with persistent low back pain. A Cochrane review of randomized trials of various exercises for persistent low back pain, including strengthening, general stretching, the McKenzie method of passive end-range stretching exercises, and conventional physical therapy (consisting of hot packs, massage, and stretching, flexibility, and coordination exercises), showed that these strategies appeared equivalent and seemed to be more effective than the usual care by a general practitioner.33
In general, exercise programs, such as two or three one-hour sessions a week until normal activities are resumed, in four to six weeks, have moderate effects — 10 to 15 points on a 100-point pain scale, or a 5 to 10 percent improvement on scales that assess functional disability, as compared with placebo or usual care.33,36 Similarly, randomized trials and systematic reviews have not shown a clear advantage of any particular treatment method over another, including physical therapy, exercise, massage, manipulation by chiropractors or other practitioners of manual medicine, low-impact aerobics, reconditioning on training machines, or back school (classroom-style educational programs for patients with back pain) (Table 1).32,34,48,49
Available data suggest that a combination of medical care with either physical therapy49 or manipulation35 may be moderately more effective in reducing pain and self-rated disability than is a single method of treatment. The difference may reflect the patient's confidence in the treatment prescribed. In a trial comparing chiropractic care with medical care with and without physical therapy, the patient's initial confidence in the assigned treatment correlated directly with outcome, whereas the treatment assignment per se did not.49,50 At the other extreme, patients with persistent pain should avoid rest or confinement to bed.51
Given the marginal effect on functional outcomes of most of these interventions when used alone, and given the evidence that psychosocial factors may be important obstacles to recovery, more comprehensive approaches have been developed. Functional restoration programs incorporate physical therapy and medical treatment strategies with a cognitive behavioral approach that focuses on achieving specific functional goals (e.g., certain walking distances and speeds or certain weights lifted and numbers of repetitions). These programs, as compared with usual care by a general practitioner, also seem to decrease the amount of sick leave taken.37,38,44,45,52
However, none of these methods of rehabilitation have consistently been shown to have generalized applicability (partly because of compliance issues in distressed patients), and it is unknown if effects are sustained for the long term.28 Neither less intensive rehabilitation programs, especially those not accompanied by a strong component of behavioral therapy, nor pain-management programs relying on spinal injections and analgesic drugs seem to offer clear advantages over usual care for improving functional outcomes.