Should Physical Therapy for Low Back Pain Focus on Pain Relief or Returning to Work?

Special Report

Theodore R. Fields, MD, FACP

Attending Physician, Hospital for Special Surgery
Professor of Clinical Medicine, Weill Cornell Medical College

Low back pain has great financial implications for our healthcare system. Lost work time, moving to "light duty" and medical service utilization are major issues, and the overall burden to society is very high. Treatments are often less than satisfactory. For back pain which threatens patients' occupational function, there has been a controversy about whether physical therapy should focus on pain relief or on rapidly getting patients back to work.

This controversy has led to mixed messages being given to patients. Physical therapists will often tell them to "work through the pain." Yet, therapists repeatedly return to the "1 to 10" scale of pain, if for no other reason than that the JCAHO mandates it. While therapists do not dismiss the importance of pain, many believe that an excessive focus on pain slows patients' progress.

Staal et al[1] recently reported in an approach to this issue in the Annals of Internal Medicine. They noted that a prior study of Swedish auto workers[2] showed less absence from work with a graded exercise program than with "usual care." The importance of the graded exercise program is that it attempted to progress the patients in activities mimicking those in their workplace and to move the therapy ahead regardless of pain.

Staal et al asked the question: Will a physical therapy program that focuses on functional milestones lead to a more rapid return to work than "routine care"? In other words, is a behavioral approach superior to a pain-focused approach?

Their randomized, controlled study involved KLM Airport workers in Amsterdam who had "nonspecific back pain," meaning that they had no sciatica below the knee and no neurologic deficit. Of the 150 initially recruited, 134 were ultimately randomized into the study. (Patients were excluded if they had legal action pending with the airline.) Patients had to have been out of work for 4 weeks due to back pain to be eligible. Participants were stratified, based on what type of work did for the airline, into 5 groups: baggage and aircraft turnaround, passenger services, engineering and maintenance, cargo, and cabin and cockpit. They were then randomized to graded activity group or regular care. Research assistants were blind to the allocation they gave to patients in sealed envelopes.

The intervention group received 1 hour of treatment twice a week, for up to 3 months. The control group did general exercises, e.g. rowing and cycling, abdominal, upper and lower back exercises, according to the low back pain guidelines of the Dutch College of General Practitioners. The intervention group was given individual exercises chosen to match their occupational or home disability, e.g. lifting suitcases. The two groups were comparable pre-study in age, sex, type of work done at the airline, number of weeks of work absence prior to randomization, functional status and pain level.

Operant Conditioning

The graded exercise intervention used the principles of operant conditioning, in which a reinforcer of a particular activity is used to produce a conditioned response. In this case, the hypothesis was that stimulating exercise, especially with an activity which simulated the activity the patient had difficulty with at work, should decrease pain behavior, and allow an earlier return to work.

In the graded exercise group, key steps were followed:

  1. patients were reassured as to the benign nature of their problem;
  2. the goal of exercise was explained as primarily aimed to help the patient to feel safe in moving, not to increase strength;
  3. patients were told that relief of pain was not a pre-requisite for returning to work;
  4. pain relief was de-emphasized as the main goal; rather, the focus was on "abilities, not disabilities";
  5. the patient, with the therapist, planned a gradually increasing quota for each exercise, to be reached with each session, which the patient was encouraged to achieve despite pain;
  6. The patient, with the therapist, proposed a date for return to work.


The study measured, following study entry, the number of days of absence from work, functional status and severity of pain (0-10 scale). The analysis was on an intention-to-treat basis (and only 3 dropouts were noted from the graded activity group). The mean number of days of work absence was 58 in the graded activity group and 87 in the usual care group, with hazard ratio of 1.9 (P=.009). If the three non-adherent patients were included, the hazard ratio was 2.0 (P=.004). This significant benefit for the graded activity group was only present after 50 days of treatment. For pain reduction and functional status at 3 and 6 months, the results favored the graded exercise group, but in both cases the result was non-significant.

The authors of the study felt that the workers' changed perception of their condition seems to be the chief mechanism of success, since they went back to work earlier, despite having similar pain and functional levels as the control group.

In an editorial in the same issue of the Annals of Internal Medicine, Dr. James Weinstein[3] noted how patients can "unlearn" pain behaviors. He also pointed out that athletes and professionals may be different from injured airline workers, since motivational factors may be quite different.

The Staal study has the strengths of having complete records of absence from work due to complete occupational health records in the Dutch system. Randomization, blinding of research assistants, objective measures of outcome, high rate of study completion and an intention-to-treat analysis were positive factors, and the data was also analyzed "per protocol" and found not to vary from the intention-to-treat analysis.

However, the Staal study also leaves many questions unanswered. As Dr. Weinstein[3] notes, issues such as depression, education and arthritis in other joints were not measured and may have differed between the two groups. The six-month follow-up was short in terms of the overall natural history of low back pain, and the long-term benefits of the graded exercise program are unknown.

Whether one can generalize from this data is not completely clear. The Dutch medical system pays for a year of disability regardless of the cause, and cases with "legal involvement" were eliminated from the study, so results might be quite different in the adversarial Workers' Compensation system in the U.S. The back pain patients in this study were also a very narrow slice of the overall back pain population, having "nonspecific" back pain and already out of work for 4 weeks prior to study entry.

There remains controversy as to whether the focus on "back-to-work" is the correct one for patients with back pain. Pain and functional status are still important. In the Workers' Compensation system in the U.S., a strong focus on "back to work" has potential for possible abuse by employers.

Nonetheless, at Hospital for Special Surgery in New York, the Physical Therapy group has for some time embraced the operant conditioning model and attempted to focus patients on getting back to their work and leisure activities. The Staal study helps by providing objective data on the benefits of focusing on key activities in the therapeutic plan. This study, and future work in this area, should help to assure that referring physicians and therapists are giving their patients the same message. Patients need to appreciate that pain need not stop them from getting back to activity and that they should be actively involved in developing their exercise regimens and back-to-work plans. The Staal study is a step in the right direction - toward managing patient expectations and helping them return to work and leisure activities.

[1] Staal JB, Hlobil H, Twisk JWR, Smid T, Köke AJA. van Mechelen W. Graded Activity for Low Back Pain in Occupational Health Care. Annals Int Med 2004 Jan;140(2):77-84.

[2] Lindstrom I, Ohlund C, Eek C, Wallin L, Peterson LE, Fordyce WE, Nachemson AL. The effect of graded activity on patients with subacute low back pain: a randomized prospective clinical study with an operant-conditioning behavioral approach. Phys Ther 1992;72(4):279-90.

[3] Weinstein J. Absent from Work: Nature versus Nurture. Annals Int Med 2004 Jan:140(2):142-143.

^ Back to Top
Request an Appointment