Research Review By Dr. Michael Haneline ©

Audio:

Download MP3

Date Posted:

July 2011

Study Title:

A systematic review on the effectiveness of physical and rehabilitation interventions for chronic non-specific low back pain

Authors:

van Middelkoop M, Rubinstein S, Kuijpers T, et al.

Author's Affiliations:

Department of General Practice, Erasmus MC, University Medical Center, Rotterdam, The Netherlands.

Publication Information:

European Spine Journal 2011; 20:19–39.

Background Information:

Low back pain (LBP) is typically considered to be chronic after 3 months, since injured tissues normally heal within 6 to12 weeks. It is thought that about one third of patients who seek treatment for LBP will continue to have their problem after 1 year and of those who remain disabled for more than 6 months, less than half of them will return to work.

In the USA, chronic LBP is the most common cause of disability in those who are under 45 years of age. Although only a minority of LBP patients become chronic, they account for most of the economic burden associated with LBP.

The objective of this systematic review was to determine the effectiveness of physical and rehabilitation interventions for chronic LBP, including the following: exercise therapy, back school, transcutaneous electrical nerve stimulation (TENS), low level laser therapy, education, massage, behavioral treatment, traction, multidisciplinary treatment, lumbar supports, and heat/cold therapy.

Pertinent Results:

A total of 114 full text articles were screened for eligibility from Cochrane reviews, resulting in the inclusion of 58 studies. Database searches netted an additional 1,825 potential studies which were screened for inclusion, but only 35 of them fulfilled the inclusion criteria. Some of the studies were duplicates, so ultimately only 83 studies were included in the review.

The breakdown of the 83 studies with regard to which subject they dealt with is as follows:
  • exercise therapy (n = 37),
  • back schools (n = 5),
  • TENS (n = 6),
  • low-level laser therapy (n = 3),
  • massage [65–67] (n = 3),
  • behavioral treatment (n = 21),
  • patient education (n = 1),
  • traction (n = 1), and
  • multidisciplinary treatment (n = 6).
There were 8,816 patients included in all of the studies combined. The studies that dealt with exercise therapy involved the most subjects (3,957 patients), the behavioral studies were next (2,062 patients), and then multidisciplinary studies (1,229 patients).

The authors established a threshold for studies to be considered as having a low risk of bias wherein they had to meet six or more of the criteria (out of 11 possible). However, only 28 studies (33.7%) met or exceeded the established threshold.

A summary of the effectiveness of the interventions is as follows:

Exercise therapy versus:
  • Waiting list controls/no treatment: There is low quality evidence from 8 studies pointing to no statistically significant difference in pain reduction and improvement of disability between groups.
  • Usual care/advise to stay active: The results of 6 studies showed that there is low quality evidence that exercise therapy is effective in reducing pain intensity and disability.
  • Back school/education: Very low quality evidence from 4 studies showed that there was no statistically significant difference in effect on pain and disability at short- and intermediate follow-up between groups.
  • Behavioral treatment: There is low quality evidence from 3 studies that there were no statistically significant differences between groups regarding pain intensity and disability at short- and long-term follow-up.
  • TENS/laser therapy/ultrasound/massage: Low quality evidence from 5 studies pointed to no statistically significant difference in effect between groups regarding pain and disability at short-term follow-up.
  • Manual therapy/manipulation: Low quality evidence from 5 studies indicates that there was no statistically significant difference in effect regarding pain intensity and disability between groups at short and long-term follow-up.
  • Psychotherapy: One study with a high risk of bias showed a statistically significant difference in favor of exercise regarding disability scores, but not for pain intensity. Both disability and pain intensity scores were lower in the exercise group at the 6 months follow-up, although the differences between groups were not statistically significant.
  • Other forms of exercise therapy: Out of 11 studies, 8 of them did not find any differences between groups; whereas an aerobic exercise training program was shown to be statistically significantly superior to a lumbar flexion exercise program in 1 study with a high risk of bias after 3 months, a study that investigated a motor control exercise group compared to a general exercise group reported significantly better outcomes at 8 weeks follow-up associated with the motor control exercise, and another study that compared a 12-week yoga program to a 12-week conventional exercise program reported that function was superior in the yoga group.
Back school versus:
  • Waiting list controls/no treatment/usual care: No statistically significant difference could be shown between groups regarding pain and disability in 3 studies that were of low quality.
  • Active treatment: Low quality evidence from 2 studies pointed to no statistically significant difference between groups.
  • Education/information: One study having a high risk of bias showed a statistically significant difference in pain intensity and disability in favor of the back school group at 6 months follow-up, but only for disability at 12 months follow-up.
Transcutaneous electrical nerve stimulation (TENS) versus:
  • Sham TENS treatment: Low quality evidence from 5 studies pointed to no statistically significant differences between groups regarding post-treatment pain intensity and disability.
  • Percutaneous electrical nerve stimulation (PENS)/acupuncture: Very low quality evidence from 5 studies pointed to PENS/acupuncture as being more effective than TENS for short-term pain relief.
  • Active treatments: Two studies, 1 with a high risk of bias, showed that there was no statistically significant difference between groups regarding pain intensity.
  • Biphasic new wave TENS: One study with a high risk of bias reported no statistically significant difference between groups for the outcomes of pain and disability.
Low-level laser therapy (LLLT) versus:
  • Sham treatment: One study with a low risk of bias provided low quality evidence that LLLT was more effective in pain relief at intermediate follow-up.
  • Sham plus exercise: Very low quality evidence from 2 studies showed that LLLT plus exercise was more effective that sham LLLT plus exercise on pain intensity at short-term follow-up, but not for disability.
  • Exercise: Only 1 study compared the effectiveness of LLLT with exercise therapy, but there was no statistically significant difference between the groups.
Patient education:
  • Versus active non-educational interventions: Low quality evidence from 3 studies that compared with physiotherapy, 1 study involving yoga and conventional exercise and 1 involving Swedish Back School pointed to no statistically significant differences between groups.
  • One study having a high risk of bias compared patient education that focused on anatomy with education that focused on the neurosystem. There was a significant reduction in disability when the focus on the was on the neurosystem, but no differences on pain perception.
Massage therapy:
Three studies that offered low quality evidence pointed to no statistically significant differences on pain intensity post-treatment when massage therapy was compared to passive interventions.

Traction
Only 1 study with a high risk of bias was found in this category which compared motorized traction plus standard physiotherapy to standard physiotherapy alone. There were no statistically significant differences found between groups on pain intensity, disability, and recovery immediately after treatment and at 3 months follow-up.

Behavioral treatment:
  • Versus no treatment/waiting list controls/placebo: Twelve studies using various types of behavioral therapies provided low quality evidence for the effectiveness of behavioral therapy when compared to no treatment/waiting list controls/placebo for pain intensity and disability at short-term follow-up.
  • Versus other kinds of treatment: There is low to moderate quality evidence, based on 6 studies, that indicates no difference between groups regarding pain intensity and disability at short- and long-term follow-up.
  • Plus another treatment versus the other treatment alone: Seven studies provided low to moderate quality evidence that points to no significant effect of behavioral therapy plus another treatment when compared to the other treatment alone in pain intensity and disability at short- and long-term follow-up.
  • Comparisons between different types of behavioral treatment: cognitive versus operant – patients in one study who received operant therapy reported significantly greater improvement in function, but not in pain intensity; cognitive versus respondent – based on 2 studies with a high risk of bias, low quality evidence points to no effect on pain and disability between groups who received these therapies; operant therapy versus a graded activity program – one study with a low risk of bias found no significant differences on pain intensity and disability between the groups; group versus individual cognitive-behavioral treatment – based on one study, there were no significant differences between groups regarding pain intensity and disability.
Multidisciplinary treatment:
  • Versus no treatment/waiting list controls: There was moderate quality evidence from 3 studies that multidisciplinary treatment was significantly better than no treatment/waiting list as regards short-term pain intensity but not in the long-term, also that treated patients had fewer days of sick leave than a those in a non-treated group.
  • Versus other kinds of active treatment:Four studies were identified that compared multidisciplinary treatment to inpatient exercises, physiotherapy, usual care, and exercise therapy. The conclusion was that there is moderate evidence for the effectiveness of multidisciplinary treatment compared to other kinds of active treatment on pain intensity in the short-term, but not in the long-term.
  • Outpatient versus inpatient: No statistically significant differences were found between groups, based on one study with a high risk of bias that compared a 3-week inpatient back school rehabilitation program with a 15-session outpatient back school rehabilitation program.

Clinical Application & Conclusions:

This review concluded that the physical and rehabilitation interventions with the most support for treating chronic LBP patients are multidisciplinary treatment and behavioral treatment. Additionally, exercise therapy was found to reduce pain intensity and disability significantly better than usual care.

The data were insufficient to draw solid conclusions about back schools, low-level laser therapy, patient education, massage, traction, superficial heat/cold, and lumbar supports.

As a result, the authors suggested that multidisciplinary and behavioral treatments, as well as exercise therapy, are the only conservative treatments that should be used in daily practice in the treatment of chronic LBP.

Undoubtedly, many RRS readers utilize some of the “unsupported” interventions listed above and find them to be effective, at least in selected chronic LBP patients. For instance, patients with concomitant paraspinal muscle tension may benefit from heat application or massage simply because it provides at least temporary pain relief and relaxation which facilitates manipulation or other treatment.

A significant part of evidence-based practice is “clinical expertise”, which refers to the knowledge, skills and past experience of the practitioner. Given that chronic LBP patients are often difficult to handle and the evidence for treatment is conflicting, a practitioner may have to try different modalities and combinations of modalities to determine what works best for a particular patient. In reality, the best evidence that a particular management strategy is effective for a particular patient is whether or not the patient improves; even if some of the improvement is related to the placebo effect (1).

Study Methods:

The search strategy first of all screened the references of Cochrane reviews that had considered the interventions of interest for studies that fulfilled the inclusion criteria. Next the MEDLINE, EMBASE, CINAHL, CENTRAL and PEDro databases were similarly searched. The authors also contacted experts on these topics in order to identify any additional studies.

The titles and abstracts of the retrieved citations were independently screened for inclusion by three of the authors based on the review’s inclusion criteria. The full text articles that endured the screening process were obtained and were further screened for inclusion by two authors. Disagreements were resolved by consensus and a third author was consulted if disagreements persisted.

Study inclusion criteria were as follows:
  • must be a RCT;
  • patients had to be adults, older than 18 years; and
  • patients had to have had non-specific chronic LBP for ?12 weeks.
RCTs that investigated the following physical and rehabilitation interventions were included in the review:
  • exercise therapy,
  • back schools,
  • transcutaneous electrical nerve stimulation (TENS),
  • superficial heat or cold,
  • low-level laser therapy (LLLT),
  • individual patient education,
  • massage,
  • behavioral treatment,
  • lumbar supports,
  • traction, and
  • multidisciplinary rehabilitation.
The above interventions should be familiar to RRS readers except “multidisciplinary rehabilitation” which was defined as including multidisciplinary bio-psychosocial rehabilitation that involved at least one physical dimension plus one of the other dimensions (psychological or social or occupational).

The primary outcome measures were pain intensity and physical functional status (i.e., back-specific disability). Other outcome measures that were assessed included:
  • perceived recovery,
  • return to work, and
  • side effects.
Studies were excluded if:
  • the subjects had specific LBP that was caused by a pathology (e.g., spinal stenosis, ankylosing spondylitis, scoliosis and coccydynia);
  • the subjects had post-partum LBP or pelvic pain due to pregnancy;
  • the treatment was for post-operative pain;
  • they were for prevention; and
  • follow-up was for less than one day.
The risk of bias was independently assessed by two reviewers using the 11-item criteria list recommended by the Cochrane Back Review Group. Studies were considered “high quality” when they fulfilled at least 6 of the 11 internal validity criteria. Data were extracted from the included studies by the same two reviewers.

The quality of evidence was determined using the “Grades of Recommendation, Assessment, Development, and Evaluation (GRADE)” scheme and the results were used to develop the strength of the recommendations.

The quality of evidence supporting a given intervention was considered to be high when there were RCTs with a low risk of bias that provided consistent, sufficient, and precise results. Quality ratings were downgraded stepwise when one or more of the factors listed below were not met.
  1. Study limitations (e.g., inadequate study design)
  2. Consistency of results
  3. Indirectness (e.g. generalizability of the findings)
  4. Precision (e.g. sufficient data)
  5. Other considerations
In addition to the “high quality” rating, evidence could also be rated as “moderate quality”, “low quality” and “very low quality”.

Study Strengths / Weaknesses:

One limitation of this review has to do with the fact that when studies reported outcomes only by way of graphs, means and standard deviations were estimated from the graphs. Also, standard deviations were estimated in some cases based on data from other studies. These extrapolations may have introduced a degree of error, although it is not possible to determine how much.

The authors listed several other possible limitations, as follows:
  • Biases related to the literature search and article selection procedure that resulted in relevant studies being missed. Normally these would be unpublished trials and small studies with non-significant results.
  • Since trials with a positive result are more likely to be referred to in other publications, the screening of reference sections of identified articles may have over-represented positive studies in the review, leading to reference bias.
  • Studies published languages other than English, Dutch or German were not included.
Although the GRADE method was utilized in an attempt to account for the risk of bias in the studies, very few studies were rated as high quality which may have resulted in an overestimation of effect.

Additional References:

  1. Bialosky J, et al. Placebo response to manual therapy: something out of nothing? Journal of Manual and Manipulative Therapy, 2011;19(1):11-19.