PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of bmjclinevidLink to Publisher's site
 
Clin Evid (Online). 2011; 2011: 1102.
Published online May 9, 2011.
PMCID: PMC3217769
Low back pain (acute)
Greg McIntosh, Epidemiologist, Manager of Clinical Research# and Hamilton Hall, Medical Director#
Greg McIntosh, CBI Health Group Research Dept, Toronto, ON, Canada;
#Contributed equally.
GM and HH declare that they have no competing interests.
Introduction
Low back pain affects about 70% of people in resource-rich countries at some point in their lives. Acute low back pain can be self-limiting; however, 1 year after an initial episode, as many as 33% of people still have moderate-intensity pain and 15% have severe pain. Acute low back pain has a high recurrence rate; 75% of those with a first episode have a recurrence. Although acute episodes may resolve completely, they may increase in severity and duration over time.
Methods and outcomes
We conducted a systematic review and aimed to answer the following clinical questions: What are the effects of oral drug treatments for acute low back pain? What are the effects of local injections for acute low back pain? What are the effects of non-drug treatments for acute low back pain? We searched: Medline, Embase, The Cochrane Library, and other important databases up to December 2009 (Clinical Evidence reviews are updated periodically, please check our website for the most up-to-date version of this review). We included harms alerts from relevant organisations such as the US Food and Drug Administration (FDA) and the UK Medicines and Healthcare products Regulatory Agency (MHRA).
Results
We found 49 systematic reviews, RCTs, or observational studies that met our inclusion criteria. We performed a GRADE evaluation of the quality of evidence for interventions.
Conclusions
In this systematic review we present information relating to the effectiveness and safety of the following interventions: acupuncture, advice to stay active, analgesics (paracetamol, opioids), back exercises, back schools, bed rest, behavioural therapy, electromyographic biofeedback, epidural corticosteroid injections, lumbar supports, massage, multidisciplinary treatment programmes, muscle relaxants, non-steroidal anti-inflammatory drugs (NSAIDs), spinal manipulation, temperature treatments (short-wave diathermy, ultrasound, ice, heat), traction, and transcutaneous electrical nerve stimulation (TENS).
Low back pain is pain, muscle tension, or stiffness, localised below the costal margin and above the inferior gluteal folds, with or without referred or radicular leg pain (sciatica), and is defined as acute when pain persists for <12 weeks.
  • Low back pain affects about 70% of people in resource-rich countries at some point in their lives.
  • Acute low back pain may be self-limiting, although there is a high recurrence rate with less-painful symptoms recurring in 50% to 80% of people within 1 year of the initial episode; 1 year later, as many as 33% of people still experience moderate-intensity pain and 15% experience severe pain.
NSAIDs have been shown to effectively improve symptoms compared with placebo. However, their use is associated with gastrointestinal adverse effects.
  • Muscle relaxants may also reduce pain and improve overall clinical assessment, but are associated with some severe adverse effects including drowsiness, dizziness, and nausea.
  • The studies examining the effects of analgesics such as paracetamol or opioids were generally too small to detect any clinically important differences.
We found no studies examining the effectiveness of epidural injections of corticosteroids in treating people with acute low back pain.
With regard to non-drug treatments, advice to stay active (be it as a single treatment or in combination with other interventions such as back schools, a graded activity programme, or behavioural counselling) may be effective.
  • We don't know whether spinal manipulation improves pain or function compared with sham treatments.
  • We found insufficient evidence to judge the effectiveness of acupuncture, back schools, behavioural therapy, massage, multidisciplinary treatment programmes (for either acute or subacute low back pain), lumbar supports, TENS, or temperature treatments in treating people with acute low back pain.
  • We found no evidence examining the effectiveness of electromyographic biofeedback or traction in the treatment of acute low back pain.
  • Back exercises may decrease recovery time compared with no treatment, but there is considerable heterogeneity among studies with regard to the definition of back exercise. There is a large disparity among studies in the definition of generic versus specific back exercise.
  • Bed rest does not improve symptoms any more effectively than other treatments, but does produce a number of adverse effects including joint stiffness, muscle wasting, loss of bone mineral density, pressure sores, and venous thromboembolism.
Definition
Low back pain is pain, muscle tension, or stiffness, localised below the costal margin and above the inferior gluteal folds, with or without referred or radicular leg pain (sciatica).[1] For this review, acute low back pain is defined as pain that persists for <12 weeks. Non-specific low back pain is a term some people use to indicate back pain not attributed to a recognisable pathology or symptom pattern (such as infection, tumour, osteoporosis, rheumatoid arthritis, fracture, or inflammation).[1] This review excludes acute low back pain with symptoms or signs at presentation that suggest a specific underlying pathoanatomical condition. People with solely sciatica (lumbosacral radicular syndrome), herniated discs, or both are also excluded. Unless otherwise stated, people included in this review have acute low back pain (i.e., of <12 weeks' duration). Some included RCTs further subdivided acute low back pain of <12 weeks' duration into acute (<6 weeks' duration) or subacute (6–12 weeks' duration).
Incidence/ Prevalence
Over 70% of people in resource-rich countries will experience low back pain at some time in their lives.[2] Each year, 15% to 45% of adults suffer low back pain, and 1/20 (5%) people present to a healthcare professional with a new episode. Low back pain is most common between the ages of 35 to 55 years.[2] About 30% of European workers reported that their work caused low back pain but in a Canadian study, 67% of people not involved in workers' compensation claims could not attribute their symptoms to any specific cause or precipitating event.[3] Prevalence rates from different countries range from 13% to 44%. The longer the period of sick leave, the less likely return to work becomes.[2] [4]
Aetiology/ Risk factors
Symptoms, pathology, and radiological appearances are poorly correlated. An anatomical source of pain cannot be identified in about 80% of people. About 4% of people with low back pain in primary care have compression fractures and only about 1% have a tumour.[5] The prevalence of prolapsed intervertebral disc is about 1% to 3%.[2] Ankylosing spondylitis and spinal infections are less common.[5] Risk factors for the development of back pain include heavy physical work; frequent bending, twisting, or lifting; and prolonged static postures including sitting. Psychosocial risk factors include anxiety, depression, and mental stress at work.[2] [6]
Prognosis
Acute low back pain may be self-limiting, although acute low back pain has a high recurrence rate with symptoms recurring in 50% to 80% of people within 1 year;[7] 1 year after the initial episode, as many as 33% of people still endure moderate-intensity pain and 15% experience severe pain.
Aims of intervention
Aims include: to relieve pain, to improve function, to reduce time taken to return to work, to develop coping strategies for pain, with minimal adverse effects from treatment; and to prevent the development of chronic back pain (see definition in review on low back pain [chronic]).[8] [9]
Outcomes
Symptom improvement: pain intensity (visual analogue or numerical rating scale); overall improvement (self-reported or observed); medication use; intervention-specific outcomes (such as coping and pain behaviour for behavioural treatment, strength and flexibility for exercise, and muscle spasm for muscle relaxants and electromyographic biofeedback). Functional improvement: back pain-specific functional status (such as Roland Morris questionnaire, Oswestry questionnaire). Return to work: impact on employment (days of sick leave, number of people returned to work). Adverse effects of treatments. Treatment effects: some people have argued that the small effects of treatments are a consequence of the favourable natural history of non-specific low back pain.[10] The theory is that control groups have improved substantially and so there is not "room" for large treatment effects. To evaluate this argument, one review examined the baseline and follow-up scores from the acute trials in a meta-analysis.[10] The study found that the theory of no "room" for improvement does not seem consistent with the data; there is scope for treatment effects (i.e., mean between-group differences as large as 40 points that can be demonstrated in acute non-specific low back pain trials). Another argument used to explain the small treatment effects found in the non-specific low back pain literature is that most trials are conducted on samples from clinically heterogeneous populations. It is probable that specific treatments have large treatment effects on specific subgroups of patients with non-specific low back pain.[10]
Methods
Clinical Evidence search and appraisal December 2009. The following databases were used to identify studies for this systematic review: Medline 1966 to December 2009, Embase 1980 to December 2009, and The Cochrane Database of Systematic Reviews 2009, Issue 4 (1966 to date of issue). An additional search within The Cochrane Library was carried out for the Database of Abstracts of Reviews of Effects (DARE) and the Health Technology Assessment (HTA) database. We also searched for retractions of studies included in the review. Most earlier RCTs of acute low back pain treatments were small (fewer than 50 people/intervention group), short term (mostly <6 months' follow-up), and of low overall quality. The quality of many recent RCTs is higher. In this review, we have excluded studies done solely in people with sciatica or disc herniation. We have included studies in people with acute low back pain in which the study does not describe whether people had radiation, or in which the study included people without radiation. Abstracts of the studies retrieved from the initial search were initially assessed by an information specialist. Selected studies were then sent to the contributors for additional assessment, using predetermined criteria to identify relevant studies. Study design criteria for inclusion in this review were: published systematic reviews of RCTs and RCTs in English language. RCTs had to be at least single blinded, unless blinding was impossible (e.g., physical treatments). We excluded all studies described as "open", "open label", or not blinded unless blinding was impossible. RCTs had to contain 20 or more individuals. There was no minimum length of follow-up or maximum loss to follow-up. We included systematic reviews of RCTs and RCTs where harms of an included intervention were studied applying the same study design criteria for inclusion as we did for benefits. In addition we use a regular surveillance protocol to capture harms alerts from organisations such as the FDA and the MHRA, which are added to the reviews as required. To make numerical data in our reviews more readable, we round many percentages to the nearest whole number. Readers should be aware of this approximation when relating percentages to summary statistics such as relative risks (RRs) and odds ratios (ORs). We have performed a GRADE evaluation of the quality of evidence for interventions included in this review (see table ). The categorisation of the quality of the evidence (into high, moderate, low, or very low) reflects the quality of evidence available for our chosen outcomes in our defined populations of interest. These categorisations are not necessarily a reflection of the overall methodological quality of any individual study, because the Clinical Evidence population and outcome of choice may represent only a small subset of the total outcomes reported, and population included, in any individual trial. For further details of how we perform the GRADE evaluation and the scoring system we use, please see our website (www.clinicalevidence.com).
Table 1
Table 1
GRADE evaluation of interventions for low back pain (acute)
Glossary
Acupuncture Needle puncture of the skin at traditional "meridian" acupuncture points. Modern acupuncturists also use non-meridian points and trigger points (tender sites occurring in the most painful areas). The needles may be stimulated manually or electrically. Placebo acupuncture is needling of traditionally unimportant sites or non-stimulation of the needles once placed.
Back school Traditionally, this is a series of group education sessions on low back pain. Sessions are usually supervised by a physiotherapist or physician and often include information on an exercise programme.
Cesar therapyExercise programme to improve posture and so reduce back pain caused by poor posture.
Cognitive behavioural therapy This aims to identify and modify people's understanding of their pain and disability using cognitive restructuring techniques (such as imagery and attention diversion) or by modifying maladaptive thoughts, feelings, and beliefs.
Electromyographic biofeedback A person receives external feedback of their own electromyogram (using visual or auditory scales), and uses this to learn how to control the electromyogram and hence the tension within their own muscles. Electromyogram biofeedback for low back pain aims to relax the paraspinal muscles.
Generic back exercise (low back pain)In this review, generic back exercise denotes undifferentiated exercise/movements performed in multiple directions or planes without emphasis on the person’s pattern of pain or directional preference for pain control.
Low-quality evidenceFurther research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
McKenzie (direction-specific) exercise A method of physiotherapy that involves a comprehensive mechanical diagnosis and treatment to assess the effects on patient symptoms of end-range repetitive movements, static positioning, or both. The mechanical diagnosis enables physiotherapists to prescribe individual exercises in a specific preferred direction. The emphasis is on patient responsibility and self-treatment. Mobilisation techniques are used in more difficult mechanical cases until patients can perform the prescribed exercises on their own.
Moderate-quality evidenceFurther research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Multidisciplinary treatment Intensive physical and psychosocial training by a team (e.g., a physician, physiotherapist, psychologist, social worker, and occupational therapist). Training is usually given in groups and does not involve passive physiotherapy.
Sciatica Radicular leg pain emanating from irritation in one of the roots of the sciatic nerve and following the nerve's distribution.
Very low-quality evidenceAny estimate of effect is very uncertain.

Notes
Disclaimer
The information contained in this publication is intended for medical professionals. Categories presented in Clinical Evidence indicate a judgement about the strength of the evidence available to our contributors prior to publication and the relevant importance of benefit and harms. We rely on our contributors to confirm the accuracy of the information presented and to adhere to describe accepted practices. Readers should be aware that professionals in the field may have different opinions. Because of this and regular advances in medical research we strongly recommend that readers' independently verify specified treatments and drugs including manufacturers' guidance. Also, the categories do not indicate whether a particular treatment is generally appropriate or whether it is suitable for a particular individual. Ultimately it is the readers' responsibility to make their own professional judgements, so to appropriately advise and treat their patients.To the fullest extent permitted by law, BMJ Publishing Group Limited and its editors are not responsible for any losses, injury or damage caused to any person or property (including under contract, by negligence, products liability or otherwise) whether they be direct or indirect, special, incidental or consequential, resulting from the application of the information in this publication.
Contributor Information
Greg McIntosh, CBI Health Group Research Dept, Toronto, ON, Canada.
Hamilton Hall, CBI Health Group, Toronto, ON, Canada.
1. Van der Heijden GJMG, Bouter LM, Terpstra-Lindeman E. The efficacy of traction for low back pain: results of a randomized blinded pilot study. Ned T Fysiotherapie 1991;101:37–43. [In Dutch]
2. Andersson GBJ. The epidemiology of spinal disorders. In: Frymoyer JW, ed. The adult spine: principles and practice. 2nd ed. New York: Raven Press, 1997:93–141.
3. Hall H, McIntosh G, Wilson L, et al. Spontaneous onset of back pain. Clin J Pain1998;14:129–133. [PubMed]
4. Waddell G. The back pain revolution. Edinburgh: Churchill Livingstone; 1998.
5. Deyo RA, Rainville J, Kent DL. What can the history and physical examination tell us about low back pain? JAMA1992;268:760–765. [PubMed]
6. Bongers PM, de Winter CR, Kompier MA, et al. Psychosocial factors at work and musculoskeletal disease. Scand J Work Environ Health1993;19:297–312. [PubMed]
7. Frymoyer JW. Back pain and sciatica. N Engl J Med1988;318:291–300. [PubMed]
8. Bigos S, Bowyer O, Braen G, et al. Acute low back problems in adults. Clinical Practice Guideline no. 14. AHCPR Publication No. 95-0642. Rockville MD: Agency for Health Care Policy and Research, Public Health Service, US, Department of Health and Human Services. December 1994. Search date not reported.
9. Evans G, Richards S. Low back pain: an evaluation of therapeutic interventions. Bristol: Health Care Evaluation Unit, University of Bristol, 1996. Search date 1995.
10. Machado LA, Kamper SJ, Herbert RD, et al. Analgesic effects of treatments for non-specific low back pain: a meta-analysis of placebo-controlled randomized trials. Rheumatology2009;48:520–527. [PubMed]
11. van Tulder MW, Touray T, Furlan AD, et al. Muscle relaxants for non-specific low back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK; John Wiley & Sons, Ltd. Search date 2002.
12. Moll W. Therapy of acute lumbovertebral syndromes through optimal muscle relaxation using diazepam. Results of a double-blind study on 68 cases. Med Welt1973;24:1747–1751. [In German] [PubMed]
13. Ralph L, Look M, Wheeler W, et al. Double-blind, placebo-controlled trial of carisoprodol 250-mg tablets in the treatment of acute lower-back spasm. Curr Med Res Opin2008;24:551–558. [PubMed]
14. Hoiriis KT, Pfleger B, McDuffie FC, et al. A randomized clinical trial comparing chiropractic adjustments to muscle relaxants for subacute low back pain. J Manipulative Physiol Ther2004;27:388–398. [PubMed]
15. Boyles W, Glassman J, Soyka J. Management of acute musculoskeletal conditions: thoracolumbar strain or sprain. Double-blind evaluation comparing the efficacy and safety of carisoprodol with diazepam. Today's Ther Trends 1983;1:1–16.
16. Rollings H. Management of acute musculoskeletal conditions – thoracolumbar strain or sprain: a double-blind evaluation comparing the efficacy and safety of carisoprodol with cyclobenzaprine hydrochloride. Curr Ther Res l983;34:917–928.
17. Hennies O. A new skeletal muscle relaxant (DS 103–282) compared to diazepam in the treatment of muscle spasm of local origin. Int Med Res1981;9:62–68. [PubMed]
18. Rusinyol FCP. Effects of two different doses of eperisone in the treatment of acute low back pain. J Appl Res 2009;9:23–29.
19. Roelofs PDDM, Deyo RA, Koes BW, et al. Non-steroidal anti-inflammatory drugs for low back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2007.
20. Laws D. Double blind parallel group investigation in general practice of the efficacy and tolerability of acemetacin, in comparison with diclofenac, in patients suffering with acute low back pain. Br J Clin Res 1994;5:55–64.
21. Zippel H, Wagenitz A. A multicentre, randomised, double-blind study comparing the efficacy and tolerability of intramuscular dexketoprofen versus diclofenac in the symptomatic treatment of acute low back pain. Clin Drug Investig2007;27:533–543. [PubMed]
22. Henry D, Lim LLY, Rodriguez LAG, et al. Variability in risk of gastrointestinal complications with individual non-steroidal anti-inflammatory drugs: results of a collaborative meta-analysis. BMJ1996;312:1563–1566. Search date 1994. [PMC free article] [PubMed]
23. Dequeker J, Hawkey C, Kahan A, et al. Improvement in gastrointestinal tolerability of the selective cyclooxygenase (COX)-2 inhibitor, meloxicam, compared with piroxicam: results of the Safety and Efficacy Large-scale Evaluation of COX-inhibiting Therapies (SELECT) trial in osteoarthritis. Br J Rheumatol1998;37:946–951. [PubMed]
24. Harms alert for Bextra: European suspension of Bextra. MHRA Press Release, 2005.
25. Van Tulder MW, Koes BW, Bouter LM. Conservative treatment of acute and chronic nonspecific low back pain: a systematic review of randomized controlled trials of the most common interventions. Spine1997;22:2128–2156. Search date 1995. [PubMed]
26. Perrot S, Krause D, Crozes P, et al. Efficacy and tolerability of paracetamol/tramadol (325 mg/37.5 mg) combination treatment compared with tramadol (50 mg) monotherapy in patients with subacute low back pain: a multicenter, randomized, double-blind, parallel-group, 10-day treatment study. Clin Ther2006;28:1592–1606. [PubMed]
27. Koes BW, Scholten RJPM, Mens JMA, et al. Epidural steroid injections for low back pain and sciatica: an updated systematic review of randomized clinical trials. Pain Digest 1999;9:241–247. Search date 1998.
28. Waddell G, Feder G, Lewis M. Systematic reviews of bed rest and advice to stay active for acute low back pain. Br J Gen Pract1997;47:647–652. Search date not reported. [PMC free article] [PubMed]
29. Anema JR, Steenstra IA, Bongers PM, et al. Multidisciplinary rehabilitation for subacute low back pain: Graded activity or workplace intervention or both? A randomized controlled trial. Spine2007;32:291–298. [PubMed]
30. Steenstra IA, Anema JR, Bongers PM, et al. The effectiveness of graded activity for low back pain in occupational healthcare. Occup Environ Med2006;63:718–725. [PMC free article] [PubMed]
31. Karjalainen K, Malmivaara A, van Tulder M, et al. Multidisciplinary biopsychosocial rehabilitation for subacute low back pain among working age adults. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2002. [PubMed]
32. Norlund A, Ropponen A, Alexanderson K, et al. Multidisciplinary interventions: review of studies of return to work after rehabilitation for low back pain. J Rehab Med2009;41:115–121. [PubMed]
33. Assendelft WJJ, Morton SC, Yu EI, et al. Spinal manipulative therapy for low-back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2000. [PubMed]
34. Juni P, Battaglia M, Nuesch E, et al. A randomised controlled trial of spinal manipulative therapy in acute low back pain. Ann Rheum Dis2009;68:1420–1427. [PubMed]
35. Assendelft WJJ, Bouter LM, Knipschild PG. Complications of spinal manipulation: a comprehensive review of the literature. J Fam Pract1996;42:475–480. [PubMed]
36. Childs JD, Flynn TW, Fritz JM. A perspective for considering the risks and benefits of spinal manipulation in patients with low back pain. Man Ther2006;11:316–320. [PubMed]
37. Chou R, Huffman LH, American Pain Society, et al. Nonpharmacologic therapies for acute and chronic low back pain: a review of the evidence for an American Pain Society/American College of Physicians clinical practice guideline. Ann Intern Med2007;147:492–504. [PubMed]
38. Waddell G, Feder G, McIntosh A, et al. Low back pain evidence review. London: Royal College of General Practitioners, 1999. Search date 1999.
39. Furlan AD, van Tulder MW, Cherkin DC, et al. Acupuncture and dry-needling for low back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2003. [PubMed]
40. Ernst E, White A. Life-threatening adverse reactions after acupuncture? A systematic review. Pain1997;71:123–126. Search date 1996. [PubMed]
41. Heymans MW, van Tulder MW, Esmail R, et al. Back schools for non-specific low-back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2003. [PubMed]
42. Newcomer KL, Vickers Douglas KS, Shelerud RA, et al. Is a videotape to change beliefs and behaviors superior to a standard videotape in acute low back pain? A randomized controlled trial. Spine J2008;8:940–947. [PubMed]
43. Gohner W, Schlicht W. Preventing chronic back pain: evaluation of a theory-based cognitive-behavioural training programme for patients with subacute back pain. Patient Educ Couns2006;64:87–95. [PubMed]
44. Calmels P, Queneau P, Hamonet C, et al. Effectiveness of a lumbar belt in subacute low back pain: an open, multicentric, and randomized clinical study. Spine2009;34:215–220. [PubMed]
45. Furlan AD, Imamura M, Dryden T, et al. Massage for low back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2008. [PubMed]
46. French SD, Cameron M, Walker BF, et al. Superficial heat or cold for low back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2005. [PubMed]
47. Tao XG, Bernacki EJ. A randomized clinical trial of continuous low-level heat therapy for acute muscular low back pain in the workplace. J Occup Environ Med/Am Coll Occup Environ Med2005;47:1298–1306. [PubMed]
48. Kettenmann B, Wille C, Lurie-Luke E, et al. Impact of continuous low level heatwrap therapy in acute low back pain patients: subjective and objective measurements. Clin J Pain2007;23:663–668. [PubMed]
49. Hayden JA, Tulder MW van, Malmivaara A, et al. Exercise therapy for treatment of non-specific low back pain. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2004.
50. Machado LA, de Souza MS, Ferreira PH, et al. The McKenzie method for low back pain: a systematic review of the literature with a meta-analysis approach. Spine2006;31:E254–E262. [PubMed]
51. Clare HAA. A systematic review of efficacy of McKenzie therapy for spinal pain. Austr J Physiother2004;50:209–216. [PubMed]
52. Hlobil H, Staal JB, Twisk J, et al. The effects of a graded activity intervention for low back pain in occupational health on sick leave, functional status and pain: 12-month results of a randomized controlled trial. J Occup Rehab2005;15:569–580. [PubMed]
53. Suni J, Rinne M, Natri A, et al. Control of the lumbar neutral zone decreases low back pain and improves self-evaluated work ability: a 12-month randomized controlled study. Spine2006;31:E611–E620. [PubMed]
54. Dettori JR, Bullock SH, Sutlive TG, et al. The effects of spinal flexion and extension exercises and their associated postures in patients with acute low back pain. Spine1995;20:2303–2312. [PubMed]
55. Delitto A, Cibulka MT, Erhard RE, et al. Evidence for use of an extension-mobilization category in acute low back syndrome: a prescriptive validation pilot study. Phys Ther1993;73:216–228. [PubMed]
56. Stankovic R, Johnell O. Conservative treatment of acute low-back pain. A prospective randomized trial: McKenzie method of treatment versus patient education in "mini back school". Spine1990;15:120–123. [PubMed]
57. Stankovic R, Johnell O. Conservative treatment of acute low back pain. A 5-year follow-up study of two methods of treatment. Spine1995;20:469–472. [PubMed]
58. Erhard RE, Delitto A, Cibulka MT. Relative effectiveness of an extension program and a combined program of manipulation and flexion and extension exercises in patients with acute low back syndrome. Phys Ther1994;74:1093–1100. [PubMed]
59. Hagen KB, Hilde G, Jamtvedt G, et al. Bed rest for acute low back pain and sciatica. In: The Cochrane Library, Issue 4, 2009. Chichester, UK: John Wiley & Sons, Ltd. Search date 2003. [PubMed]

Muscle relaxants
Summary
SYMPTOM IMPROVEMENT Benzodiazepine muscle relaxants compared with placebo: Benzodiazepines may be more effective than placebo at reducing pain ( very low-quality evidence ). Non-benzodiazepine muscle relaxants compared with placebo: Non-benzodiazepine muscle relaxants may be more effective than placebo at reducing pain and may be more effective at improving global assessment ( low-quality evidence ). Muscle relaxants compared with NSAIDs: We don't know whether chlormezanone plus paracetamol is more effective than mefenamic acid at increasing the proportion of people with global improvement in people with acute low back pain (very low-quality evidence). Muscle relaxants compared with each other: We don't know if any one muscle relaxant is consistently more effective than all other muscle relaxants at improving symptoms, as we found insufficient evidence (low-quality evidence). FUNCTIONAL IMPROVEMENT Non-benzodiazepine muscle relaxants compared with placebo: We don't know whether non-benzodiazepine muscle relaxants are more effective than placebo at improving disability at 4 weeks, as we found insufficient evidence (low-quality evidence). NOTE Benzodiazepine and non-benzodiazepine muscle relaxants have been associated with an increase in adverse effects compared with placebo, particularly central nervous system effects (such as dizziness, nausea, and possibly drowsiness).
Benefits
Benzodiazepine muscle relaxants versus placebo:
We found one systematic review (search date 2002),[11] which identified one poor-quality RCT (68 people).[12] The RCT found that intramuscular diazepam followed by oral diazepam for 5 days significantly reduced pain and increased the rate of overall improvement (rating scales used to assess overall improvement not reported) compared with placebo (overall effect rated good or very good: 21/33 [64%] with diazepam v 6/35 [17%] with placebo; P value and pain results not reported in the review). However, treatment groups were not comparable at baseline.
Non-benzodiazepine muscle relaxants versus placebo:
We found two systematic reviews (search dates 2001[11] and 2006[10]) and one subsequent RCT.[13] The first review identified 9 RCTs comparing non-benzodiazepines (tizanidine, cyclobenzaprine, carisoprodol, baclofen, orphenadrine) versus placebo.[11] Meta-analysis of RCTs with adequate data found that oral non-benzodiazepines (cyclobenzaprine, tizanidine, and orphenadrine) significantly reduced pain and improved global assessment after 2 to 4 days (presence of pain: 4 RCTs, 294 people; RR 0.80, 95% CI 0.71 to 0.89; global assessment at 2–4 days, dichotomous, assessed by patient: 4 RCTs, 222 people; RR 0.49, 95% CI 0.25 to 0.95).
The second review identified 8 RCTs reporting on the analgesic effects of muscle relaxants estimated by placebo-controlled trials in people with non-specific acute low back pain.[10] This review excluded RCTs in which a contemporary treatment had been given as placebo (i.e., what might now be considered to be an active treatment, e.g., educational booklets, non-specific exercises, TENS, low-dose lidocaine injections, diphenhydramine tablets, low-force spinal manipulation and soft tissue massage, massage with placebo ointment). The review extracted data on outcomes from the first assessment after the end of therapy, as the review considered that this time point was where the largest analgesic effects would be observed. It did not include a description of individual interventions used in each RCT. The review found that muscle relaxants significantly improved pain (measured by analgesic efficacy [100-point scale]) compared with placebo (8 RCTs, 777 people; RR presented graphically; absolute numbers and figures for point estimate of RR and CI not reported; individual RCTs in analysis not reported). The review reported that the point estimate of the effects was moderate for muscle relaxants, but the confidence intervals were not narrow enough to exclude small effects, and noted that generally in all RCTs included in the review, that large effects were observed in only single trials.[10]
One RCT (192 people) included in the second review compared chiropractic adjustments, muscle relaxants, and placebo, and found no significant difference among groups in disability at 4 weeks.[14]
The subsequent RCT (562 people) was a 7-day trial that compared carisoprodol (277 people) versus placebo (285 people).[13] The RCT found that carisoprodol was significantly more effective than placebo at improving participant-rated global impression of change at 3 days (scale 0 = worsening to 4 = marked improvement: 2.24 with carisoprodol v 1.70 with placebo; P <0.0001) and participant-rated relief from initial backache (scale 0 = no relief to 4 = complete relief: 1.83 with carisoprodol v 1.12 with placebo; P <0.0001). Time until onset of moderate or marked improvement was significantly reduced with carisoprodol compared with placebo (3 days with carisoprodol v 6 days with placebo; P <0.0001). The RCT undertook a subgroup analysis to see if the effects were dependent on the sedative effects of carisoprodol, and there was no significant differences found between people who reported sedation (13.4% of participants) and those who did not (86.6% of participants).[13]
Muscle relaxants versus NSAIDs:
See benefits of NSAIDs.
Muscle relaxants versus each other:
We found one systematic review (search date 2001),[11] which identified three RCTs.[15] [16] [17] and one subsequent RCT.[18]
The RCTs in the review found no clinically important differences in effect among muscle relaxants (cyclobenzaprine, carisoprodol, diazepam, and tizanidine), although the results were not pooled in the review. The first RCT (80 people) found that carisoprodol significantly increased overall improvement compared with diazepam, but found no significant difference in pain at 7 days (improvement rated as very good or excellent; 70% with carisoprodol v 45% with diazepam; pain on 100-mm visual analogue scale [VAS]: 58 mm with carisoprodol v 48 mm with diazepam; P values not reported in the review).[15] The second RCT (78 people) found no significant difference between carisoprodol and cyclobenzaprine in pain or overall improvement after 8 days (pain on 100-mm VAS: 30 mm with carisoprodol v 28 mm with cyclobenzaprine; overall improvement good or excellent: 70% with carisoprodol v 70% with cyclobenzaprine; P values not reported in review).[16] The third RCT (30 people with acute back pain, 20% with concomitant acute neck pain) was small and found no significant difference between diazepam and tizanidine in pain or function at 7 days (pain relief: 77.4% with tizanidine v 48.0% with diazepam; improvement in daily activities: 87% with tizanidine v 93% with diazepam; P values not reported in review).[17]
The subsequent RCT (90 people [86 assessed], acute low back pain <48 hours, with mild to severe intensity muscle contracture) compared diazepam 15 mg daily versus eperisone 150 mg daily versus eperisone 300 mg daily, given orally for 7 days.[18] The RCT found that, compared with eperisone 150 mg daily, eperisone 300 mg daily significantly improved pain at rest or palpation, muscular contracture, impaired working capacity, and hand-to-floor distance (a measure of flexion) at 7 days (all outcomes, P <0.01). It found that, compared with diazepam, eperisone 300 mg daily significantly improved muscular contracture at 3 days but not at 7 days, and significantly improved impaired working capacity (measured by limitations of activity) at 3 days and 7 days (all significant outcomes, P <0.01).[18]
Harms
Benzodiazepine or non-benzodiazepine muscle relaxants versus placebo:
The first review found that muscle relaxants (both benzodiazepines and non-benzodiazepines) significantly increased adverse effects, particularly central nervous system effects, compared with placebo (all adverse effects: 8 RCTs, 724 people; RR 1.50, 95% CI 1.14 to 1.98; nervous system effects: 8 RCTs, 724 people; RR 2.04, 95% CI 1.23 to 3.37).[11] The most common adverse effects were drowsiness, dizziness, and nausea. The second review did not report on adverse effects.[10]
The subsequent RCT reported that treatment-emergent adverse events occurring in 3% or more of participants included: drowsiness, dizziness, and headache (drowsiness: 13.4% with carisoprodol v 4.6% with placebo; dizziness: 9.7% with carisoprodol v 3.2% with placebo; headache: 3.6% with carisoprodol v 1.4% with placebo; statistical analysis between groups not reported).[13] Eight (2.9%) people in the carisoprodol group and 5 (1.8%) people in the control group discontinued the study because of treatment-emergent adverse events (statistical analysis between groups not reported). The RCT reported that most adverse events were mild to moderate including those leading to discontinuation. No participant discontinued treatment with carisoprodol because of drowsiness, and no serious adverse events or clinically significant effects on laboratory values or vital signs were detected.[13]
Muscle relaxants versus NSAIDs:
See harms of NSAIDs.
Muscle relaxants versus each other:
The subsequent RCT reported that in the diazepam group, 23 people reported adverse reactions.[18] These included somnolence (19), tachycardia with vertigo (1), epigastric pain (2), and diarrhoea (1). In the eperisone 150 mg group, it reported 5 adverse reactions which were: epigastric pain (3), somnolence (1), and headache (1). The RCT reported that there were 6 adverse reactions in the eperisone 300 mg group: somnolence (2), epigastric pain (1), vertigo (1), urinary retention (1), and anorexia (1). It did not report a statistical analysis between groups.[18]
Comment
None.
Substantive changes
Muscle relaxants New evidence added.[10] [13] [18] Categorisation unchanged (Trade-off between benefits and harms).

Summary
SYMPTOM IMPROVEMENT Compared with placebo: NSAIDs may be more effective than placebo at improving pain in people with acute low back pain ( very low-quality evidence ). Compared with each other: We don't know if any one NSAID is consistently more effective than all other NSAIDs at improving symptoms in people with acute low back pain ( low-quality evidence ). Compared with paracetamol (acetaminophen): We don't know whether NSAIDs and paracetamol (acetaminophen) differ in effectiveness at improving pain or global improvement in people with acute low back pain as we found insufficient evidence (very low-quality evidence). Compared with muscle relaxants: We don't know whether mefenamic acid is more effective than chlormezanone plus paracetamol at increasing the proportion of people with global improvement in people with acute low back pain (very low-quality evidence). Compared with non-drug treatments (physiotherapy or spinal manipulation): We don't know whether diflunisal, physiotherapy, and spinal manipulation differ in effectiveness at improving pain at 4 and 12 days in people with acute low back pain (low-quality evidence). Compared with NSAIDs plus adjuvant treatment: We don't know whether naproxen is more effective than naproxen plus cyclobenzaprine at improving pain in people with acute low back pain (low-quality evidence). Compared with heat wrap: Ibuprofen may be less effective than heat wraps at improving pain at 1 and 4 days (low-quality evidence). FUNCTIONAL IMPROVEMENT Compared with each other: We don't know if dexketoprofen is more effective than diclofenac at improving disability in people with acute low back pain as we found insufficient evidence (low-quality evidence). Compared with heat wrap: Ibuprofen may be less effective than heat wraps at improving disability at 4 days (low-quality evidence). Compared with specific back exercises: We don't know whether NSAIDs are more effective than McKenzie treatment at 3 months at improving short-term disability as we found insufficient evidence (low-quality evidence). NOTE NSAIDs have been associated with an increase in gastrointestinal and other adverse effects compared with placebo.
Benefits
We found two systematic reviews (search dates 2007[19] and 2006[10]), one additional RCT,[20] and one subsequent RCT.[21]
The first review included RCTs in people with acute, chronic, and mixed back pain, and also RCTs in people with no sciatica, all participants with sciatica, mixed populations, and RCTs in which the presence or absence of sciatica was not specified.[19] We have reported RCTs exclusively undertaken in acute back pain that included a homogeneous population of people without sciatica.
NSAIDs versus placebo:
We found two systematic reviews (search dates 2007[19] and 2006[10]).
The first review included three RCTs (724 people) in people with acute low back pain without sciatica and pooled results.[19] The RCTs included piroxicam, ibuprofen, and tenoxicam (see comment below). The review found that NSAIDs significantly improved pain compared with placebo (pain measured on 100-mm visual analogue scale; change from baseline: 3 RCTs, 577 people; mean difference –7.69, 95% CI –12.08 to –3.30; P = 0.00059). Two of the included RCTs were described as being of high methodological quality, while the third was described as being of low quality. One included RCT reported data on the need for additional analgesic use; analgesics were not permitted in one included RCT.[19]
The second review identified three RCTs evaluating the analgesic effects of NSAIDs estimated by placebo-controlled trials in people with non-specific low back pain.[10] In this review, RCTs using a placebo consisting of what might now be considered to be an active treatment were excluded (see benefits of muscle relaxants). The review extracted data on outcomes from the first assessment after the end of therapy, because the review considered that this time point was where the largest analgesic effects would be observed. It did not include a description of individual interventions used in each RCT. The review included two RCTs included in the analysis in the first review, and one further RCT that may have included people with sciatica. The review found that NSAIDs significantly improved pain (measured by analgesic efficacy [100-point scale]) compared with placebo (3 RCTs, 427 people; RR presented graphically; absolute numbers and figures for point estimate of RR and CI not reported; individual RCTs in analysis not reported). The review reported that the point estimate of the effects for NSAIDs was small, and noted that generally in all RCTs included in the review, large effects were observed in only single trials.[10]
NSAIDs versus each other:
We found one systematic review (search date 2007, 6 RCTs, 1353 people),[19] one additional RCT,[20] and one subsequent RCT.[21]
Overall, the review found 21 RCTs comparing different NSAIDs versus each other, of which 15 RCTs found no significant difference between groups.[19] However, some of these RCTs included either a proportion of people, or all the study population, with sciatica. We have therefore reported the 6 included RCTs that compared different traditional NSAIDs in people with acute low back pain without sciatica. The review did not pool data. In the review, 4 high-quality RCTs and two low-quality RCTs compared different types of traditional NSAIDs versus each other. The review reported that none of the high-quality RCTs evaluating NSAIDs found any significant differences between groups (further statistical analysis not reported).[19]
The additional RCT (194 people) found no significant difference in pain or global assessment between acemetacin and diclofenac (absolute numbers and P value not reported).[20]
The subsequent RCT (323 people, acute low back pain, 1 week or less) compared intramuscular dexketoprofen versus diclofenac.[21] The RCT measured pain on a 100-mm visual analogue scale. The RCT found no significant difference between groups in pain intensity (sum of analogue pain intensity difference [SAPID] score at 6 hours after the first dose administration: 111.8 mm/hour with dexketoprofen v 112.7 mm/hour with diclofenac; adjusted ratio of means 0.993, lower 95% CI 0.79, upper 95% CI not reported). The RCT reported that there was no significant difference between groups in disability (median change in Roland Disability Questionnaire score –6 points for both groups; P = 0.69).[21]
NSAIDs versus paracetamol (acetaminophen):
We found one systematic review (search date 2007, 3 RCTs).[19] The three included low-quality RCTs compared some type of NSAID versus paracetamol (acetaminophen). None of the RCTs found a significant difference between groups in outcomes (pain intensity [various scales]: first RCT, 219 people; ibuprofen v paracetamol, SMD –0.09, 95% CI –0.35 to +0.18; second RCT, 30 people; phenylbutazone v paracetamol, SMD –0.58, 95% CI –1.31 to +0.16; proportion of people experiencing global improvement: third RCT, 48 people; ibuprofen v paracetamol; RR 1.23, 95% CI 0.77 to 1.96). There was no reporting of allocation concealment in any of the RCTs, randomisation was according to participant's military identification numbers in one RCT, and two RCTs were undertaken solely in male military participants.[19]
NSAIDs versus muscle relaxants:
We found one systematic review (search date 2007, 1 RCT, 122 people).[19] The RCT found no significant difference between mefenamic acid plus placebo and chlormezanone plus paracetamol plus placebo in the proportion of people with global improvement (80 people; RR 1.05, 95% CI 0.86 to 1.29). The randomisation procedure was not described and allocation concealment was unclear.[19]
NSAIDs versus non-drug treatments (physiotherapy or spinal manipulation):
We found one systematic review (search date 2007, 1 RCT, 112 people).[19] The included RCT found no significant difference between NSAIDs (36 people) and physiotherapy (34 people) or spinal manipulation (38 people) in pain after 4 and 12 days (mean change in pain intensity on 4-point scale; 4 days: –0.9 with diflunisal v –0.9 with physiotherapy v –1.1 with spinal manipulation, McKenzie therapy, or both; 12 days: –1.7 with diflunisal v –1.6 with physiotherapy v –1.7 with spinal manipulation, McKenzie therapy, or both; reported as no significant difference; P value not reported).[19] However, the study lacked power because of the small groups.
NSAIDs versus NSAIDs plus adjuvant treatment:
We found one systematic review (search date 2007, 1 RCT).[19] The review reported that the low-quality RCT (40 people) indicated that the combination of an NSAID with a muscle relaxant was better than the NSAID alone, although there were no statistically significant differences between groups (days to resolution of pain: 12.5 days with naproxen v 8.5 days with naproxen plus cyclobenzaprine; days to sit without pain: 7 days with naproxen v 5 days with naproxen plus cyclobenzaprine; further statistical analysis and P value not reported).[19]
NSAIDs versus heat wrap:
See benefits of temperature treatments.
NSAIDs versus back exercises:
See benefits of back exercises.
Harms
NSAIDs may cause gastrointestinal and other complications (see review on NSAIDs).
The first review reported that an earlier meta-analysis of controlled epidemiological studies concluded that ibuprofen was associated with the lowest relative risk of serious gastrointestinal complications, but this was mainly attributable to the relatively low doses prescribed (see NSAIDs versus placebo below).[19]
The review reported that for all RCTs included in the review, adverse effects were frequently reported; they included: abdominal pain, diarrhoea, oedema, dry mouth, rash, dizziness, headache, and tiredness. The review reported that according to the authors of the RCTs, most adverse effects were considered mild to moderately severe, but the sample sizes were relatively small, and therefore no clear conclusion could be drawn from these studies regarding the risks for gastrointestinal and other adverse effects of NSAIDs.[19]
NSAIDs versus placebo:
One systematic review of harms of NSAIDs found no significant difference in adverse effects between NSAIDs as a class and placebo (pooled OR for adverse effects v placebo 1.30, 95% CI 0.91 to 1.80).[22]The review reported that ibuprofen and diclofenac had the lowest gastrointestinal complication rate, mainly because of the low doses used in practice.
NSAIDs versus each other:
The review reported that there was no clear difference in the reported number or severity of adverse effects between the different types of NSAIDs.[19] Only one low-quality RCT reported statistically significant differences in adverse effects.
The additional RCT found that a similar proportion of people reported adverse effects in the acemetacin and diclofenac groups (30/94 [32%] with acemetacin v 39/100 [39%] with diclofenac; significance not assessed; P value not reported).[20] No other information on adverse effects was reported. One RCT included in the review found that similar proportions of people reported adverse effects in the diclofenac and ibuprofen groups (13% with diclofenac v 12% with ibuprofen; absolute numbers not reported; significance not assessed; P value not reported).[19]
The subsequent RCT reported that dexketoprofen was well tolerated, with a reported incidence of adverse events similar to that of diclofenac (percentage of participants with adverse effect: 27% with dexketoprofen v 31% with diclofenac; P = 0.43).[21] No serious adverse events were reported in either treatment group. Six people withdrew from the study as a result of adverse effects (dexketoprofen, 4 people: liver pain, hot flushes and headache, vomiting and stomach ache; diclofenac, 2 people: dyspepsia and nausea; statistical analysis between groups not reported). With regard to intensity, the RCT reported that adverse effects were mild or moderate in the majority of cases. Headache, pain at the site of injection, nausea, and abdominal pain were the most common adverse effects in both treatment groups.[21]
COX-2 NSAIDS:
The review noted that debates on the increased risk for cardiovascular events associated with the use of selective COX-2 inhibitors are important, and the increased risks of COX-2 NSAIDs (when compared with placebo) had been demonstrated in large RCTs.[19]These findings had resulted in the removal of some COX-2 inhibitors from the market. The review noted that the current question is whether the increased risk pertains to all selective COX-2 inhibitors and traditional NSAIDs too. It concluded that more data are needed from large epidemiological studies to answer this question and that such studies must consider the dose, frequency, and duration of NSAID intake. It reported that in many people with acute low back pain, the intake is of short duration and might not reach the level associated with increased cardiovascular risks.[19]
NSAIDs versus paracetamol (acetaminophen):
The review reported that NSAIDs were associated with significantly more adverse effects compared with paracetamol (proportion of people with adverse effects: 3 RCTs [2 RCTs may have included people with sciatica], 309 people; RR 1.76, 95% CI 1.12 to 2.76; P = 0.014).[19]
NSAIDs versus muscle relaxants:
The RCT found no significant difference between groups in the proportion of people with adverse effects (80 people; RR 0.59, 95% CI 0.30 to 1.18).[19]
NSAIDs versus non-drug treatments (physiotherapy or spinal manipulation):
The review gave no information on adverse effects for this comparison.[19]
NSAIDs versus NSAIDs plus adjuvant treatment:
The review reported that adverse effects were more frequent in the combination groups (4 people with naproxen v 12 people with naproxen plus cyclobenzaprine; further details not reported; statistical analysis between groups not reported).[19]
NSAIDs versus heat wrap:
See harms of temperature treatments.
NSAIDs versus back exercises:
See harms of back exercises.
Comment
Originally, COX-2 inhibitors, such as valdecoxib, were associated with fewer gastrointestinal adverse effects in osteoarthritic and rheumatoid arthritis studies;[23]however, they have been associated with serious cardiovascular adverse effects. Valdecoxib has been removed from the market in some countries because of concerns about its potential association with increased risk of MI and stroke.[24] Piroxicam is no longer recommended for the treatment of short-term painful and inflammatory conditions. Although piroxicam can still be used for the symptomatic relief of osteoarthritis, rheumatoid arthritis, and ankylosing spondylitis, it is not recommended as a first-line treatment for these conditions. Treatment should be used in the lowest dose (no more than 20 mg/day) and for the shortest duration possible. Piroxicam could be associated with a higher risk of gastrointestinal adverse effects and serious skin reactions compared with other non-selective NSAIDs.
Substantive changes
NSAIDs New evidence added.[10] [19] [21] Categorisation unchanged (Trade-off between benefits and harms).

Analgesics (paracetamol, opioids)
Summary
SYMPTOM IMPROVEMENT Compared with NSAIDs: We don't know whether paracetamol (acetaminophen) and NSAIDs differ in effectiveness at improving pain or global improvement in people with acute low back pain as we found insufficient evidence ( very low-quality evidence ). Compared with non-drug treatments: We don't know whether paracetamol or analgesics (not specified) are more effective than electroacupuncture or ultrasound treatment at relieving pain as we found insufficient evidence (very low-quality evidence). Analgesics alone compared with combination analgesics: Paracetamol plus tramadol may be no more effective at 10 days than tramadol alone at reducing pain intensity, but may cause fewer adverse effects (very low-quality evidence). Compared with heat wrap: Paracetamol (acetaminophen) may be less effective at 1 and 4 days at improving pain ( low-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with heat wrap: Paracetamol (acetaminophen) may be less effective at improving disability at 4 days (low-quality evidence). NOTE We found no direct information about whether analgesics (paracetamol, opioids) are better than no active treatment in people with acute low back pain.
Benefits
Analgesics versus placebo:
We found two reviews (search dates 1995[25] and 1997[19]), which identified no RCTs.
Paracetamol (acetaminophen) versus NSAIDs:
See benefits of NSAIDs.
Analgesics versus non-drug treatments:
We found one systematic review (search date 1995), which identified one RCT (40 people) that found that electroacupuncture significantly reduced pain after 6 weeks compared with paracetamol (change in pain scores from baseline [on a 100-point visual analogue scale]: from 54.4 to 13.7 with paracetamol v from 52.7 to 3.3 with electroacupuncture; reported as significant; P value not reported).[25] The review identified a second RCT (73 people), which found that ultrasound treatment significantly increased the proportion of people who were pain free after 4 weeks compared with analgesics (unspecified) (41% with ultrasound v 7% with analgesics; reported as significant; P value not reported).[25]
Combination analgesics versus analgesic alone:
We found one RCT (119 people with moderate-intensity low back pain [40 mm or more on a 100-mm visual analogue scale] for 10–42 days before enrolment), which compared 10 days' treatment with paracetamol 325 mg plus tramadol 37.5 mg versus tramadol 50 mg alone.[26] The RCT found no significant difference between groups in change in pain intensity after 10 days' treatment (measured on a 100-mm visual analogue scale: from 67.5 mm to 27.9 mm with combination v from 65.3 mm to 24.8 mm with tramadol alone; P = 0.455).
Analgesic versus heat wrap:
See benefits of temperature treatments.
Harms
See review on paracetamol (acetaminophen) poisoning. RCTs have found adverse effects (constipation and drowsiness) with analgesics in about 50% of people. One earlier systematic review (search date 1995) found that combinations of paracetamol plus weak opioids increased the risk of adverse effects compared with paracetamol alone (15 single-dose studies; OR 1.1, 95% CI 0.8 to 1.5; 3 multiple-dose studies; OR 2.5, 95% CI 1.5 to 4.2).
Analgesics versus placebo:
The reviews identified no RCTs.[25] [19]
Paracetamol (acetaminophen) versus NSAIDs:
See harms of NSAIDs.
Analgesics versus non-drug treatments:
The review gave no information on adverse effects for this comparison.[25]
Combination analgesics versus analgesic alone:
The RCT found that a significantly smaller proportion of people receiving combination treatment reported adverse effects compared with those receiving tramadol alone (30/59 [51%] with paracetamol plus tramadol v 44/60 [73%] with tramadol alone; P = 0.019).[26] The most common adverse effects reported were nausea, dizziness/vertigo, and sleepiness. The RCT found that the incidences of nausea and dizziness/vertigo were significantly lower in the combination group compared with the tramadol-alone group (nausea: 8/59 [14%] with combination v 21/60 [35%] with tramadol alone; P = 0.012; dizziness/vertigo: 3/59 [5%] with combination v 15/60 [25%] with tramadol alone; P = 0.006). However, there was no significant difference between groups in incidence of sleepiness (7/59 [12%] with combination v 15/60 [25%] with tramadol alone; P = 0.198).
Analgesic versus heat wrap:
See harms of temperature treatments.
Comment
None.
Substantive changes
Analgesics (paracetamol, opioids) New evidence added.[19] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Epidural corticosteroid injections
Summary
We found no clinically important results from RCTs about the effects of epidural corticosteroid injections in people with acute low back pain. Epidural corticosteroid injections have been associated with serious adverse effects.
Benefits
We found one systematic review (search date 1998), which identified no RCTs on the effects of epidural corticosteroid injections in people with acute low back pain without sciatica.[27]
Harms
We found no RCTs.
Comment
Clinical guide:
Epidural corticosteroid injections may be associated with serious adverse effects and should only be administered under specific indications. Epidural corticosteroid injections are only indicated for people who are on a waiting list for surgery with leg-dominant pain and root irritation. Epidurals are most effective for potential surgical candidates; however, even in such cases, epidural injections lead to only marginal benefit. Epidurals give a short period of improvement and are ineffective in the long term. Epidural corticosteroid injections are not effective for those with only acute back pain and no leg pain.
Substantive changes
No new evidence

Advice to stay active
Summary
SYMPTOM IMPROVEMENT Compared with bed rest: Advice to stay active is more effective at reducing pain at 3 to 12 weeks post episode ( moderate-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with no advice or traditional medical treatment (including analgesics as required, advice to rest, and "let pain be your guide"): Advice to stay active with or without other treatments may be more effective than traditional medical treatment alone (including analgesics as required, advice to rest, and "let pain be your guide") at reducing chronic disability at up to 1 year. However, evidence was weak and the size of effects was unclear ( very low-quality evidence ). Compared with bed rest: Advice to stay active is more effective at improving functional outcomes at 3 to 12 weeks post episode (moderate-quality evidence). RETURN TO WORK Compared with no advice or traditional medical treatment (including analgesics as required, advice to rest, and "let pain be your guide"): Advice to stay active with or without other treatments may be more effective than traditional medical treatment alone (including analgesics as required, advice to rest, and "let pain be your guide") at reducing sick leave. However, evidence was weak and the size of effects was unclear (very low-quality evidence). Compared with bed rest: Advice to stay active seems to be more effective than bed rest at reducing initial sick leave and sick leave at 3 to 4 weeks and 12 weeks in people with acute low back pain (moderate-quality evidence).
Benefits
We found one systematic review (search date 1996, 6 RCTs, 1957 people).[28]
Advice to stay active versus no advice or traditional medical treatment:
The review did not pool data, but reported consistent findings among included RCTs.[28] The review compared advice to stay active with or without other treatments versus those other treatments alone. The review found that advice to stay active significantly reduced sick leave (significance not assessed; reported as significant) and reduced chronic disability at up to 1 year compared with traditional medical treatment (including analgesics as required, advice to rest, and "let pain be your guide"). See comment.[28]
Advice to stay active versus bed rest:
See benefits of bed rest.
Harms
Advice to stay active versus no advice or traditional medical treatment:
The review[28] gave no information on adverse effects.
Advice to stay active versus bed rest:
See harms of bed rest.
Comment
Limitations in methods preclude meaningful quantification of effect sizes. Advice to stay active was provided either as a single treatment or in combination with other interventions such as back schools, a graded activity programme, or behavioural counselling.
The distinction between placebo effects and specific treatment effects is often ill-defined in non-pharmaceutical treatment trials. Thus, the selection of a comparison group often requires considerable thought to ensure that the placebo intervention does not share some of the specific therapeutic components of the experimental intervention. This issue is more of a concern when placebos are designed to resemble the experimental intervention. In some placebo-controlled trials, the placebo treatment is actually used in clinical practice as a treatment.[10]
Substantive changes
No new evidence

Multidisciplinary treatment programmes (acute low back pain)
Summary
SYMPTOM IMPROVEMENT Compared with usual care: We don't know whether graded activity is more effective at 26 weeks at reducing pain intensity as we found insufficient evidence ( very low-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with usual care: We don't know whether graded activity is more effective at improving functional status as we found insufficient evidence (very low-quality evidence). RETURN TO WORK Compared with usual care: People undergoing graded activity (even when combined with workplace intervention) may occasionally take longer to return to work (very low-quality evidence).
Benefits
We found one RCT assessing the effects of a multidisciplinary treatment programme in people with acute low back pain analysed in two publications.[29] [30] The RCT (196 people with low back pain who had been on sick leave for 2–6 weeks) randomised people initially to a workplace intervention (96 people) or usual care (100 people).[29] At 8 weeks after the start of the person's sick leave, people (112 people) underwent a second round of randomisation to either graded activity or usual care.
One report analysed the effects of the combination of graded activity plus workplace intervention (27 people) versus the effects of either treatment alone and usual care as a group (85 people): the study did not correct for the effects of the workplace intervention or graded activity in the control comparator group.[29] At 12 months' follow-up, the study found no significant difference in the number of days off work between groups receiving both the workplace intervention and graded activity compared with those receiving either treatment alone or usual care (median number of days off work: 143 with combined treatment v 126 without combined treatment; P = 0.49). The RCT also found no significant difference between groups in pain intensity and functional status (improvement in pain intensity [measured using a visual analogue scale, where 0 = no pain and 10 = severe pain]: 2.9 with combined treatment v 3.3 without combined treatment; improvement in functional status [measured using Roland Morris questionnaire]: 8.3 with combined treatment v 8.7 without combined treatment; number of people in analysis not reported; both comparisons reported as not significant; P values not reported).
The second analysis of this study assessed the effects of graded activity versus usual care.[30] At 26 weeks, the RCT found that people in the graded activity group had a small, but significant, worsening in pain intensity compared with the usual-care group (mean improvement from baseline on a 10-point visual analogue scale: 92 people analysed: 3.7 with graded activity v 3.2 with usual care; reported by the authors to be a significant difference in favour of usual care; P value not reported).[30] People undergoing graded activity took significantly longer to return to work compared with those receiving usual care (intention-to-treat analysis: median time taken to return to work: 139 days with graded activity v 111 days with usual care; HR 0.52, 95% CI 0.32 to 0.86; P <0.01). However, there was no significant difference between groups in functional status (mean improvement from baseline on Roland Morris questionnaire: 91 people analysed: 7.9 with graded activity v 7.5 with usual care; reported as not significant; P value not reported). The RCT reported that, of the 55 people assigned to graded activity, 27 received workplace intervention, and of the 57 assigned to usual care, 26 received the workplace intervention. Subgroup analysis of those who had not received workplace intervention (59 people) found no significant difference in median number of days taken to return to work between graded activity and usual care (HR 0.86, 95% CI 0.40 to 1.84; P = 0.69). The RCT did not carry out a subgroup analysis for those who received the workplace intervention. Graded activity comprised physiotherapist-supervised exercise programmes (26 sessions lasting 1 hour/week) emphasising return to work based on operant conditioning principles. The workplace intervention consisted of ergonomic workplace assessment, modifications plus case management, and additional treatments (physiotherapy, manual therapy, Cesar therapy, and chiropractor care). The results presented should be interpreted with caution. The number of people who received both the workplace intervention and graded activity is unclear. Of the 55 people randomised to graded activity, 19 did not receive the clinical intervention, and, of the 36 people receiving graded activity, it is unclear how many had previously received the workplace intervention and were followed up for 12 months.
Harms
The RCT gave no information on adverse effects.[29] [30]
Comment
There was a considerable time lag between randomisation and the start of the graded activity programme, which, together with poor compliance in this group, could explain the observed delay in return to work.[30]
Clinical guide:
Multidisciplinary rehabilitation programmes are typically expensive and may not be necessary for uncomplicated acute low back problems. Multidisciplinary programmes typically include treatments provided by two or more healthcare providers with different professional training to obtain different perspectives and approaches to recovery. The term multidisciplinary does not imply a mandatory roster of specialists and does not dictate the nature of the treatment.
Substantive changes
No new evidence

Multidisciplinary treatment programmes (subacute low back pain)
Summary
RETURN TO WORK Compared with usual care: Multidisciplinary treatment, with or without a workplace visit, may be more effective than usual care at reducing sick leave in people with subacute low back pain. However, evidence was weak and interventions varied between studies ( very low-quality evidence ).
Benefits
We found two systematic reviews (search date 2002, 2 RCTs, 233 people with subacute low back pain, duration between 4 weeks and 3 months;[31] 1998–2006, 2 RCTs, 928 people with subacute low back pain, duration 5–12 weeks).[32]
The first review found that multidisciplinary treatment, including a workplace visit, significantly reduced sick leave compared with usual care (time to return to work: 10 weeks with multidisciplinary treatment v 15 weeks with usual care in first RCT; RR for return to work rate 2.4, 95% CI 1.2 to 4.9 in second RCT).[31] However, both studies identified in the review were of low quality; methodological deficiencies included lack of blinding, reporting of co-interventions, and unclear reporting of loss to follow-up.
The second review included two studies excluded from the first review as not being multidisciplinary (see comments).[32] The first included RCT (457 people, low back pain 8–12 weeks) included in the review used an intervention of light mobilisation and individualised information on prognosis and activity in the setting of a university clinic, while the control was people in primary care.[32] The review reported that the intervention group had an earlier return to work at 1 year (OR 1.60; CI, P value, and absolute numbers not reported), but differences between groups diminished over the second year (reported as not significant; P value not reported). In the third year, sick leave was 127.7 days with the intervention compared with 169.6 days with control (statistical analysis between groups not reported). In the second included study (489 people, described as subchronic low back pain, initial sick leave of 5–11 weeks), consecutive people were assigned by alternate allocation to intervention or control. Control was not specified in the review. The intervention group received a light mobilisation programme based on education and advice, and monitoring of conventional treatment. The review reported that the intervention significantly improved return to work compared with control at 5 years (based on data from an insurance office: 81% with intervention v 66% with control; absolute numbers not reported; P <0.001). The review reported that during the follow-up period, 72% of people in the intervention group and 74% of people in the control group had sickness absence because of low back pain (statistical analysis between groups not reported). However, this study was by alternate allocation and the results should be interpreted with caution.[32]
Harms
The reviews gave no information on adverse effects.[31] [32]
Comment
The first review included inpatient and outpatient programmes that were multidisciplinary.[31] To be multidisciplinary they had to consist of a physician's consultation plus either a psychological, social, or vocational intervention, or any combination. Trials in which rehabilitation was exclusively or predominantly medical were excluded, and back schools were also excluded from the review.[31] However, multidisciplinary programmes do not always include a psychosocial aspect as is evident in the second review.[32] The second review defined multidisciplinary interventions as those involving two or more healthcare disciplines.[32]
Clinical guide:
Multidisciplinary rehabilitation programmes are typically expensive and may not be necessary for uncomplicated acute low back problems. Multidisciplinary programmes are typically taken to comprise treatments provided by two or more healthcare providers with different professional training to obtain different perspectives and approaches to recovery. The term multidisciplinary does not imply a mandatory roster of specialists and does not dictate the nature of the treatment.
Substantive changes
Multidisciplinary treatment programmes (subacute low back pain) New evidence added.[32] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Spinal manipulation
Summary
SYMPTOM IMPROVEMENT Compared with placebo or sham treatment: Spinal manipulative therapy may be more effective than sham therapy at reducing pain at 6 weeks. However, results were inconsistent between studies and varied by the analysis undertaken. We don't know whether spinal manipulative therapy is more effective than placebo or sham therapy at reducing pain in the longer term ( very low-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with placebo or sham treatment: We don't know whether spinal manipulation and chiropractic adjustment are more effective than placebo or sham therapy at improving disability in either the short or long term ( low-quality evidence ). Compared with specific back exercise: Spinal manipulation may be less likely than McKenzie treatment to increase disability at 5 days and at 4 weeks (low-quality evidence).
Benefits
We found two systematic reviews (search dates 2000, 39 RCTs;[33] 2006, 4 RCTs, 149 people[10]) and one subsequent RCT.[34]
Spinal manipulation versus placebo or sham treatment:
The first review found that spinal manipulative therapy significantly reduced pain in the short term (<6 weeks) compared with sham therapy, but found no significant difference in the longer term (short-term difference in pain on 100-mm visual analogue scale: 10 mm, 95% CI 2 mm to 17 mm).[33] The review found no significant difference in either short-term or long-term disability between groups (difference in disability on Roland Disability Questionnaire: short term: +2.8 mm, 95% CI –0.1 mm to +5.6 mm; long term: further data not reported).
The second review identified 4 RCTs (149 people) evaluating the analgesic effects of spinal manipulation estimated by placebo-controlled trials in people with non-specific low back pain.[10] In this review, RCTs using a placebo consisting of what might now be considered to be an active treatment were excluded (see muscle relaxants; see comment). The review extracted data on outcomes from the first assessment after the end of therapy, because the review considered that this time point was where the largest analgesic effects would be observed. It did not include a description of individual interventions used in each RCT. The review found no significant difference between spinal manipulation and placebo in pain (pain measured by analgesic efficacy [100-point scale]: 4 RCTs, 149 people; RR presented graphically; absolute numbers and figures for point estimate of RR and CI not reported; individual RCTs in analysis not reported; see comment).[10] One RCT (192 people) included in the second review compared chiropractic adjustments, muscle relaxants, and placebo, and found no significant difference in disability at 4 weeks among groups.[14]
The subsequent RCT (101 people) compared manipulation plus standard care (general advice and paracetamol, diclofenac, or dihydrocodeine as required, other treatments not allowed) versus standard care alone for 2 weeks.[34] Manipulation was initiated within 24 hours of randomisation, with people undergoing a maximum of 5 sessions within 2 weeks. Manipulation was performed by a specialist in manual medicine, chiropractice, and rheumatology; a specialist in physical medicine; or an osteopath. The RCT found no significant difference between groups in pain at 14 days or 6 months (pain measured by 11-point box scale [BS-11]; 14 days: difference +0.5, 95% CI –0.2 to + 1.2; P = 0.13; 6 months: difference +0.6, 95% CI –0.4 to +1.6; P = 0.22). It found no significant difference between groups in analgesic consumption at 14 days (calculated as diclofenac mg equivalents: difference –18, 95% CI –43 to +7; P = 0.17). There was no significant difference between the proportion of people who were pain free at 6 months (22/50 [44%] with manipulation plus standard care v 30/51 [59%] with standard care alone; difference −15%, 95% CI −34% to +4%; P = 0.17).[34]
Spinal manipulation versus other treatments:
The review found no significant difference in pain or function between spinal manipulative therapy and general-practitioner care, physiotherapy, exercises, or back school (results presented graphically).[33]
Spinal manipulation versus back exercises:
See benefits of back exercises.
Harms
Spinal manipulation versus placebo or sham treatment:
The systematic reviews gave no information on adverse effects.[33] [10] Another systematic review assessed harms of spinal manipulation.[35] In RCTs identified by the review that used a trained therapist to select people and perform spinal manipulation, the risk of serious complications was low, with an estimated risk of vertebrobasilar strokes of 1/20,000 to 1/1,000,000 people and risk of cauda equina syndrome of less than 1/1,000,000 manipulations.[36]
One systematic overview of non-pharmacological therapies reported that 5 systematic reviews consistently found that serious adverse events after spinal manipulation (such as worsening lumbar disc herniation or the cauda equina syndrome) were very rare.[37]
In the subsequent RCT, two serious adverse events occurred in the experimental group (4%) and two in the control group (4%).[34] In the manipulation group, there was one person with acute pancreatitis and one person with an acute loss of motor and sensory function of the left lumbar segment L5 due to a herniated disk after randomisation, but before any manipulation treatment was initiated. In the standard care group, one person had a symptomatic cholelithiasis and one person had a femoroacetabular impingement syndrome. The RCT reported that neither of these events seemed to be related to the allocated treatment strategies.[34]
Spinal manipulation versus other treatments:
The review gave no information on adverse effects related to treatment.[33]
Spinal manipulation versus back exercises:
See harms of back exercises.
Comment
Spinal manipulation is not advised in people with severe or progressive neurological deficit.[38] The first review included RCTs that compared manipulation or mobilisation for low back pain with another treatment or control (the review noted that manipulation differed from mobilisation in that manipulation focused on a different range of motion of the involved joint — the review reported that both hands-on treatments were included in the review).[33] The second review did not specify what was included under the term spinal manipulative therapy.[10]
The distinction between placebo effects and specific treatment effects is often ill-defined in non-pharmaceutical treatment trials. Thus, the selection of a comparison group often requires considerable thought to ensure that the placebo intervention does not share some of the specific therapeutic components of the experimental intervention. This issue is more of a concern when placebos are designed to resemble the experimental intervention. In some placebo-controlled trials, the placebo treatment is actually used in clinical practice as a treatment.[10]
The lack of identification of RCTs in each analysis in the second review and the lack of descriptions of trials are omissions that make interpretation of results very difficult.[10] On the basis of published studies, there is little evidence of benefit of spinal manipulation.
Substantive changes
Spinal manipulation One systematic review (149 people)[10] and one subsequent RCT (101 people)[34]added at this update. The review and RCT found no significant difference between spinal manipulation and placebo or usual care in pain. One further study added in harms which reports on adverse effects after spinal manipulation.[37] Categorisation of spinal manipulation changed from Likely to be beneficial to Unknown effectiveness.

Acupuncture
Summary
SYMPTOM IMPROVEMENT Compared with sham needling or other treatments: We don't know whether acupuncture is more effective at reducing pain as we found insufficient evidence ( very low-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with sham needling or other treatments: We don't know whether acupuncture is more effective at improving functional status as we found insufficient evidence (very low-quality evidence).
Benefits
We found one systematic review (search date 2003; see comment) that identified three RCTs of acupuncture in people with acute low back pain.[39] The review did not pool data. The first included RCT (40 people) found no significant difference in pain or function (measured immediately after the session) between one session of acupuncture on the SI3 acupoint bilaterally, and sham needling of the same point (see comment). The second included RCT (60 people) found no significant difference in pain between acupuncture and naproxen. The third identified RCT (100 people with low back pain, 5 days to 6 months' duration, worse in cold or rainy weather), which was of poor methodological quality, compared acupuncture plus moxibustion (burning a herb at the end of the needle) plus Chinese herbal medicine versus Chinese herbal medicine alone, making it difficult to draw reliable conclusions on the effects of acupuncture alone.
Harms
One systematic review (search date 1996) found that serious, rare, adverse effects included infections (HIV, hepatitis, bacterial endocarditis) and visceral trauma (pneumothorax, cardiac tamponade).[40]
Comment
The first included RCT was reported in abstract form only. The authors of the systematic review obtained additional material from the authors of the RCT.[39] The review concluded that, because of the small sample sizes and low methodological quality of the studies, the data did not allow firm conclusions about the effectiveness of acupuncture in acute low back pain.[39] Many studies of acupuncture are either non-English language papers (which we excluded) or were published in difficult-to-access journals, and thus were not available to review.
Substantive changes
No new evidence

Back schools
Summary
SYMPTOM IMPROVEMENT Compared with placebo or usual care: We don't know whether back schools are more effective at improving pain as we found insufficient evidence ( very low-quality evidence ). FUNCTIONAL IMPROVEMENT Back schools plus usual treatment compared with usual treatment alone: Back schools plus usual treatment may be no more effective at improving functional status (very low-quality evidence). RETURN TO WORK Compared with placebo or usual care: We don't know whether back schools are more effective at reducing sick leave as we found insufficient evidence (very low-quality evidence).
Benefits
We found one systematic review (search date 2003, 4 RCTs, see comment below)[41]and one subsequent RCT.[42]
The review did not pool data because of data deficiencies and heterogeneity of trial design.[41] The review assessed the quality of included RCTs against standard criteria and categorised them as being of higher or lower methodological quality (high quality: score of 6 or more on a methodological scale of 0–10). One low-quality RCT (217 people working in a car factory, pain with or without radiation; see comment) identified by the review compared back school, combined physiotherapy (including manual therapy), and placebo (short waves at the lowest intensity). The review found that back school significantly reduced the duration of sick leave compared with placebo (mean days until recovery: 14.8 with back school v 28.7 with placebo; median days of absence from work: 20.5 with back school v 26.5 with placebo; P value not reported). It found no significant difference between groups in pain at 6 weeks or recurrences during 1 year (P values not reported). A second high-quality RCT (170 people attending a private outpatient clinic, reporting inability to work and receiving compensation) identified by the review compared back school plus usual treatment versus usual treatment alone (including rest, analgesics, NSAIDs as appropriate, daily physiotherapy) and measured outcomes at 8 weeks, 6 months, and 12 months. The review found no significant differences between groups in pain, functional status, median time to return to work, or compensated recurrences over 1 year. A third low-quality RCT (56 people attending a general practitioner, in pain with or without radiation to the thigh; see comment) identified by the review compared back school versus a control treatment (advice not to strain the back, analgesics when required). The review found no significant difference between groups in the proportion of people pain free at 1, 3, or 6 weeks. The fourth high-quality RCT (975 people referred to a spine clinic, on sick leave from work for 8–12 weeks, in pain with or without radiation; see comment) identified by the review compared back school versus usual care. The review found that back school significantly reduced sick leave compared with usual care at 200 days and 5 years (200 days: 30% with back school v 60% with usual care; 5 years: 19% with back school v 34% with usual care; P values not reported).
In the subsequent RCT (220 people), people were randomly assigned to receive a behavioural videotape or a control videotape (each about 20 minutes in length).[42] The behavioural video indicated that people with acute low back pain may require several weeks of decreased activity while healing. However, after this period, moderate activity and exercises were the best treatment. Return to work as soon as possible was also emphasised. The control video described proper technique and posture when performing daily activities but did not target beliefs or self-management skills. Other than the videotape, usual care was provided to each person. At 12 months' follow-up, the RCT found no significant differences between groups in outcomes (as measured by short form-36 scores [P value not reported], Oswestry Disability Index [P = 0.36], fear avoidance beliefs [FABs] questionnaire [P = 0.32], Pain and Impairment Relationship Scale [P = 0.70], or Spielberger's State-Trait Anxiety Inventory [P =  0.79]). The behavioural participants viewed the videotape significantly more frequently during the final 3 months of the study than did controls (0.86 times with behavioural video v 0.55 times with control video; P = 0.008).[42] However, results were based on 111/220 (55%) people initially randomised, so results should be interpreted with caution.
Harms
The review and subsequent RCT gave no information on adverse effects.[41] [42]
Comment
The systematic review included RCTs in which a back-school type intervention was included. A back school was defined as consisting of an educational and skills-acquisition programme, including exercises, in which all lessons were given to groups of people and supervised by a paramedical therapist or medical specialist.[41] The back-school programmes in the 4 included RCTs varied considerably between trials, as did the included populations, making generalisations difficult. Three RCTs included people with radiating back pain (not further defined), but subgroup analysis of back pain without radiation was not possible.[41] With the explosion in the ways in which information can be disseminated, formal back schools have become far less common than they were previously. The emphasis currently focuses more on general education, often through less-traditional methods such as the internet. In a future update, we will include education on low back pain as a separate intervention.
The distinction between placebo effects and specific treatment effects is often ill-defined in non-pharmaceutical treatment trials. Thus, the selection of a comparison group often requires considerable thought to ensure that the placebo intervention does not share some of the specific therapeutic components of the experimental intervention. This issue is more of a concern when placebos are designed to resemble the experimental intervention. In some placebo-controlled trials, the placebo treatment is actually used in clinical practice as a treatment.[10]
Substantive changes
Back schools New evidence added.[42] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Behavioural therapy
Summary
SYMPTOM IMPROVEMENT CBT compared with usual care: We don't know whether CBT is more effective than traditional care (analgesics plus back exercises until pain subsides) at reducing low back pain at 9 to 12 months as we found insufficient evidence ( very low-quality evidence ). CBT plus generic back exercise compared with no exercise or CBT alone: CBT plus neuromuscular training may be more effective at reducing pain intensity at 7 days ( low-quality evidence ). CBT plus back exercise compared with back exercise alone: We don't know whether CBT plus generic back exercise is more effective than generic back exercise alone at improving pain or severity at up to 6 months' follow-up in people with subacute low back pain (low-quality evidence). FUNCTIONAL IMPROVEMENT CBT compared with usual care: We don't know whether CBT is more effective than traditional care (analgesics plus back exercises until pain subsides) at improving perceived disability at 9 to 12 months as we found insufficient evidence (very low-quality evidence). CBT plus generic back exercise compared with no exercise or CBT alone: CBT plus neuromuscular training may be no more effective at improving disability at 12 months' follow-up (low-quality evidence).
Benefits
CBT versus usual care:
We found one systematic review (search date 1995, 1 RCT, 107 people).[25] The poor-quality RCT identified by the review found that CBT significantly reduced pain and perceived disability compared with traditional care (analgesics plus back exercises until pain had subsided) at 9 to 12 months' follow-up (mean score on pain drawing: 1.98 with CBT v 3.06 with control; mean claimed impairment: 4.84 with CBT v 6.25 with control; scales not defined; P values not reported).[25]
CBT plus back exercise versus no exercise or versus CBT:
See benefits of back exercises.
CBT plus back exercise verses exercise alone:
See benefits of back exercises.
Harms
CBT versus usual care:
The review did not report on harms.[25]
CBT plus back exercise versus no exercise or versus CBT:
See harms of back exercises.
CBT plus back exercise verses exercise alone:
See harms of back exercises.
Comment
None.
Substantive changes
Behavioural therapy New evidence added.[43] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Electromyographic biofeedback
Summary
We found no direct information from RCTs about the effects of electromyographic biofeedback in people with acute low back pain.
Benefits
We found no systematic review or RCTs of electromyographic biofeedback in people with acute low back pain.
Harms
We found no evidence on harms.
Comment
None.
Substantive changes
No new evidence

Lumbar supports
Summary
SYMPTOM IMPROVEMENT Compared with no lumbar support: An elastic lumbar belt may be more effective than no belt at improving pain (measured by visual analogue scale) at 30 and 90 days in people with subacute low back pain lasting 1 to 3 months. However, evidence was weak ( very low-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with no lumbar support: An elastic lumbar belt may be more effective than no belt at improving functional capacity (measured by EIFEL score) at 30 and 90 days in people with subacute low back pain lasting 1 to 3 months. However, evidence was weak (very low-quality evidence). NOTE We found no evidence from RCTs in people with <30 days of low back pain.
Benefits
We found one RCT (197 people with subacute low back pain, episode lasting 1−3 months), which compared an elastic lumbar belt (with crossed bands and posterior metallic reinforcement) versus control (no belt).[44] The RCT found that the lumbar belt significantly improved functional capacity (as measured by the EIFEL score) at 30 and 90 days (reduction from baseline, score 0−24; 30 days: 5.4 with belt v 4.0 with no belt; P = 0.022; 90 days: 7.6 with belt v 6.1 with no belt; P = 0.023). The RCT found that the lumbar belt significantly improved pain (as measured by a visual analogue scale [VAS]) at 30 and 90 days (reduction from baseline, VAS 0−100 mm; 30 days: 26.8 with belt v 21.3 with no belt; P = 0.038; 90 days: 41.5 with belt v 32.0 with no belt; P = 0.002). It also found a significant reduction in the proportion of people who took medication during the trial (no medication: 60.8% with belt v 40% with no belt; P = 0.029; absolute numbers not reported). The RCT was not blinded and people in the control group were asked not to purchase a lumbar belt or to wear a lumbar belt during the study.[44]
Harms
Harms associated with prolonged lumbar support use include possible muscle atrophy, sweating, and discomfort. The RCT did not report on adverse effects.[44]
Comment
Supportive studies for usage usually address comfort over biomechanical advantage or protection.
Substantive changes
Lumbar supports New evidence added.[44] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Summary
SYMPTOM IMPROVEMENT Compared with placebo massage, sham massage, no massage, or usual care: Massage (with or without usual care) may be more effective than usual care alone or placebo massage at improving pain at short-term follow-up (1 week) in people with acute low back pain, but we don't know about at longer term follow-up, and evidence was weak ( very low-quality evidence ). Specific back exercise compared with passive treatments: A combined analysis of educational booklets, bed rest, ice packs, and massage may be less effective at 7 days than McKenzie treatment at reducing pain ( low-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with placebo massage, sham massage, no massage, or usual care: We don't know whether massage is more effective at improving function at short-term follow-up (1 week) in people with acute low back pain (very low-quality evidence). Specific back exercise compared with passive treatments: A combined analysis of educational booklets, bed rest, ice packs, and massage may be less effective than McKenzie treatment at 7 days but not at 4 weeks at reducing disability (low-quality evidence).
Benefits
Massage versus placebo massage, sham massage, no massage, or usual care:
We found one systematic review (search date 2008, 13 RCTs).[45] The review included massage in both acute and chronic low back pain. Of the 13 RCTs included in the review, two RCTs met our inclusion criteria. The types of massage technique, duration, and frequency of treatments varied among the studies.
The first three-armed RCT (60 people) included in the review compared massage (20 people) versus an inert control group comprising placebo massage therapy (20 people) and waiting list control (20 people). Massage was applied with a mechanical device (one 30-minute session of deep cross-friction massage with the aid of a copper myofascial T-bar [roptrotherapy] applied to the lumbar pelvic region). The review reported that massage was significantly better than placebo and no treatment (waiting list) for reducing pain and improving function at 1 week (mean pain measured by 100-mm visual analogue scale [VAS]; massage v placebo massage: SMD −1.08, 95% CI −1.74 to −0.41; massage v waiting list: reported as significant; P value not reported; mean function measured by Oswestry Disability Index; massage v placebo massage: SMD −2.52, 95% CI −3.37 to −1.67; massage v waiting list: reported as significant; P value not reported).[45] The review noted that allocation sequence generation and concealment were unclear and the study was not blinded.
The second RCT (61 people) included in the review compared the addition of massage to usual care (not described in the RCT) versus usual care alone. Massage was acupressure with a specific oil for 8 sessions with relaxation with a digital electronic muscle stimulator on acupoints before the massage. The RCT found that massage plus usual care significantly improved pain at short-term follow-up but not function (pain measured by VAS, mean difference −0.38, 95% CI −0.54 to −0.22; function measured by range of measures such as flexion, walking time, daily activities, mean difference −0.10, 95% CI −0.21 to +0.01).[45] The review reported that the acupuncture massage group had 39% greater reduction in pain intensity than the usual care group at 1 week after the end of treatment (P = 0.0001). It reported that electrical stimulation on acupuncture points followed by acupressure with aromatic lavender oil had no significant effects on housework, work, or leisure time. The review reported that allocation concealment was not clear, participants and carers were not blinded to intervention and assessment, co-interventions were not described, and 16% of people were lost to follow-up.
Massage versus back exercises:
See benefits of back exercises.
Harms
Massage versus placebo massage, sham massage, no massage, or usual care:
The review gave no information on adverse effects.[45]
Massage versus back exercises:
See harms of back exercises.
Comment
The review defined massage as soft tissue manipulation using the hands or a mechanical device (examples include Shiatsu, Rolfing [soft tissue manipulation], Swedish massage, reflexology, craniosacral therapy, as part of physiotherapy, copper myofascial T-bar).[45] Massage could be applied to any body part (lumbar region only or to the whole body) and any technique could be used (e.g., cyriax, friction, kneading, and hacking).
Substantive changes
Massage New evidence added.[45] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Temperature treatments (short-wave diathermy, ultrasound, ice, and heat)
Summary
SYMPTOM IMPROVEMENT Heat wrap compared with placebo or non-heated wrap: Heat wrap is more effective at improving pain relief at 5 days ( moderate-quality evidence ). Heat wrap compared with oral analgesic: Heat wraps may be more effective than paracetamol (acetaminophen) at improving pain at 1 to 4 days ( low-quality evidence ). Heat wrap compared with NSAID (ibuprofen): Heat wraps may be more effective at improving pain at 1 and 4 days (low-quality evidence). Heat wrap plus NSAID compared with NSAID alone: Heat wrap plus NSAID (not specified; taken on an as required basis) may be more effective than NSAID alone (taken on an as required basis) at reducing pain at days 2, 3, and 4 post treatment. However, evidence was weak ( very low-quality evidence ). Heat wrap plus education compared with education alone: Heat wrap plus education may be more effective at reducing pain intensity but not pain relief at 14 days (low-quality evidence). Heat wrap alone compared with McKenzie treatment: We don't know whether heat wrap is more effective at relieving pain at 2 to 7 days as we found insufficient evidence (low-quality evidence). FUNCTIONAL IMPROVEMENT Heat wrap compared with placebo or non-heated wrap: Heat wrap is more effective at improving disability at 5 days (moderate-quality evidence). Heat wrap compared with oral analgesic: Heat wraps may be more effective than paracetamol (acetaminophen) at improving disability at 4 days (low-quality evidence). Heat wrap compared with NSAID (ibuprofen): Heat wraps may be more effective at improving disability at 4 days (low-quality evidence). Heat wrap plus education compared with education alone: Heat wrap plus education may be more effective at improving disability at 14 days (low-quality evidence). Heat wrap alone compared with McKenzie treatment: We don't know whether heat wrap is more effective at improving function at 2 to 7 days as we found insufficient evidence (low-quality evidence).
Benefits
We found one systematic review (search date 2005, 5 RCTs, 856 people with acute or subacute low back pain)[46] and two subsequent RCTs[47] [48] assessing the effects of heat treatments on acute low back pain. The review reported that only a small proportion of the data were suitable for pooling (pooling was not possible for most outcomes and comparisons). We found no review or RCTs on the effects of short-wave diathermy, ultrasound, or cold therapies in people with acute low back pain.
Heat wrap versus placebo or non-heated wrap:
The review found that heat wrap therapy significantly improved pain relief, reduced pain, and improved disability at 5 days compared with placebo or non-heated wrap (pain relief [scale range 0–5, higher score favours heat]: 2 RCTs, 258 people; WMD 1.06, 95% CI 0.68 to 1.45; pain [measured using a visual analogue scale, range 0–100, lower score favours heat]: 1 RCT, 90 people; WMD –32.20, 95% CI –38.69 to –25.71; disability [measured using Roland Morris questionnaire, lower score favours heat]: 2 RCTs, 258 people; WMD –2.12, 95% CI –3.07 to –1.18).[46]
Heat wrap versus paracetamol (acetaminophen):
The review found that heat wrap significantly improved pain relief at both 1 and 4 days' treatment, and improved disability at 4 days' treatment compared with paracetamol (acetaminophen) (1 RCT, 226 people; pain relief at 1 day: WMD 0.90, 95% CI 0.50 to 1.30; pain relief at 4 days: WMD 0.74, 95% CI 0.31 to 1.17; disability at 4 days: WMD 2.00, 95% CI 0.86 to 3.14).[46]
Heat wrap versus NSAID (ibuprofen):
The review found that heat wrap significantly improved pain relief at both 1 and 4 days' treatment and improved disability at 4 days' treatment compared with ibuprofen (1 RCT, 226 people; pain relief at 1 day: WMD 0.65, 95% CI 0.25 to 1.05; pain relief at 4 days: WMD 1.05, 95% CI 0.62 to 1.48; disability at 4 days: WMD 2.20, 95% CI 1.11 to 3.29).[46]
Heat wrap plus NSAID versus NSAID alone:
The second subsequent RCT (30 people) compared heat wrap therapy plus oral analgesics (NSAIDs; not specified) as required versus control (oral analgesics [NSAIDs] as required alone).[48] The heat wrap therapy had to be used once a day for at least 4 hours on 4 consecutive days. The heat-delivering device was a single-use commercial product designed like a belt, which wrapped around the stomach and delivered heat to the low back area, providing at least 8 hours of continuous heat at a constant temperature of 40 °C. Both groups were investigated on 5 successive days. Pain was assessed as part of an overall questionnaire that was based on other published validated questionnaires. The RCT found that the heat wrap significantly reduced pain in the evening of the third day (P = 0.20) and fourth day (P = 0.042) and significantly reduced the proportion of people who woke with pain on the second day (P = 0.016), third day (P = 0.002), and fourth day (P = 0.001; all results presented graphically; absolute numbers not reported). The RCT reported that of 38 people enrolled in the study, only 30 completed all measurements. It was unclear whether the additional 8 participants had been initially randomised. The RCT reported that assignment to the two groups followed a randomisation list in the order that people were recruited by the orthopaedist; no further details were reported.[48]
Heat wrap plus education versus education alone:
The first subsequent RCT (43 people) compared topical heat wrap (worn during daytime hours for 3 consecutive days) plus education versus education alone.[47] At 14 days after initial treatment, the RCT found that combined treatment of heat wrap plus education significantly reduced pain intensity and significantly improved disability compared with education alone (difference between groups adjusted for sex, age, baseline pain intensity, and pain medication use: pain intensity: –1.75, 95% CI –3.33 to –0.18; P = 0.030: disability: –4.33, 95% CI –8.41 to –0.27; P = 0.038). The RCT found that heat wrap plus education significantly increased pain relief at 4 days compared with education alone, but the difference between groups was not significant at 14 days (difference between groups adjusted for sex, age, baseline pain intensity, and pain medication use: 4 days: 1.13, 95% CI 0.11 to 2.14; P = 0.03: 14 days: +0.80, 95% CI –0.33 to +1.93; P = 0.157). Pain intensity was measured as a change in a visual analogue scale (VAS [where 0 = no pain and 10 = pain as bad as it could be]) and disability was measured using the Roland Morris questionnaire. Education comprised distribution of written material describing low back pain covering, for example, the recognition and treatment of symptoms.
Heat wrap alone versus McKenzie treatment:
The review found no significant difference between heat wrap and McKenzie treatment in pain relief or function at 2 or 7 days' follow-up (1 RCT, 50 people; pain relief [higher score favours heat]; 2 days: 1.40 with heat wrap v 1.00 with McKenzie treatment; WMD +0.40, 95% CI –0.15 to +0.95; 7 days: 2.30 with heat wrap v 2.00 with McKenzie treatment; WMD +0.30, 95% CI –0.68 to +1.28; function: 2 days: –0.90 with heat wrap v –0.20 with McKenzie treatment; WMD –0.70, 95% CI –2.09 to +0.69; 7 days: –2.80 with heat wrap v –2.30 with McKenzie treatment; WMD –0.50, 95% CI –2.72 to +1.72).[46]
Harms
Heat wrap versus placebo or non-heated wrap:
The review reported that skin pinkness, which resolved quickly, was reported as an adverse effect of heat wrap therapy.[46] The first subsequent RCT reported that no serious adverse effects were associated with heat wrap treatment.[47]
Heat wrap versus paracetamol (acetaminophen):
For harms associated with paracetamol, see harms of analgesics.
Heat wrap versus NSAID (ibuprofen):
For harms associated with ibuprofen, see harms of NSAIDs.
Heat wrap plus NSAID versus NSAID alone:
The RCT gave no information on adverse effects.[48]
Heat wrap plus education versus education alone:
The RCT gave no information on adverse effects.[47]
Heat wrap alone versus McKenzie treatment:
The review gave no information on specific adverse effects for this comparison.[46]
Comment
Of the 5 RCTs identified in the review, one was in people with acute low back pain, and 4 were in people with subacute low back pain.[46] Four RCTs declared receipt of industry funding.
Substantive changes
Temperature treatments (short-wave diathermy, ultrasound, ice, and heat) New evidence added.[48] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Summary
We found no clinically important results from RCTs about the effects of traction in people with acute low back pain.
Benefits
We found no systematic review or RCTs of traction in people with acute low back pain.
Harms
We found no RCTs.
Comment
None.
Substantive changes
No new evidence

Summary
SYMPTOM IMPROVEMENT Compared with placebo: We don't know whether TENS is more effective than placebo at improving pain in people with acute low back pain ( very low-quality evidence ).
Benefits
TENS versus placebo:
We found one systematic review (search date 2006),[10] which identified two RCTs reporting on the analgesic effects of TENS estimated by placebo-controlled trials in people with non-specific acute low back pain. In this review, RCTs using a placebo consisting of what might now be considered to be an active treatment were excluded (see benefits of muscle relaxants). The review extracted data on outcome from the first assessment after the end of therapy, because the review considered that this time point was where the largest analgesic effects would be observed. It did not include a description of individual interventions used in each RCT. The review found no significant difference between TENS and placebo in pain (measured by analgesic efficacy [100-point scale]) compared with placebo (2 RCTs, 121 people; RR presented graphically; absolute numbers and figures for point estimate of RR and CI not reported; individual RCTs in analysis not reported).
Harms
TENS versus placebo:
The review did not report on harms.[10]
Comment
None.
Substantive changes
TENS New evidence added.[10] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Back exercises
Summary
SYMPTOM IMPROVEMENT Generic back exercise compared with usual care or no treatment (acute low back pain of <6 weeks' duration): We don't know whether exercise is more effective at improving pain as we found insufficient evidence ( very low-quality evidence ). Generic back exercise compared with non-exercise interventions (acute and subacute low back pain): We don't know whether exercise is more effective at improving pain as we found insufficient evidence ( low-quality evidence ). Generic back exercise plus CBT compared with no exercise or CBT alone: Neuromuscular training plus CBT may be more effective at reducing pain intensity at 7 days (low-quality evidence). Generic back exercise compared with CBT plus exercise: We don't know whether generic back exercise plus CBT is more effective than generic back exercise alone at improving pain or severity at up to 6 months' follow-up in people with subacute low back pain (low-quality evidence). Specific back exercise compared with passive treatments: McKenzie treatment may be more effective than a combined analysis of educational booklets, bed rest, ice packs, and massage at reducing pain at 7 days (low-quality evidence). Specific back exercise compared with advice to stay active: McKenzie treatment is no more effective at reducing pain intensity at 12 weeks ( moderate-quality evidence ). Specific back exercise compared with flexion exercises: We don't know whether McKenzie treatment is more effective at reducing pain at 8 weeks as we found insufficient evidence (low-quality evidence). Specific back exercise compared with back school: McKenzie treatment may be more effective at improving pain at 1 year (low-quality evidence). FUNCTIONAL IMPROVEMENT Generic back exercise compared with usual care or no treatment (acute and subacute back pain): We don't know whether exercise is more effective at improving function as we found insufficient evidence (very low-quality evidence). Generic back exercise compared with non-exercise interventions (acute and subacute low back pain): We don't know whether exercise is more effective at improving function (low-quality evidence). Generic back exercise plus CBT compared with no exercise or CBT alone: Neuromuscular training plus CBT may be no more effective at improving disability at 12 months' follow-up (low-quality evidence). Specific back exercise compared with passive treatments: McKenzie treatment may be more effective at reducing disability at 7 days but not at 4 weeks compared with a combined analysis of educational booklets, bed rest, ice packs, and massage (low-quality evidence). Specific back exercise compared with advice to stay active: McKenzie treatment seems to increase disability at 12 weeks (moderate-quality evidence). Specific back exercise compared with flexion exercises: McKenzie treatment may be more effective at improving disability scores at 5 days (very low-quality evidence). Specific back exercise compared with spinal manipulation: McKenzie treatment may increase disability at 5 days and at 4 weeks (low-quality evidence). Specific back exercise compared with NSAIDs: We don't know whether McKenzie treatment is more effective at 3 months at improving short-term disability as we found insufficient evidence (low-quality evidence). RETURN TO WORK Generic back exercise compared with usual care or no treatment (subacute low back pain of 6–12 weeks' duration): We don't know whether exercise is more effective at reducing absenteeism in the work place or at reducing time taken to return to work as we found insufficient evidence (very low-quality evidence).
Benefits
We found three systematic reviews[49] [50] [51] and three subsequent RCTs.[52] [53] [43] The first review (search date 2004, 17 RCTs, see comment) included RCTs of back exercises versus placebo, no treatment, or other conservative treatments.[49] The second review (search date 2003, 6 RCTs, 518 people) included RCTs of McKenzie treatment versus passive treatment, advice to stay active, flexion exercises, spinal manipulation, back school, or strengthening.[50] The third review (3 RCTs, number of people included not clear) included RCTs of McKenzie treatment versus the NSAID ketoprofen, massage/advice, or passive movement/mobilisation.[51] The methodological quality of RCTs identified by the first review was assessed by the adequacy of 4 criteria: randomisation, allocation concealment, follow-up, and outcome blinding.[49] Studies were classed as high quality if they met all 4 criteria. The review identified 11 RCTs in people with acute back pain and 6 RCTs in people with subacute back pain; one RCT in each group was categorised as being of high quality. Methodological quality in the second and third reviews was based on the PEDro scale.[50] The second review identified 5 RCTs in people with acute low back pain, one RCT in people with subacute low back pain, and three RCTs in a mixed population of acute/subacute low back pain; all but one of the identified RCTs were high quality (score of 5+/10).[50] In the third review, two of the three RCTs identified were high quality (5+/10).[51] The first and second reviews identified 6 RCTs, one of which was also identified by the third review (see comment). The second and third reviews identified three RCTs assessing the effects of McKenzie treatment (see comment).[50] [51] All three reviews defined the included RCTs as either acute (<6 weeks' duration), subacute (6–12 weeks' duration), or duration not subgrouped (<12 weeks). The first review used both a qualitative rating system and a quantitative pooling of data where possible.[49] The second review pooled data (only statistically homogeneous RCTs) to compare the McKenzie treatment versus passive therapy (combined data on educational booklet, ice packs, massage, and bed rest) and advice to stay active (random effects model).[50] The third review transformed pain and disability scores to a score ranging from 0 to 100. To describe treatment effect for individual studies, mean and 95% confidence intervals were calculated for between-group differences (see comment).[51] The second review pooled data based on treatments, whereas the third review pooled data based on outcomes, and so, here, we report meta-analyses from only the second review.
Generic back exercise versus usual care or no treatment for acute low back pain (<6 weeks' duration):
The first review reported that 10 of 11 RCTs identified had non-exercise comparisons.[49] The review found no significant difference between generic exercise and no treatment in change in pain or function measured at the earliest follow-up (scale 0–100; pain: 3 RCTs, 491 people; WMD –0.59, 95% CI –12.9 to +11.51; function: 3 RCTs, 491 people; WMD –2.82, 95% CI –15.35 to +9.71; see comment). One high-quality RCT in an occupational setting found that mobilising home exercises were less effective than usual care, and one low-quality RCT in a healthcare setting found that a therapist-delivered endurance programme improved short-term functioning more than no treatment. Of the remaining 8 RCTs, 6 studies identified by the review found no statistically significant or clinically important difference between exercise therapy and usual care/no treatment, and the results of two RCTs were unclear.[49]
Generic back exercise versus usual care or no treatment for subacute low back pain (6–12 weeks' duration):
The first review reported that, in 6 included RCTs, 7 exercise groups (total number of exercise groups not reported) had a non-exercise comparison.[49] One high-quality and one low-quality RCT found that a graded exercise intervention reduced absenteeism outcomes in the workplace compared with usual care, and one low-quality RCT found improved functioning with exercise plus behavioural therapy compared with usual care. Two poor-quality RCTs found no difference in outcomes between exercise and the comparative treatments (including usual care), and one poor-quality RCT reported unclear results. One subsequent RCT (134 people with low back pain for at least 4 weeks before inclusion) compared graded exercise versus usual care.[52] The RCT found no significant difference in pain severity (11-point visual analogue scale [VAS]: 0 = no pain to 10 = very severe pain) or functional status (Roland Disability Questionnaire) between graded exercise and usual care, although there were greater improvements in both outcomes with graded exercise (between-group difference at 12 months: pain severity [favours graded exercise]: –0.2, 95% CI –1.2 to +0.8; P = 0.67: functional status [favours graded exercise]: –0.6, 95% CI –2.8 to +1.5; P = 0.56). The RCT found that people assigned to the graded-exercise group returned to work faster than those assigned to usual care (median duration of first continuous period of sick leave after randomisation: 54 days with graded activity v 67 days with usual care; significance not assessed). Graded exercise consisted of twice-weekly exercise sessions lasting 60 minutes each until the people either achieved full return to work, or the maximum therapy duration of 3 months had been completed.
Generic back exercise versus non-exercise interventions for acute low back pain (<6 weeks' duration):
The first review found no significant difference between exercise and other conservative treatments (advice to stay active, education, and usual care) in change in pain or function measured at the earliest follow-up (scale 0–100; pain: 7 RCTs, 606 people; WMD +0.31, 95% CI –0.10 to +0.72; function: 6 RCTs, 534 people; WMD –1.34, 95% CI –5.5 to +2.81).[49] Results were similar at intermediate and long-term follow-up.
Generic back exercise versus non-exercise interventions for subacute low back pain (6–12 weeks' duration):
The first review found no significant difference between exercise and all other comparisons (including no treatment, usual care, advice to stay active, and education) in change in pain or function measured at the earliest follow-up (scale 0–100; pain: 5 RCTs, 608 people; WMD –1.89, 95% CI –4.91 to +1.13; function: 4 RCTs, 579 people; WMD –1.07, 95% CI –5.32 to +3.18).[49] Results were similar at intermediate follow-up. The review concluded that there was insufficient evidence to support or refute the effectiveness of exercise for pain or function in subacute low back pain.
Generic back exercise plus CBT versus no exercise or CBT alone:
The second subsequent RCT (106 men with low back pain during the 3 months before study enrolment) compared neuromuscular training plus CBT versus no exercise or CBT.[53] At 12 months, the RCT found that neuromuscular training plus CBT significantly decreased pain intensity (visual analogue scale [VAS]) for the 7 days before assessment compared with no treatment (80 people; change in VAS from baseline: from 9.9 to 5.5 with neuromuscular training plus CBT v from 11.8 to 10.2 with no treatment; P = 0.032). There was no significant difference between groups in intensity of back pain for the 2 months before assessment, although a greater improvement in pain was reported by the group receiving neuromuscular training plus CBT (80 people; change in VAS from baseline: from 15.3 to 8.6 with neuromuscular training plus CBT v from 15.8 to 14.3 with no treatment; P = 0.052). The RCT found no significant difference between treatments in disability (Oswestry Disability Index [ODI]) at 12 months (84 people; change in ODI from baseline: from 5.6 to 4.8 with neuromuscular training plus CBT v from 5.8 to 5.0 with no treatment; P = 0.88). Neuromuscular training plus CBT consisted of neuromuscular training plus counselling with cognitive-behavioural goals for improved lumbar stability (2 sessions/week, one of which was physiotherapist-led and the other independent): the exercise programme consisted of 10 generic exercises.
Generic exercise versus CBT plus exercise:
The third subsequent RCT (47 people with subacute low back pain) compared a control group who only received physiotherapy in the form of generic exercises versus an intervention group who received the exercise treatment plus a CBT programme.[43] The CBT programme was conducted by a psychologist and took place independently of the physiotherapy treatment. For pain intensity (measured by 10-point VAS), there were no significant differences between groups after the last physiotherapy session (P = 0.81), 3 months later (P = 0.12), or 6 months later (P = 0.075). For overall severity (measured by the question "how severe do you think your back problems are?" and scored 1–4), there was no significant difference between groups after the last physiotherapy session (P = 0.22), or at 3 months (P = 0.98), although there was a significant difference between groups at 6 months (P = 0.004).[43]
Specific back exercise (McKenzie treatment) versus usual care or no treatment:
The reviews identified no RCTs for this comparison.[50] [51]
Specific back exercise (McKenzie treatment) versus passive treatments (combined analysis of educational booklets, bed rest, ice packs, and massage):
The second review (4 RCTs, 681 people) found that McKenzie treatment significantly decreased pain and disability at 1 week compared with passive therapy (combined data on educational booklets, bed rest, ice packs, and massage) (2 RCTs, 470 people; pain: WMD –4.16, 95% CI –7.12 to –1.20; disability: WMD –5.22, 95% CI –8.28 to –2.16; absolute numbers and P values not reported).[50] However, there was no significant difference between groups in disability at 4 weeks (3 RCTs, 495 people; WMD –1.06, 95% CI –3.21 to +1.10; absolute numbers and P value not reported).
Specific back exercise (McKenzie treatment) versus advice to stay active:
The second review found a significant increase in disability after 12 weeks' treatment with the McKenzie treatment compared with advice to stay active (2 RCTs, 261 people; WMD [0–100 point scale] 3.85, 95% CI 0.30 to 7.39; absolute numbers not reported; P value not reported).[50] There was no significant difference between groups in pain intensity at 12 weeks (WMD +5.02, 95% CI –1.19 to +11.22; absolute numbers not reported).
Specific back exercise (McKenzie treatment) versus flexion exercises:
The second review did not pool data for this comparison because of clinical and statistical heterogeneity among studies.[50] The review identified two RCTs that met Clinical Evidence inclusion criteria. One high-quality RCT (149 people with acute low back pain with or without radiation) identified by the review found no significant difference between treatment groups in pain at 8 weeks (data presented graphically; reported as not significant; P value not reported).[54] One low-quality RCT (24 people)[55] identified by the review[50] found a greater improvement in mean disability scores (ODI) at 5 days' follow-up with McKenzie treatment compared with flexion exercise (data presented graphically in the RCT; no further details reported: mean difference [0 to 100-point scale] between groups reported in the review: –22 points, 95% CI –26 points to –18 points).[50] [55]
Specific back exercise (McKenzie treatment) versus back school:
The second review identified one RCT (100 people with acute or subacute low back pain and with or without radiating pain) that met Clinical Evidence inclusion criteria.[50] The RCT found that McKenzie treatment decreased pain at 1 year compared with back school (absolute numbers not reported; P <0.001).[56] A 5-year follow-up study of the RCT identified by the review found that McKenzie treatment significantly decreased the proportion of people on sick leave at 5 years compared with back school (24/47 [51%] with McKenzie treatment v 31/42 [74%] with back school; P <0.03).[57]
Specific back exercise (McKenzie treatment) versus spinal manipulation:
The second review identified one high-quality RCT (24 people with acute or subacute low back pain[58]) that met Clinical Evidence inclusion criteria.[50] The RCT did not carry out a statistical analysis.[58] The review found a significant increase in disability (ODI) with McKenzie treatment at 5 days and 4 weeks compared with spinal manipulation (mean difference [0 to 100-point scale]; 5 days: 17 points, 95% CI 8 points to 27 points; 4 weeks; 22 points, 95% CI 10 points to 33 points).[50]
Specific back exercise (McKenzie treatment) versus NSAIDs:
The third review (1 RCT, 260 people) found no significant difference in short-term disability between McKenzie treatment and the NSAID ketoprofen (follow-up at <3 months), although results favoured McKenzie treatment (mean AR –4.2, 95% CI –9.8 to +1.4; absolute numbers not reported).[51]
Harms
Generic back exercise versus usual care or no treatment for acute low back pain:
The first review reported that few identified RCTs reported on harms (about 26% of RCTs).[49] Overall, in the review (which included RCTs on acute, subacute, and chronic low back pain), 12 RCTs reported mild negative reactions associated with the exercise programme, such as increased low back pain, and soreness in a minority of people;[49]although this is a natural and innocuous reaction, particularly in those starting an exercise programme for the first time or after prolonged inactivity. No further details were provided. The subsequent RCTs gave no information on adverse effects.[52] [53]
Generic back exercise versus usual care or no treatment for subacute low back pain:
See harms of back exercises versus usual care or no treatment for acute low back pain.
Generic back exercise versus non-exercise interventions for acute low back pain:
The review gave no information on adverse effects for this comparison (see harms of back exercises versus usual care or no treatment for acute low back pain).[49]
Generic back exercise versus non-exercise interventions for subacute low back pain:
The review gave no information on adverse effects for this comparison (see harms of back exercises versus usual care or no treatment for acute low back pain).[49]
Generic back exercises plus CBT versus no exercise or CBT alone:
The RCT gave no information on adverse effects.[53]
Generic exercise versus CBT plus exercise:
The RCT gave no information on adverse effects.[43]
Specific back exercise (McKenzie treatment) versus usual care or no treatment:
The reviews identified no RCTs for this comparison.[50] [51]
Specific back exercise (McKenzie treatment) versus passive treatments (combined analysis of educational booklets, bed rest, ice packs, and massage):
The review gave no information on adverse effects for this comparison.[50]
Specific back exercise (McKenzie treatment) versus advice to stay active:
The review gave no information on adverse effects for this comparison.[50]
Specific back exercise (McKenzie treatment) versus flexion exercises:
The review gave no information on adverse effects for this comparison.[50]
Specific back exercise (McKenzie treatment) versus back school:
The review gave no information on adverse effects for this comparison.[50]
Specific back exercise (McKenzie treatment) versus spinal manipulation:
The review gave no information on adverse effects for this comparison.[50]
Specific back exercise (McKenzie treatment) versus NSAIDs:
The review gave no information on adverse effects for this comparison (see review on NSAIDs).[51]
Comment
There was considerable variation in the exercise programmes undertaken in RCTs identified by the reviews. In the first review, subgroup meta-analysis for different specific types of exercise, or comparisons against specific individual conservative treatments were not reported.[49] The review included RCTs of exercise, this being defined as "a series of specific movements with the aim of training or developing the body by a routine practice or as physical training to promote good physical health". Individual RCT outcome data for pain and function were converted to a scale from 0 to 100 points to allow the pooling of data. The review considered that a 20-point (out of 100) improvement in pain and a 10-point (out of 100) improvement in functional outcomes were clinically important differences. The review categorised populations of included RCTs as being healthcare (primary, secondary, or tertiary), occupational (occupational healthcare, in compensatory situations), and general or mixed (e.g., people recruited through advertisement for trials), to differentiate those studies in people in typical treatment settings (healthcare, occupational) from those in people who may not normally present for treatment. The review noted that, overall, the methodological quality of included RCTs was poor, with only 54% adequately describing the exercise intervention. The second review concluded that, when evaluating treatment effects of individual RCTs, the McKenzie approach was as effective at all follow-up times as an educational booklet, advice to stay active, and strengthening exercises. Comparisons with flexion exercises and spinal manipulative therapy yielded statistically significant differences favouring McKenzie treatment; however, no placebo-controlled trial was identified.[50] In the first subsequent RCT, it is not clear which component of the complex intervention — the graded activity instruction, the exercises, or the combination of both modalities — is the most important. Because no placebo therapy was used, the attention of the therapist may have had a role in the positive effects.[52] A possible criticism of generic-exercise studies is that all patients in the exercise groups receive the same treatment, regardless of a patient's preference for extension or flexion exercises. According to the McKenzie system, this type of pre-selection is essential to determine a directional preference for certain exercises.
Clinical guide:
For specific exercises, there is a growing, but still limited, evidence for short-term pain reduction and increased function. Given the methodological flaws mentioned above, and the lack of relevant detail of the primary studies, it is not possible to either support or oppose the use of exercise in patients with acute low back pain.
Substantive changes
Back exercises New evidence added.[43] Categorisation unchanged (Unknown effectiveness), as there remains insufficient evidence to judge the effects of this intervention.

Summary
SYMPTOM IMPROVEMENT Compared with advice to stay active: Bed rest is less effective at reducing pain at 3 to 12 weeks post episode ( moderate-quality evidence ). FUNCTIONAL IMPROVEMENT Compared with advice to stay active: Bed rest is less effective at improving functional outcomes at 3 to 12 weeks post episode (moderate-quality evidence). Different lengths of bed rest compared with each other: Three days and 7 days of bed rest may be equally effective at reducing pain intensity ( low-quality evidence ). RETURN TO WORK Compared with advice to stay active: Bed rest seems less effective than advice to stay active at reducing initial sick leave and sick leave at 3 to 4 weeks and 12 weeks in people with acute low back pain (moderate-quality evidence). NOTE Bed rest has been associated with joint stiffness, muscle wasting, loss of bone mineral density, pressure sores, and venous thromboembolism.
Benefits
We found one systematic review (search date 2003, 11 RCTs, 1963 people; see comment).[59] The systematic review assessed the methodological quality of included RCTs against standard criteria and categorised them as being of low, moderate, or high risk of bias (see comment).[59]
Bed rest versus advice to stay active:
The systematic review included two RCTs at moderate/low risk of bias in a meta-analysis (see comment). The review found that advice to stay active significantly reduced pain and significantly improved functional status at 3 to 4 weeks' follow-up and at 12 weeks' follow-up compared with bed rest (pain: 2 RCTs, 400 people; 3–4 weeks: SMD 0.22, 95% CI 0.02 to 0.41; 12 weeks: SMD 0.25, 95% CI 0.05 to 0.45; functional status: 2 RCTs, 400 people; 3–4 weeks: SMD 0.29, 95% CI 0.09 to 0.49; 12 weeks: SMD 0.24, 95% CI 0.04 to 0.44).[59] The first RCT identified by the review found that advice to stay active significantly reduced sick leave at 3 to 4 weeks' follow-up and at 12 weeks' follow-up compared with bed rest (3–4 weeks: WMD 3.4 days, 95% CI 1.64 days to 5.16 days; 12 weeks: WMD 4.5 days, 95% CI 1.37 days to 7.63 days). The second RCT identified by the review found that bed rest increased initial sick leave compared with advice to stay active in people followed up at 12 weeks (86% with bed rest v 52% with advice to stay active; P <0.001).[59]
Different lengths of bed rest:
One included RCT (47 people) at low risk of bias found no significant difference in pain intensity between 3 days and 7 days of bed rest measured 2 days after the end of treatment.[59]
Bed rest versus exercise:
The review identified two RCTs at low risk of bias.[59] It reported that the first RCT found no significant difference between advice to rest in bed and exercise in pain or restrictions in activities of daily living at 6 weeks, 12 weeks, and 1 year of follow-up.[59] The review reported that the second RCT found no significant difference between advice to rest in bed and exercise in pain, functional status, or sick leave at 3 and 12 weeks' follow-up.[59]
Bed rest versus other treatments:
One included RCT at low risk of bias compared advice to rest in bed versus bed rest plus exercise plus education versus no instruction. The review found no significant difference in pain or restrictions of daily activities between any of the treatment groups (statistical analysis not reported).[59] The review reported that one other included RCT at high risk of bias found no difference in improvement on a combined pain, disability, and physical exam score between bed rest and manipulation, drug therapy, physiotherapy, back school, or placebo (statistical analysis not reported).[59]
Harms
The review gave no information on adverse effects.[59] One previous systematic review assessing harms[28] found that adverse effects of bed rest included joint stiffness, muscle wasting, loss of bone mineral density, pressure sores, and venous thromboembolism (see review on thromboembolism).
Comment
The review based classification of bias on 4 criteria: concealment of allocation, co-interventions, intention-to-treat analysis or losses to follow-up, and blinding of outcome assessor.[59] The review separately analysed: RCTs that included people with acute low back pain, with or without radiating pain, but excluded people with neurological deficits (called the acute simple low back pain group); RCTs that included people with verified neurological deficits (called the sciatica group); and RCTs that had included people with and without verified neurological deficits (called the mixed low back pain group).[59] We have only reported the results for the acute simple low back pain group here. However, within this group the proportion of people with radiating pain to the legs varied from none in some RCTs to 30% of the study population in others.
Bed rest versus advice to stay active:
In the analysis comparing advice to stay active versus bed rest for pain, one RCT that found significantly better pain outcomes for bed rest compared with advice to stay active was excluded from the meta-analysis: the RCT was categorised as being at high risk of bias, and the applicability of the included population (80 male combat trainees in an army hospital) to the general population was questionable.[59] This RCT also found that bed rest significantly reduced length of sick leave compared with advice to stay active.[59]
Substantive changes
No new evidence
Articles from Clinical Evidence are provided here courtesy of
BMJ Publishing Group