Table 2Included systematic reviews and meta-analyses of skeletal muscle relaxants in patients with musculoskeletal conditions

Author YearAimsTime period covered and sources used in literature searchEligibility criteriaExclusion criteriaFunding source and roleMethod of appraisalCharacteristics of identified articlesPopulation characteristicsMain ResultsAdverse eventsInternal validity
Systematic reviews
Schnitzer 200460Assess the efficacy and safety of low back pain medicationsThrough October 2002

Multiple databases including Medline, EMBASE, Cochrane
RCTs of low back pain in adults that used quantitative clinical endpoints of efficacy and/or safetyNot specifiedMerck & Company, New Jersey; role of funder not reportedAbstracted information on study characteristics, quality using Koes criteria (0–100)50 of 110 identified RCTs met inclusion criteria; 6 evaluated skeletal muscle relaxants

6 placebo-controlled trials (1 baclofen, 3 tizanidine, 1 chlormezanone, 1 tetrazepam)

931 patients included in 6 trials
Low back pain patients, age and severity varied between studiesOverall quality rated as 'moderate', limited evidence was found on effectiveness of drug treatments for low back pain and comparative assessments were not attempted. No head-to- head trials found.Insufficient data to adequately assess.GOOD.
Tofferi 200462Assess the efficacy and safety of cyclobenzaprine for fibromyalgiaThrough November 2000

Multiple databases including MEDLINE, EMBASE, DARE, Cochrane, Psyclit
Placebo-controlled RCTs with measurable outcomesNot specifiedNot reportedAbstracted information on study characteristics, quality using Jadad scale, efficacy and safety outcomes5 of 27 identified RCTs met inclusion criteria

5 placebo-controlled trials of cyclobenzaprine with 312 patients, longest 24 weeks
Low back pain patients, age and severity varied between studiesOverall quality of studies fair, with average quality score 4.4 (range 0–8).

Patients on cyclobenzaprine more likely to report themselves to be 'improved' (OR 3.0, 95% CI 1.6– 5.6), NNT 4.8.

Sleep improved similarly in cyclobenzaprine and placebo patients.

Pain improved in cyclobenzaprine patients at week 4 only (SMD 0.35).

No improvements in fatigue or tender points in cyclobenzarpine or placebo groups.
Not assessed.GOOD.
Van Tulder 200348, 49Systematic review of effectiveness of skeletal muscle relaxants in the treatment of back painthrough October 2001 (MEDLINE, EMBASE) or 2002 (Cochrane Library)

MEDLINE, Cochrane Library, EMBASE
Randomized controlled trials and double-blind controlled clinical trials of patients with nonspecific low back pain receiving skeletal muscle relaxants of benzodiazepenes, reporting specified outcome measuresStudies of chlormezanone and botulinum toxinUniversity of Toronto and VU University Medical Center AmsterdamIndependently assessed by two reviewers using criteria (11-item instrument) recommended by the Cochrane Back Review Group.27 studies excluded

30 trials of 2884 patients included (14 of these studies did not meet our inclusion criteria because they were non-English or evaluated excluded interventions)
Acute or chronic low back pain of varying degrees; age, race and gender reported for individual studiesAll studies had at least two criteria for which it was rated inadequate. Mean quality score 6 (range 3- 9, scale 0–11).

Nonbenzodiazepines versus placebo (11 studies, pooled relative risks)
Pain relief after 2 to 4 days: 0.80 (95% CI, 0.71– 0.89)
Global efficacy after 2 to 4 days: 0.49 (95% CI, 0.25–0.95)
Nonbenzodiazepines versus placebo (11 studies, pooled relative risks)
Overall adverse events: 1.50 (95% CI, 1.14–2.98)
Central nervous system adverse events: 2.04 (95% CI, 1.23–3.37)
GOOD.
Browning 200165Systematic review of cyclobenzaprine's effectiveness in the treatment of back pain1966–1999

MEDLINE, PsycLit, CINAHL, EMBASE, AIDSLINE, HEALTHSTAR, CANCERLIT, Micromedix, Cochrane Library and Cochrane Database of Systematic Reviewers, Federal Research in Progress, reference lists, pharmaceutical companies contacted
Randomized, placebo controlled, at least one group receiving cyclobenzaprine, and measurable outcomes reportedNot reportedNoneIndependently assessed by two reviewers using 6- item instrument7 trials excluded

14 randomized placebo- controlled trials of 3315 patients on cyclobenzaprine; 6 studies also had diazepam as a control, 1 diflunisal, and 1 methocarbamol
Acute back pain and muscle spasm of varying degrees; age, race, and gender not reportedAll studies had at least one problem with rated quality. Mean quality score 4.3 (scale 1–8).

Cyclobenzaprine vs. placebo: Global improvement (10 studies, pooled risk difference): 0.37 (95% CI, 0.24–0.50) No statistically different results (though trends favored cyclobenzaprine) for local pain, muscle spasm, tenderness to palpation, range of motion, and ADL at 3 days, 1 or 2 weeks.
Cyclobenzaprine vs. placebo (percentages)
Drowsiness: 20% vs. 2%, p<0.001
Dry mouth: 8% vs. 2%, p=0.02
Dizziness: 7% vs. 4%, p=0.04
Nausea: 2% vs. 2%, p=0.70
Any: 53% vs. 28%, p=0.002
GOOD.
Meta-analysis
Nibbelink 197870Assess the therapeutic response of cyclobenzaprine compared to diazepam and placeboTime period covered not clear

Not clear what methods used to identify relevant studies, but appears to include unpublished studies performed at Merck
Controlled clinical studies of patients with skeletal muscle spasm treated with cyclobenzaprine, diazepam, or placebo.Studies outside the United States (3 studies) because of differences in protocol and data collection.Authors employed by Merck. Not reported if funder held data.Not reported20 double-blind randomized trials of 1153 patients (434 cyclobenzaprine, 280 diazepam, 439 placebo)

46% posttraumatic, 14% musculoskeletal strain, 10% idiopathic, 8% postoperative, 6% osteoarthritis, 3% cervical root syndrome, 1% miscellaneous.
46% posttraumatic, 14% musculoskeletal strain, 10% idiopathic, 8% postoperative, 6% osteoarthritis, 3% cervical root syndrome, 1% miscellaneous.

Gender 535/1065 female, 186/1153 >50 years, race not reported
Cyclobenzaprine vs. diazepam vs. placebo
Global response: Cyclobenzaprine and diazepam significantly better than placebo, no significant differences between cyclobenzaprine and diazepam.

Cyclobenzaprine vs. diazepam (symptoms absent or mild at week 2)
Muscle spasms: 42% vs. 29% (p=0.035)
Local pain: 24% vs. 33% (NS)
Tenderness on palpation: 26% vs. 39% (p=0.044)
Limitation of motion: 30% vs. 50% (p=0.006)
Limitation of daily living: 31% vs. 48% (p=0.030)
Muscle relaxants 'similar in Cyclobenzaprine vs. diazepam vs. placebo
Drowsiness: 39% vs. 33% vs. 12%
Dry mouth: 24% vs. 8% vs. 4%
Ataxia/dizziness: 10% vs. 17% vs. 6%
Bad taste: 3% vs. 1% vs. 0.4%
Nausea: 2% vs. 1% vs. 3%
Withdrawals not reported for different interventions
FAIR. No evaluation for heterogeneity. Insufficient detail of included studies. Not clear if studies summarized appropriately: combined all individual patient data for comparisons between interventions.

From: Evidence Tables

Cover of Drug Class Review: Skeletal Muscle Relaxants
Drug Class Review: Skeletal Muscle Relaxants: Final Report [Internet].
Chou R, Peterson K.
Portland (OR): Oregon Health & Science University; 2005 May.
Copyright © 2005, Oregon Health & Science University, Portland, Oregon.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.