Benefits and harms of spinal manipulative therapy for the treatment of chronic low back pain: systematic review and meta-analysis of randomised controlled trials.

The short version

This systematic review and meta-analysis pooled 47 randomised controlled trials with 9211 adults to assess the benefits and harms of spinal manipulative therapy (SMT), hands-on manipulation or mobilisation of the spine, for chronic low back pain (pain lasting more than three months). SMT is widely used but is not currently a first-line recommendation, and earlier reviews gave mixed results. The review updated a previous one, added 21 new trials, and searched databases and trial registries up to May 2018. Trials compared SMT with recommended treatments such as exercise and pain medicines, with non-recommended ones such as light massage, waiting lists or electrotherapy, with sham (placebo) manipulation, and with SMT added to another therapy. Techniques were high-velocity thrusts, low-velocity passive movements, or both, delivered mostly by chiropractors, physical therapists and osteopaths. Two reviewers independently selected trials and judged risk of bias. Pain was converted to a 0-100 scale, function was pooled as standardised mean differences, and outcomes were examined at one, six and twelve months. Against recommended therapies, SMT gave similar short-term pain relief (-3.17 points, 95% CI -7.85 to 1.51; moderate-quality evidence) and a small, clinically better improvement in function (SMD -0.25, -0.41 to -0.09). Against non-recommended therapies, high-quality evidence showed small pain reductions (-7.48, -11.50 to -3.47) and small to moderate functional gains (SMD -0.41, -0.67 to -0.15) at one month, with similar results later. Evidence from sham-controlled trials was low to very low quality, so those effects are uncertain, and only 45% of trials had adequate randomisation and allocation concealment. High-velocity and low-velocity techniques did not differ. About half of the trials reported adverse events, mostly transient, mild to moderate musculoskeletal symptoms; one trial of 183 people found no increased risk versus sham (RR 1.24, 0.85 to 1.81), and one serious event was judged possibly related to SMT. The authors conclude that SMT produces similar effects to recommended therapies and seems better than non-recommended interventions for short-term function, and that clinicians should inform patients of possible adverse events. The authors note that safety data were limited because most trials did not register adverse events systematically. Limitations include unexplained heterogeneity, trials that varied widely in size (21 to 1334 participants), only five sham-controlled trials, weaker data on adjuvant use, and limited, inconsistently recorded information on adverse events.

Key findings

• This systematic review combined 47 randomised trials with 9,211 adults who had low back pain lasting more than three months to test spinal manipulative therapy (SMT, hands-on manipulation or mobilisation of the spine).
• SMT had similar effects on short-term pain as other recommended therapies such as exercise (mean difference -3.17 on a 0-100 scale, 95% CI -7.85 to 1.51) and a small, clinically better effect on function (standardised mean difference -0.25).
• Compared with non-recommended therapies, high-quality evidence showed small pain reductions (-7.48) and small to moderate functional improvement (standardised mean difference -0.41) at one month.
• Evidence for sham (placebo) manipulation was low to very low quality, so those effects are uncertain.
• About half of the trials assessed adverse events, mostly unsystematically; reported events were mainly transient musculoskeletal symptoms, and one serious event was judged possibly related to SMT.

Common questions