Objectives: Many interventions are available to manage chronic pain; understanding the durability of treatment effects may assist with treatment selection. We sought to assess which noninvasive nonpharmacological treatments for selected chronic pain conditions are associated with persistent improvement in function and pain outcomes at least 1 month after the completion of treatment.

Data sources: Electronic databases (Ovid MEDLINE®, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews) through November 2017, reference lists, and ClinicalTrials.gov.

Review methods: Using predefined criteria, we selected randomized controlled trials of noninvasive nonpharmacological treatments for five common chronic pain conditions (chronic low back pain; chronic neck pain; osteoarthritis of the knee, hip, or hand; fibromyalgia; and tension headache) that addressed efficacy or harms compared with usual care, no treatment, waitlist, placebo, or sham intervention; compared with pharmacological therapy; or compared with exercise. Study quality was assessed, data extracted, and results summarized for function and pain. Only trials reporting results for at least 1 month post-intervention were included. We focused on the persistence of effects at short term (1 to <6 months following treatment completion), intermediate term (≥6 to <12 months), and long term (≥12 months).

Results: Two hundred eighteen publications (202 trials) were included. Many included trials were small. Evidence on outcomes beyond 1 year after treatment completion was sparse. Most trials enrolled patients with moderate baseline pain intensity (e.g., >5 on a 0 to 10 point numeric rating scale) and duration of symptoms ranging from 3 months to >15 years. The most common comparison was against usual care.

Chronic low back pain: At short term, massage, yoga, and psychological therapies (primarily CBT) (strength of evidence [SOE]: moderate) and exercise, acupuncture, spinal manipulation, and multidisciplinary rehabilitation (SOE: low) were associated with slight improvements in function compared with usual care or inactive controls. Except for spinal manipulation, these interventions also improved pain.

Effects on intermediate-term function were sustained for yoga, spinal manipulation, multidisciplinary rehabilitation (SOE: low), and psychological therapies (SOE: moderate). Improvements in pain continued into intermediate term for exercise, massage, and yoga (moderate effect, SOE: low); mindfulness-based stress reduction (small effect, SOE: low); spinal manipulation, psychological therapies, and multidisciplinary rehabilitation (small effects, SOE: moderate). For acupuncture, there was no difference in pain at intermediate term, but a slight improvement at long term (SOE: low). Psychological therapies were associated with slightly greater improvement than usual care or an attention control on both function and pain at short-term, intermediate-term, and long-term followup (SOE: moderate). At short and intermediate term, multidisciplinary rehabilitation slightly improved pain compared with exercise (SOE: moderate). High-intensity multidisciplinary rehabilitation (≥20 hours/week or >80 hours total) was not clearly better than non–high-intensity programs.

Chronic neck pain: At short and intermediate terms, acupuncture and Alexander Technique were associated with slightly improved function compared with usual care (both interventions), sham acupuncture, or sham laser (SOE: low), but no improvement in pain was seen at any time (SOE: llow). Short-term low-level laser therapy was associated with moderate improvement in function and pain (SOE: moderate). Combination exercise (any 3 of the following: muscle performance, mobility, muscle re-education, aerobic) demonstrated a slight improvement in pain and function short and long term (SOE: low).

Osteoarthritis: For knee osteoarthritis, exercise and ultrasound demonstrated small short-term improvements in function compared with usual care, an attention control, or sham procedure (SOE: moderate for exercise, low for ultrasound), which persisted into the intermediate term only for exercise (SOE: low). Exercise was also associated with moderate improvement in pain (SOE: low). Long term, the small improvement in function seen with exercise persisted, but there was no clear effect on pain (SOE: low). Evidence was sparse on interventions for hip and hand osteoarthritis. Exercise for hip osteoarthritis was associated with slightly greater function and pain improvement than usual care short term (SOE: low). The effect on function was sustained intermediate term (SOE: low).

Fibromyalgia: In the short term, acupuncture (SOE: moderate), CBT, tai chi, qigong, and exercise (SOE: low) were associated with slight improvements in function compared with an attention control, sham, no treatment, or usual care. Exercise (SOE: moderate) and CBT improved pain slightly, and tai chi and qigong (SOE: low) improved pain moderately in the short term. At intermediate term for exercise (SOE: moderate), acupuncture, and CBT (SOE: low), slight functional improvements persisted; they were also seen for myofascial release massage and multidisciplinary rehabilitation (SOE: low); pain was improved slightly with multidisciplinary rehabilitation in the intermediate term (SOE: low). In the long term, small improvements in function continued for multidisciplinary rehabilitation but not for exercise or massage (SOE: low for all); massage (SOE: low) improved long-term pain slightly, but no clear impact on pain for exercise (SOE: moderate) or multidisciplinary rehabilitation (SOE: low) was seen. Short-term CBT was associated with a slight improvement in function but not pain compared with pregabalin.

Chronic tension headache: Evidence was sparse and the majority of trials were of poor quality. Spinal manipulation slightly improved function and moderately improved pain short term versus usual care, and laser acupuncture was associated with slight pain improvement short term compared with sham (SOE: low).

There was no evidence suggesting increased risk for serious treatment-related harms for any of the interventions, although data on harms were limited.

Conclusions: Exercise, multidisciplinary rehabilitation, acupuncture, CBT, and mind-body practices were most consistently associated with durable slight to moderate improvements in function and pain for specific chronic pain conditions. Our findings provided some support for clinical strategies that focused on use of nonpharmacological therapies for specific chronic pain conditions. Additional comparative research on sustainability of effects beyond the immediate post-treatment period is needed, particularly for conditions other than low back pain.