Relieving patients' pain with expectation interventions: a meta-analysis.

Patients' expectations are important predictors of the outcome of analgesic treatments, as demonstrated predominantly in research on placebo effects. Three commonly investigated interventions that have been found to induce expectations (verbal suggestion, conditioning, and mental imagery) entail promising, brief, and easy-to-implement adjunctive procedures for optimizing the effectiveness of analgesic treatments. However, evidence for their efficacy stems mostly from research on experimentally evoked pain in healthy samples, and these findings might not be directly transferable to clinical populations. The current meta-analysis investigated the effects of these expectation inductions on patients' pain relief. Five bibliographic databases were systematically searched for studies that assessed the effects of brief verbal suggestion, conditioning, or imagery interventions on pain in clinical populations, with patients experiencing experimental, acute procedural, or chronic pain, compared with no treatment or control treatment. Of the 15,955 studies retrieved, 30 met the inclusion criteria, of which 27 provided sufficient data for quantitative analyses. Overall, a medium-sized effect of the interventions on patients' pain relief was observed (Hedges g = 0.61, I = 73%), with varying effects of verbal suggestion (k = 18, g = 0.75), conditioning (always paired with verbal suggestion, k = 3, g = 0.65), and imagery (k = 6, g = 0.27). Subset analyses indicated medium to large effects on experimental and acute procedural pain and small effects on chronic pain. In conclusion, patients' pain can be relieved with expectation interventions; particularly, verbal suggestion for acute procedural pain was found to be effective.