Published on Wed Aug 18 2021

Impact of rescue medication in placebo-controlled trials of pharmacotherapy for neuropathic pain and low back pain.

Lars Grøvle, Eivind Hasvik, Anne Julsrud Haugen

Rescue medication (RM) consumption is commonly used as a secondary outcome in placebo-controlled trials of chronic pain. If participants randomized to placebo take more RM than those randomized to an active drug, the difference in pain may be reduced.

1
1
2
Abstract

Rescue medication (RM) consumption is commonly used as a secondary outcome in placebo-controlled trials of chronic pain, but its validity has yet to be established. If participants randomized to placebo take more RM than those randomized to an active drug, the difference in pain between the 2 groups may be reduced, potentially masking effects of the active drug. This study assessed proportional RM consumption in the placebo and active groups according to results of 42 randomized controlled trials of neuropathic pain (NeP), and 29 trials of low back pain, which were included in 2 systematic reviews and meta-analyses. Trial results were assessed based on effect size, statistical significance, and whether the drug was recommended as first-line treatment by the systematic reviews. In trials indicating effect of the investigational drug, RM consumption was generally higher in the placebo groups than in the active groups. In trials reporting a small or a medium effect size of the investigational drug, subjects receiving placebo consumed 17% to 30% more RM than subjects receiving active drug, potentially leading to underestimation of the effects of the investigational drugs. Few trials reported a large effect size. Differences in RM consumption between participants receiving placebo and those receiving active drug were seldom taken in account by the individual trials and not at all by the systemic reviews when making treatment recommendations for NeP or low back pain. Elaboration on analytical methods to assess treatment effects in chronic pain trials using RM is warranted.