Results. Visuotactile illusions decreased pain by an average of 7.8 points (95% CI [2.0–13.5]) which corresponds to a 25% reduction in pain, but the tactile only and visual only control conditions did not (Condition × Time interaction: p = 0.028). Visuotactile illusions did not differ from incongruent control conditions where the same visual manipulation occurred, but did differ when only the same tactile input was applied. Sustained illusions prolonged analgesia, but did not increase it. Repeated illusions increased the analgesic effect with an average pain decrease of 20 points (95% CI [6.9–33.1])–corresponding to a 40% pain reduction.
Discussion. Visuotactile illusions are analgesic in people with knee OA. Our results suggest that visual input plays a critical role in pain relief, but that analgesia requires multisensory input. That visual and tactile input is needed for analgesia, supports multisensory modulation processes as a possible explanatory mechanism. Further research exploring the neural underpinnings of these visuotactile illusions is needed. For potential clinical applications, future research using a greater dosage in larger samples is warranted.
Analgesic Efficacy and Pain Stickiness? Most analgesics have an NNT for 50% pain relief in chronic pain of between 3–10, and the superior effect of the primary drug over placebo is in the region of 30% [192]. In essence the majority of people treated with pharmacological interventions do not experience the desired effect [191].
It is, however, well known that pain relief has a binomial or U-shaped distribution, meaning that describing a sample by its mean score is to choose the experience of the least number of people [191]. It is better to think more carefully about who responds and why. One reason for the ineffectiveness may relate to how drugs, when used chronically, may enhance pain. While mechanisms are not fully understood, one example is chronic opioid use associated with increased DNA methylation and increased levels of clinical pain [69]. It should be noted however, that some of these data should be assessed in the context of duration of treatment; for example, for opioid effects on chronic pain, the longest randomized controlled trial (RCT) is only 16 weeks [45].