Nocebo effects in multiple sclerosis trials: a meta-analysis
Aimed to estimate the incidence and severity of nocebo responses in trials of symptomatic treatments (STs) and disease-modifying treatments (DMTs) for multiple sclerosis (MS). Papadopoulos D, Mitsikostas DD. Mult Scler. 2010 Jul;16(7):816-28. Epub 2010 Jun 10
Nocebo effect refers to a subject-oriented adjective used to label the harmful, unpleasant, or undesirable reactions (or responses) that a subject manifests (thus, "nocebo reactions" or "nocebo responses") as a result of administering an inert dummy drug or placebo, where these responses have not been chemically generated, and are entirely due to the subject's pessimistic belief and expectation that the inert drug would produce harmful, injurious, unpleasant, or undesirable consequences.
In these cases, there is no "real" drug involved, but the actual harmful, unpleasant or undesirable physiological, behavioural, emotional, and/or cognitive consequences of the administration of the inert drug are very real.
Authors conducted a systematic Medline search for all randomised, placebo-controlled MS trials published between 1989 and 2009. Meta-analysis of the incidence of nocebo responses was performed by pooling the percentage of placebo-treated patients that exhibited adverse events. Nocebo severity was calculated from the percentage of placebo-treated patients that dropped-out due to drug-related adverse events.
Data were extracted from 56 DMT and 44 ST eligible trials. The pooled incidence of nocebo responses was 74.4% (95% CI: 69.92-88.30) in DMT trials and 25.3% (95% CI: 15.24-36.90) in ST trials and was significantly higher in the former (p < 0.0001). The pooled nocebo severity was 2.1% (95% CI: 1.6-2.67) in DMT and 2.34% (95% CI: 1.54-3.29) in ST trials. Meta-regression analysis revealed a higher nocebo incidence in parallel design ST studies compared to crossover ones (p = 0.013) and a higher nocebo severity in phase II ST studies compared to phase III ones (p = 0.0001). Nocebo severity in DMT trials exhibited an association with the year of study publication (p = 0.011) and the frequency of drug administration (p = 0.0082).
Authors conclude that nocebo responses in MS trials are substantial and appear to have increased significantly in recent years with important implications for both trial design and clinical practice. Furthermore, nocebo responses exhibit an association with medication and trial-related factors.