Het belang van interbeoordelaars­betrouwbaarheid in psychofarmacologisch onderzoek

BACKGROUND: There are concerns about the declining efficacy of antidepressants and antipsychotics in clinical trials. A potential cause may be found in poor training practices to achieve sufficient inter-rater reliability (IRR). However, it is unknown whether IRR and training procedures are currently reported. AIM: To determine the proportion of publications concerning double-blind randomized controlled trials (RCTs) investigating antipsychotics or antidepressants that report IRR and training procedures. METHOD: We extracted all double-blind RCTs from five large meta-analyses concerning antidepressants and antipsychotics. Further, we conducted a Medline-search for double-blind RCTs investigating antidepressants from January 2016 - January 2020, and antipsychotics from January 2000 - January 2019. RESULTS: In 179 double-blind RCTs with antidepressants, only 4.5% reported an IRR coefficient whereas 27.9% reported on training procedures. Further, in 207 double-blind RCTs with antipsychotics, 11.2% reported an IRR coefficient and 34.8% reported training procedures. CONCLUSION: There is a substantial lack of reporting IRR and training procedures in RCTs with antidepressants and antipsychotics. Considering the implications of insufficient IRR, it is necessary to conduct and report training procedures and IRR. Reporting IRR and training procedures should be made mandatory by editorial boards of scientific journals.