Delirium-associated medication in people at risk: A systematic update review, meta-analyses, and GRADE-profiles

Background Drug-associated delirium is a common but potentially preventable neuropsychiatric syndrome associated with detrimental outcomes. Empirical evidence for delirium-associated medication is uncertain due to a lack of high-quality studies. We aimed to further investigate the body of evidence for drugs suspected to trigger delirium. Methods A systematic update review and meta-analyses of prospective studies presenting drug associations with incident delirium in adult study populations was conducted. Two authors independently searched MEDLINE, PsycINFO, Embase, and Google Scholar dated from October 1, 2009 to June 23, 2020, after screening a previous review published in 2011. The most reliable results on drug-delirium associations were pooled in meta-analyses using the random-effects model. Quality of evidence was assessed using the GRADE-approach. This study is preregistered with OSF (DOI ). Results The 31 eligible studies, presenting results for 24 medication classes were identified. Meta-analyses and GRADE level of evidence ratings show no increased delirium risk for Haloperidol (OR: 0.96, 95% CI 0.72-1.28; high-quality evidence), Olanzapine (OR: 0.25, 95% CI 0.15-0.40), Ketamine (OR: 0.72, 95% CI 0.35-1.46) or corticosteroids (OR: 0.69, 95% CI 0.32-1.50; moderate quality evidence, respectively). Low-level evidence suggests a three-fold increased risk for anticholinergics (OR: 3.11, 95% CI 1.04-9.26). Opioids, benzodiazepines, H-1-antihistamines, and antidepressants did not reach reliable evidence levels in our analyses. Conclusion We investigated the retrievable body of evidence for delirium-associated medication. The results of this systematic review were then interpreted in conjunction with other evidence-based works and guidelines providing conclusions for clinical decision-making.