(090) Association between Recently Raised Anticholinergic Burden and Risk of Delirium – A case-case-time-control Study

Background: Although studies have revealed an association between the baseline anticholinergic burden and the risk of delirium, the evidence remains limited with concerns of the confounding by indication. Additionally, the protopathic bias warrants attention, as the early symptoms of delirium may lead to the use of psychotropic drugs which are included in the calculation of anticholinergic burden.

Objectives: To evaluate the association between anticholinergic burden and the risk of delirium among the older patients in Taiwan.

Methods: We conducted a case-case-time-control study analyzing the Taiwan National Health Insurance Research Database (NHIRD). Cases were defined as patients aged 65+ diagnosed with delirium from 2011 to 2018. We performed the case-case-time-control analysis by selecting future cases with events within 60-180 days after the matched event date. The hazard period was defined as 30 days before the delirium diagnosis, with referent periods being random 30-day intervals between 60 and 180 days prior to the diagnosis. We employed the Anticholinergic Cognitive Burden Scale (ACB) for categorizing scores into 3+, 1-2, and 0. We also considered acute conditions as time-varying variables to address the confounding by indication. Conditional logistic regression was used to assess the odds ratios (ORs) and 95% confidence interval (95% CIs) between the hazard and referent periods. Furthermore, we excluded patients receiving psychotropic drugs in the year before the index dates to reduce confounding by indication.

Results: A total of 330,910 future cases were eligible in the case-case-time-control analysis. The average age was 79.5 (SD 8.0) and 44.5% (n=147,245) were male. The case-crossover analysis indicated ACB scoring 1-2 points (OR 1.06; 95% CIs, 1.03-1.08) and 3+ points (OR 1.31; 1.29-1.34) were associated with a higher risk of delirium. After we addressed the issue of confounding by indication, the case-case-time-control OR for ACB scoring 1-2 points was at 0.93 (95% CIs, 0.91-0.95) and 3+ points at 0.99 (95% CIs, 0.97-1.01), indicating recently raised anticholinergic burden was no longer associated with an increased risk of delirium. The results of the sensitivity analysis were consistent to the primary analysis.

Conclusions: We initially found older patients with a recently raised anticholinergic burden were associated with an increased risk of delirium. However, when the confounding by indication were controlled, the association between the recently raised anticholinergic burden and increased risk of delirium was no longer observed. The finding highlighted the need to consider the confounding by indication when interpreting the results from the case-case-time-control study.