Abstract
Delirium is common in post-acute care (PAC) patients with dementia; its treatment is not established. We hypothesized that cognitive activities would reduce the duration and severity of delirium and improve cognitive and physical function to a greater extent than usual care. This was an NIH-funded single-blind randomized clinical trial conducted in eight PAC facilities. Participants were 283 community-dwelling older adults with dementia and delirium randomized to intervention or control. The intervention group received cognitive activities daily for up to 30 days. The control group received usual care. Primary outcomes were delirium duration (Confusion Assessment Method), and delirium severity (Delirium Rating Scale). Secondary outcomes were cognitive function (Digits Forward, Montreal Cognitive Assessment and CLOX) and physical function (Barthel Index). There were no differences between the groups for mean percentage of delirium-free days or delirium severity. Significant differences for secondary outcomes favoring intervention were found: executive function: 6.58 (95% CI: 6.12–7.04) vs. 5.89 (95% CI: 5.45–6.33), a difference of -0.69 (95% CI: 1.33- -0.06, p=0.03); and constructional praxis: 8.84 (95% CI: 8.83–9.34) vs. 7.53 (95% CI: 7.04–8.01), a difference of - 1.31 (95% CI: 2.01- -0.61, p=0.0003). After adjusting for baseline constructional praxis the group comparison was no longer significant. Average length of stay was shorter in intervention (36.09 days vs. 53.13 days, SE = 0.15, p = 0.01, negative binomial regression. Cognitive activities did not improve delirium but did improve executive function and reduced length of stay. Resolution of delirium may require more intense non-pharmacological management when the patient has dementia.