Proportional Sedation for Persistent Agitated Delirium in Palliative Care

Importance: Neuroleptic and benzodiazepine medications are often considered for patients with persistent agitated delirium in the last days of life; however, the risk-to-benefit ratio of these medications is ill-defined and benzodiazepine medications have not been compared to placebo. Objective: To compare the effect of scheduled haloperidol, lorazepam, haloperidol plus lorazepam, and placebo on patients with advanced cancer and delirium and experiencing restlessness and/or agitation in the palliative care setting. Design, Settings, and Participants: This multicenter randomized clinical trial was conducted at 3 acute palliative care units in Taiwan and the US with patients with advanced cancer experiencing persistent restlessness and/or agitation despite nonpharmacologic therapies and standard-dose haloperidol. Among 245 eligible patients, 111 were enrolled, and 75 received blinded treatments. Participants were randomized in a 1:1:1:1 ratio (stratified by site and Richmond Agitation-Sedation Scale [RASS] score). The study period was from July 16, 2019, to June 8, 2023, with a 30-day follow-up after medication administration. Data analysis was performed from October 10, 2023, to April 11, 2025. Interventions: Scheduled intravenous haloperidol, lorazepam, haloperidol plus lorazepam, or placebo every 4 hours until discharge, death, or withdrawal from study. Medications in all 4 groups had identical volume and appearance. Main Outcomes and Measures: Change in RASS scores during the first 24 hours. Secondary outcomes included the use of rescue neuroleptics or benzodiazepines for breakthrough restlessness or agitation during the first 24 hours, delirium severity, perceived patient comfort, and adverse events. Results: The primary outcome was assessed in 72 patients (mean [SD] age, 64 [12] years, 42 male [58%]) with a median (IQR) MDAS score of 24 (18-29). The lorazepam group had significantly lower RASS scores than the haloperidol group (mean difference, -2.1; 95% CI, -3.4 to -0.9; P < .001) and the combination group had significantly lower RASS scores than the haloperidol group (-2.0; 95% CI, -3.2 to -0.8; P = .002); however, there was no difference observed between haloperidol and placebo groups (-0.5; 95% CI, -1.7 to 0.7; P = .42) nor between the combination and lorazepam groups (0.2; 95% CI, -1.1 to 1.4; P = .79). The combination and lorazepam groups required fewer rescue medications for breakthrough restlessness or agitation compared to the haloperidol and placebo groups (32%, 37%, 56%, 83%, respectively; P = .006). Adverse events or survival did not differ between groups. Conclusions and Relevance: The results of this randomized clinical trial indicate that proactive use of scheduled sedatives, particularly lorazepam-based regimens, may reduce persistent restlessness and/or agitation in patients with advanced cancer and delirium in the palliative care setting. Trial Registration: Clinicaltrials.gov Identifier: NCT03743649.