Haloperidol and Lorazepam in Controlling Symptoms of Persistent Agitated Delirium in Patients With Advanced Cancer Undergoing Palliative Care

The effect of neuroleptic dose escalation, benzodiazepine rotation, combination therapy, and neuroleptic withdrawal will be compared on the change in the RASS score over 24 hours in patients admitted to an APCU who do not respond to low-dose haloperidol.(RASS between -2 and 0) Will use 2-sided t-tests for four pre-specified paired comparisons, if the required assumptions for this model are met. Non-parametric methods such as Wilcoxon rank-sum test may be used if any assumptions are violated.

Will conduct Bonferroni adjustment for the number of rescue doses. Will perform a comparison among the four arms using analysis of variance (ANOVA), followed by selected pairwise t-test comparisons, or the Kruskal-Wallis test followed by pairwise Wilcoxon rank-sum tests if the data violate assumptions for ANOVA. Will also compare these outcomes longitudinally using linear mixed models, with treatment as a between-patient factor and time as a within-subject factor. Appropriate transformations will be used as needed to satisfy model assumptions.

Will compare the proportions across all four treatment groups using a chi-squared test, followed by selected pairwise chi-squared tests. We will also compare these outcomes longitudinally and include all data after treatment administration using generalized linear mixed models, with treatment as a between-patient factor and time as a within-subject factor.

Will be used for patients with caregivers or nurses listing "agree" or "strongly agree" for patient comfort, proportion of patients with at least 2 on the Udvalg for Kliniske Undersogelser (UKU) scale, and proportion of patients with caregivers listing "excellent" quality of end-of-life care)

Examines both the recalled frequency of 6 delirium symptoms and associated distress in the rater: disorientation to time and place, visual/tactile/auditory hallucinations, delusional thoughts, and psychomotor agitation. It will be administered to family caregivers and nurses daily. The score for recalled frequency ranges between 0 and 4, where 0=not present, 1=a little of the time, 2=some of the time, 3=good part of the time, and 4=most or all of the time. The score for distress in the rater related to each delirium symptom also ranges from 0 to 4, where 0=no distress, 1=a little, 2=a fair amount, 3=very much, and 4=extremely distressed. This assessment will be administered to the blinded caregivers and bedside nurses daily.

The achievement of the proxy comfort goal will be assessed.

Validated and widely used in different clinical settings, including the APCU. It assesses the average symptom intensity of 10 symptoms (pain, fatigue, nausea, depression, anxiety, drowsiness, shortness of breath, appetite, sleep, and feeling of well-being) over the past 24 hours using an 11-point numeric rating scale, ranging from 0 (none) to 10 (worst). Because all patients will be delirious, caregivers will be asked to provide their proxy rating of ESAS daily.

Will perform a comparison among the four arms using ANOVA, followed by selected pairwise t-test comparisons, or the Kruskal-Wallis test followed by pairwise Wilcoxon rank-sum tests if the data violate assumptions for ANOVA. Will also compare these outcomes longitudinally using linear mixed models, with treatment as a between-patient factor and time as a within-subject factor. Appropriate transformations will be used as needed to satisfy model assumptions.

Will phone the bereaved caregivers who were most involved in the patient's care within 4-8 weeks of the patient's death to assess the patient's perceived quality of care and quality of life at the end-of-life, on the basis of questions previously used in the Coping with Cancer Study. Specifically, will ask caregivers, "Overall, how would you rate the care (the patient) received at the palliative care unit? Would you say it was excellent, very good, good, fair, or poor?" and "In your opinion, how would you rate the overall quality of (the patient)'s death or last week of life?" using a numeric rating scale from 0 (worst possible) to 10 (best possible). Will also ask them 2 questions related to control of agitation, adapted from the Quality of Death and Dying Questionnaire.

In each of the groups, will identify independent factors that are predictive of response (i.e. RASS reduction of >= 1.5 points) to treatment using multivariable logistic regression analysis. Will assess the predictive value of plasma biomarkers (i.e. IL-6, IL-8, IL-10, and S100B) in the subset of patients.