Comparing Haloperidol and Olanzapine in Treating Terminal Delirium

Our study defines "terminal delirium" as an episode of delirium that occurs during the dying phase, usually 72 hours before death. Episodes of delirium in the dying phase may be described as "terminal restlessness" or "terminal delirium". The use of the label "terminal" implies a causal relationship between the dying phase and the delirium, although the aetiology is often multi-factorial.

A prospective, open-label, randomized controlled trial was designed. When a home hospice clinician identifies a patient who meets inclusion criteria, their proxy (i.e. family caregivers) will be approached by the study team and invited to participate in the study. Patients will be randomly assigned to one of two groups as part of follow-up management: (i) Oral Transmucosal Haloperidol or (ii) Oral Transmucosal Olanzapine.

After commencement of the first dose of medication, each patient will be observed by their family caregiver and the attending clinician over 72 hours (less if the patient dies earlier). Time-points for data collection will be at 24 hours, 48 hours, and 72 hours after first commencement of medication.

This study compares two common treatments in clinical practice, both of which have demonstrated efficacy greater than placebo. Since all participants will receive treatment that is no different from accepted practice, a placebo group was deemed unnecessary and unethical, thus it was not included.

Proxy consent is a method for collecting informed consent, where the proxy (or person responsible) for the potential participant is approached to provide consent on behalf of the prospective participant. Proxy consent was suggested as a suitable method for consent collection, given the patient's absent or fluctuating capacity to consent. It was successfully used in dementia and delirium research, and was reported to have been acceptable to patients and their caregivers.

The clinical staff of the hospice service will be briefed about the study and will assist in recruitment. As they provide standard palliative care to the patient, the clinical staff will (i) identify patients who meet inclusion criteria and share preliminary information with caregivers, (ii) confirm the diagnosis of delirium, and (iii) advise caregivers on appropriate non-pharmacological management.

The primary caregiver (i.e. proxy) will be approached by a study team member who is independent from the management of the patient. They will provide more detailed information about the study to the primary caregiver. If agreeable, the proxy will sign the informed consent form and the dyad is assigned a participant ID number. Basic demographic information, including patient's age, sex, primary diagnosis and co-morbidities will be collected.

Patients may withdraw from the study anytime, without compromising further treatment using other drugs or alternative dosage regimen.

Randomization sequence is created using Microsoft Excel 2016 with 1:1 allocation using random block sizes of 2 and 4. As each participant joins, they are dispensed either haloperidol or olanzapine according to the generated randomization list.

After obtaining consent, the participant-caregiver pair will be randomized into either the Haloperidol or Olanzapine group. Instructions on how to administer the medication will be provided. Caregivers will be advised to serve breakthrough doses of the assigned anti-psychotic drug as needed.

A daily pack of medication will be prepared for the caregiver during daily clinical reviews at home. The pack will have the correct dosages of medication prepared. Subcutaneous (SC) Midazolam is prepared as rescue medication in the event delirium symptoms remain uncontrolled and distressing after the trial medication are depleted.

Per clinical practice, the first dose of medication will be given under the supervision of the study team. The caregiver will be instructed how to rate the patient using the RASS and will provide the baseline RASS rating. The study team will assess the patient's baseline symptom severity with the MDAS.

Alongside drug therapy, usual non-pharmacological interventions will be provided to all patients. These interventions include: (i) regularly orientating the patient, (ii) keeping the room bright during the day, (iii) minimizing the use of tubes, catheters, physical restraints or other immobilizing devices, and (iv) minimizing unnecessary disturbances to the patient.

Every six hours, caregivers will chart the patient's agitation using the RASS. Additionally, if breakthrough medication is required, the time of administration will be documented.

To conduct assessments as well as to ensure patient's safety, the study team will make regular home visits to review the patient 24-hours, 48-hours, and 72-hours after first dose. During each review, the study team will document their assessments of patients' symptom severity (MDAS) and observe for adverse events (NCI CTCAE). The researchers will take appropriate action to ensure patient's safety during the trial, up to and including recommending the termination of the trial.

The trial will be conducted for up to 72 hours after recruitment or until patient's death. If the patient completes the trial and survives, the family will continue to receive support from their primary nurse and attending physician. Patients who survive beyond seven days after recruitment will be excluded from per-protocol analysis, as they no longer fit the inclusion criteria for the study.

The study may be terminated at any point during the trial. In that event, the caregivers will cease assessment using the RASS, while the clinicians will record the reason for cessation, and make a final MDAS and NCI CTC assessment, where applicable. The family will continue to receive support from their primary nurse and attending physician.

The proxy may choose to decline from having the patient participate in the trial. They may also choose to withdraw the patient from the study at any point during the trial. The patient and the family will continue to receive service support from their primary nurse and attending physician thereafter. They will advise how persisting symptoms should be managed, outside of the study. The study team will assure the proxy of this fact before informed consent is collected.

SC Midazolam is prepared as rescue medication in the event the symptoms of delirium remain uncontrolled and distressing. Rescue medications will be used in any one of the following situations:

1. Within 24 hours, the total doses for Haloperidol or Olanzapine reach more than 10mg or 20mg respectively and patient remains unsettled (i.e. all prepared medication are finished any time before the next clinical review).
2. Patient is unable to tolerate Oral Transmucosal medication.
3. Patient experiences serious adverse effects from trial medication.

The study will be terminated for that patient in any one of the following situations:

1. Rescue medication is used.
2. Adverse events related to the trial medication were observed to be intolerable.
3. Caregiver asks to withdraw the patient from the study.
4. Patient is admitted to hospital.

Changes to the severity of delirium are analysed with scores from MDAS. Repeated-measures, between-factors ANOVA will be employed to analyse differences in MDAS scores over time.

If a statistically significant difference is found between medications, post-hoc analysis will be conducted to analyse the change in MDAS scores at each time-point.

Changes in the patient's agitation will be trended using the RASS score. The RASS scores from Baseline, 6 hours, 12 hours, and 24 hours will be analysed. Moreover, an average of the RASS scores for the second day and the third day will be calculated and trended.

Secondary outcomes are possible adverse effects due to the medication, which is measured using NCI CTCAE. For each group, the toxicities (Akathasia, Extrapyramidal disorder, and spasticity) for each group will be compared.

Based on validation studies by Breitbart et al. (1997) and Lawlor et al. (2000), the mean MDAS scores for patients with delirium is roughly 18 (SD = 7.64). Jain et al. (2018) reported that using either Haloperidol or Olanzapine led to nearly 55% reduction in patients' MDAS score (7-8 points). 2 points was estimated to be the minimal important difference between Haloperidol and Olanzapine.

Assuming a data collection period of 27 months and a target recruitment size of 80 participants, the study must successfully recruit 3 - 4 participants per month.

A review of HCA Hospice Care's medical notes from January 2020 to August 2020 found that 250 patients exhibited symptoms of confusion, agitation, or delirium while in an unstable or deteriorating state (approximately 30 patients per month). The duration between the date of these case notes and the date of patient's death ranged 0 days to 17 days; 187 patients died within 7 days. Hence, an average of 23 potential patients per month would be recruited for the study.