Haloperidol for the treatment of delirium in critically ill patients: an updated systematic review with meta-analysis and trial sequential analysis

Background Haloperidol is frequently used in critically ill patients with delirium, but evidence for its effects has been sparse and inconclusive. By including recent trials, we updated a systematic review assessing effects of haloperidol on mortality and serious adverse events in critically ill patients with delirium. Methods This is an updated systematic review with meta-analysis and trial sequential analysis of randomised clinical trials investigating haloperidol versus placebo or any comparator in critically ill patients with delirium. We adhered to the Cochrane handbook, the PRISMA guidelines and the grading of recommendations assessment, development and evaluation statements. The primary outcomes were all-cause mortality and proportion of patients with one or more serious adverse events or reactions (SAEs/SARs). Secondary outcomes were days alive without delirium or coma, delirium severity, cognitive function and health-related quality of life. Results We included 11 RCTs with 15 comparisons (n = 2200); five were placebo-controlled. The relative risk for mortality with haloperidol versus placebo was 0.89; 96.7% CI 0.77 to 1.03; I2 = 0% (moderate-certainty evidence) and for proportion of patients experiencing SAEs/SARs 0.94; 96.7% CI 0.81 to 1.10; I2 = 18% (low-certainty evidence). We found no difference in days alive without delirium or coma (moderate-certainty evidence). We found sparse data for other secondary outcomes and other comparators than placebo. Conclusions Haloperidol may reduce mortality and likely result in little to no change in the occurrence of SAEs/SARs compared with placebo in critically ill patients with delirium. However, the results were not statistically significant and more trial data are needed to provide higher certainty for the effects of haloperidol in these patients. Trial registration: CRD42017081133, date of registration 28 November 2017. Supplementary Information The online version contains supplementary material available at 10.1186/s13054-023-04621-4.

Haloperidol versus other antipsychotics Mortality Three trials with a total of 436 patients reported on mortality when comparing haloperidol to other antipsychotics. The control intervention was either chlorpromazine, quetiapine or ziprasidone. One trial was of low risk of bias. We found no statistically significant difference in mortality when meta-analysing data from the three trials (RR 1.07; 96.7%CI 0.82-1.41; I 2 =0%; TSAadjusted CI 0.38-2.97) ( Figure S1). TSA indicated that 16% of the required information size was accrued ( Figure S6). The quality of evidence was judged to very low due to indirectness and imprecision (Table S1).

SAEs/SARs
Four trials with a total of 460 patients reported on SAEs/SARs. The control intervention was either chlorpromazine, quetiapine or ziprasidone and one trial was of low risk of bias. All trials reported on mortality but few specified this outcome as an SAE. We categorised mortality as an SAE in accordance with International Conference on Harmonisation good clinical practice. In the two analyses estimating the number of patients with one or more SAEs/SARs we found no statistically significant difference between haloperidol and other antipsychotics (highest proportion: RR 1.06; 96.7% CI 0.81-1.40; I 2 = 0%; TSA adjusted CI 0.38 -3.00; P = 0.58, cumulated: RR 1.05; 96.7%CI 0.93-1.18; I 2 =0%; TSA adjusted CI 0.85 -1.33; P=0.32) (Figure S7 + S14). TSA for both SAE/SAR measures indicated that less than 50% of the required information size were accrued, and the cumulative z-curve did not cross any TSA monitoring boundaries. The quality of evidence was judged to very low for both SAE/SAR measures due to indirectness and imprecision (Table S1).

Days alive without delirium or coma (14 days)
Two trials with a total of 302 patients reported on days alive without delirium or coma. The comparators were either ziprasidone or quetiapine, 1 trial was of low risk of bias. We found no statistically significant difference in days alive without delirium or coma between haloperidol and other antipsychotics (mean difference (MD): -0.15; 98%CI -1.83 to 1.53; I 2 =0%; TSA-adjusted CI: -2.49 to 2.10; P=0.75). TSA indicated that less than 50% of the required information size was accrued and no monitoring boundaries were crossed. The quality of evidence was judged to low due to indirectness and imprecision (Table S1).

Delirium Severity
Three trials with a total of 248 patients reported on delirium severity. All trials were of high risk of bias. The comparators were chlorpromazine, risperidone or quetiapine. We found no statistically significant difference in delirium severity between haloperidol and other antipsychotics (standardized mean difference (SMD): 0.01; 98%CI -0.29 to 0.31; I 2 =0%; TSA-adjusted CI: too little information to conduct TSA) ( Figure S27). The quality of evidence was judged to very low due to risk of bias, indirectness and imprecision (Table S1).

Health-related quality of life (HRQL), Cognitive function and QTc prolongation
No trials reported on HRQL. 1 trial reported on cognitive function, this trial included 24 patients and had chlorpromazine as comparator ( Figure S30). 1 trial reported on QTc prolongation, this trial included 376 patients and had ziprasidone as comparator ( Figure S31). No statistically significant differences were found for any of these outcomes and the quality of evidence were judged to very low for both outcomes (Table S1).
Haloperidol versus dexmedetomidine Only 1 trial examined the effect of haloperidol versus dexmedetomidine. This trial included 64 patients and was judged high risk of bias for all reported outcomes; SAEs/SARs, delirium severity and QTc prolongation. No statistically significant differences were found for any of the reported outcomes for haloperidol versus dexmedetomidine ( Figure S7+S14+S27+S30). Quality of evidence was judged very low for all outcomes due to risk of bias, imprecision and suspected publication bias (not shown).

Haloperidol versus benzodiazepines
Only 1 trial examined the effect of haloperidol versus benzodiazepines more specifically lorazepam. This trial included 17 patients and was judged high risk of bias for all reported outcomes; mortality, SAEs/SARs, delirium severity and cognitive function. No statistically significant differences were found for any of the reported outcomes for haloperidol versus lorazepam ( Figure S1+S7+S14+S27+S30). Quality of evidence was judged very low for all outcomes due to risk of bias, imprecision and suspected publication bias (not shown).
Haloperidol versus opioids Only 1 trial examined the effect of haloperidol versus opioids more specifically morphine. This trial included 53 patients and was judged high risk of bias for all reported outcomes; mortality and SAEs/SARs. No statistically significant differences were found for any of the reported outcomes for haloperidol versus lorazepam ( Figure S1+S7+S14). Quality of evidence was judged very low for all outcomes due to risk of bias, imprecision and suspected publication bias (not shown).
Haloperidol versus antiemetics Only 1 trial examined the effect of haloperidol versus antiemetics more specifically ondansetron. This trial included 64 patients and was judged high risk of bias for all reported outcomes; SAEs/SARs, delirium severity and QTc prolongation. No statistically significant differences were found for any of the reported outcomes for haloperidol versus antiemetics ( Figure  S7+S14+S27+S30). Quality of evidence was judged very low for all outcomes due to risk of bias, imprecision and suspected publication bias (not shown).
Haloperidol versus no control Only 1 trial examined the effect of haloperidol versus no control. This trial included 10 patients and was judged high risk of bias for all reported outcomes; mortality, SAEs/SARs, delirium severity, cognitive function and QTc prolongation. No statistically significant differences were found for any of the reported outcomes for haloperidol versus no control ( Figure S1+S7+S14+S30+S31). Quality of evidence was judged very low for all outcomes due to risk of bias, imprecision and suspected publication bias (not shown).

Details on sensitivity analyses: best/worst-case and worst/best-case scenarios
To assess the potential impact of missing data for dichotomous outcomes, the two following analyses will be performed: 1. 'Best-worst-case' scenario: It will be assumed that all participants lost to follow up in the experimental group survived, had no serious adverse event, and had no morbidity; and all those with missing outcomes in the control group did not survive, had a serious adverse event, and had morbidity. 2. 'Worst-best-case' scenario: It will be assumed that all participants lost to follow-up in the experimental group did not survive, had a serious adverse event, and had morbidity; and all those with missing outcomes in the control group did survive, had no serious adverse event, and had no morbidity.
Results from both scenarios will be presented in the review.
To assess the potential impact of the missing data for continuous outcomes, the two following analyses will be performed: 1. 'Best-worst-case' scenario: It will be assumed that all participants lost to follow up in the experimental group had mean (from patients with follow-up) + 2 x SD; and all those with missing outcomes in the control group had mean (from patients with follow-up) minus 2 x SD. 2. 'Worst-best-case' scenario: It will be assumed that all participants lost to follow-up in the experimental group had mean (from patients with follow-up) minus 2 x SD; and all those with missing outcomes in the control group had mean (from patients with follow-up) + 2 x SD.
To assess the potential impact of missing SDs for continuous outcomes, the following sensitivity analyses will be performed: Where SDs are missing and not possible to calculate, SDs will be imputed from trials with similar populations and low risk of bias. If no such trials can be found, SDs will be imputed from trials with a similar population. As the final option SDs will be imputed from all trials.

Details of included trials and risk of bias assessments
Delirium screening: Delirium screening was performed by clinicians primarily ICU nurses with CAM-ICU or ICDSC. Screening was performed twice daily.

Interventions
Experimental intervention: intravenous haloperidol 2.5mg x 3 daily with as needed doses up to a maximum of 20mg daily as long as patients were delirious.

Control intervention: matching intravenous placebo (isotonic saline)
Duration: 90 day intervention period. Patients received the intervention until they were delirium free (2 negative delirium screenings on the same day), discharged from the ICU or for a maximum of 90 days. In case of recurrence of delirium in the ICU intervention was resumed.
Open-label antipsychotics: Not allowed to be given to trial participants in the ICU during the 90-day intervention period. Participants that were withdrawn from the trial could receive open-label antipsychotics. Exposure to open-label antipsychotics (any antipsychotic) in the trial population was 13.1%.

Outcomes
Outcomes: -Days alive and out of hospital to day 90 -90-day mortality -Hospital length of stay to day 90 -Days alive without delirium or coma -Serious Adverse Reaction in ICU -Number of patients using of rescue medication -Days with rescue medication per patient All outcomes were assessed at day 90

Funding
Innovation Funds Denmark, The Regional Medicines Fund, The Zealand region Research Fund, Intensiv Symposium Hindsgavl, Fogts Foundation.

Notes
Risk of bias assessment was performed by Mathieu Van der Jagt as the author (NCAR) was involved in the AID-ICU trial and thereby unfit to make such an assessment for the AID-ICU trial. Data for days alive without coma or delirium to day 14 was extracted from the database.

Randomised clinical trial
The trial was conducted in a community hospital in Turkey  Delirium screening: CAM-ICU by research team. Motor subtype by RASS by clinicians (intensivist and psychiatrist). Assessed every 12h, until discharge from the hospital or for a maximum of 10 days. Patients were considered delirium free when they were free of symptoms for 24h.

Co-interventions:
If patients were still agitated when maximum dose was reached 2.5mg of lorazepam PO was added twice daily to the intervention.
During admission to the ICU, every patient was ventilated in assist-control mode to maintain pH between 7.35 and 7.45, PaCO2 between 35 and 45 mmHg, and PaO2 4 95%. Ventilation was weaned as per ICU protocol.

Open-label antipsychotics: Not reported
Outcomes Outcomes: -Duration time of delirious behaviour -Daily total medication doses -Need for additional sedative drug -RASS scores -The percentage of patients who maintained a RASS score within the target scores -Reintubation -Redo-surgery -Length of ICU and hospital stay -Readmission to the ICU -Hospital mortality rate Timing of outcomes: until discharged from the hospital or for a maximum of 10 days following surgery.

Funding
Not reported Notes Email sent to Dr. Atalan 24 February 2023 and 13 March 2023. Reply received.

Randomised clinical trial
The trial was conducted in Saudi Arabia in a mixed ICU with trauma being the most common admission diagnosis Exclusion criteria: Patients were excluded if they had underlying neurological diseases, significant hearing loss, intracranial injury, ischemic/hemorrhagic strokes or language barrier that would confound the evaluation of delirium. Similarly, severely injured, deeply comatose, or moribund patients were excluded.
Delirium screening: Screening tool was ICDSC. Screening was performed by ICU nurses. Patients were screened twice daily for 3 days. The ICDSC was assessed 1h after study medication were given.

Duration: 3 days
Open-label antipsychotics: The treating physicians were free to prescribe additional haloperidol as rescue when clinically needed in all the three groups.
Rescue haloperidol was used in 9.4% of the patients in the haloperidol group, 15.6% of the patients in the dexmedetomidine groups and in 34.4% of the patients in the ondansetron group. A total of 19.8% of the trial population was exposed to open-label antipsychotics.

Outcomes
Outcomes: -Number of patients with delirium at day 3 -Number of patients requiring rescue haloperidol -ICDSC scores -Mean arterial blood pressure -Mean Visual Analog Scale (VAS) of pain at the time of delirium assessment -Serious adverse events -Prolongation of QTc interval

Founding sources No external funding Notes
Email sent to Dr. Bakri asking for additional data 24 February and 13 March 2023.
No reply was received.

Randomised clinical trial
The trial was conducted in the US Exclusion criteria: known hypersensitivity to neuroleptics or benzodiazepines, presence of neuroleptic malignant syndrome, concurrent treatment with neuroleptic drugs, seizure disorder, current systemic chemotherapy for Karposi's sarcoma, withdrawal syndrome or anticholinergic delirium for which a more specific treatment was indicated, current or past diagnosis of schizophrenia, schizoaffective disorder, or bipolar disorder and the study compromising medical treatment for the underlying aetiology. Efforts were made to also exclude patients in whom delirium appeared to be part of a terminal event (i.e., the patient was expected to die within 24 hours).
Delirium screening: enrolled patients were treated to the study protocol if they met DSM-III-R criteria for delirium and scored above the threshold score diagnostic for delirium (score of 13 or greater) on the delirium rating scale. The delirium rating scale is a 10 item scale specifically integrating DSM-III criteria. The maximum possible score is 32.

Interventions
Experimental intervention: a treatment protocol for study drug administration was followed. Each delirious patient was evaluated hourly with the Delirium Rating Scale -if the patient's score was 13 or greater, the next level dose of study drug was administered. After stabilization (when the patient was asleep, calm, and not hallucinating or had scored 12 or below on the Delirium Rating Scale) a maintenance dose was started on day 2 and continued for up to 6 days of treatment protocol. Mean haloperidol dose the first 24 hours was 2.8 mg. Average maintenance dose was 1.4 mg.
Control intervention: mean chlorpromazine dose the first 24 hours was 50 mg. Average maintenance dose was 36 mg. Mean lorazepam dose the first 24 hours was 3 mg. Average maintenance dose was 4.6 mg.

Open-label antipsychotics: Not reported Outcomes
-Mortality -Scores on Delirium Rating Scale -Scores on Mini-Mental State (cognitive function) Timing of outcome measurement: day 2 end of treatment (delirium and cognitive scores) and 8 days from initiation of protocol (mortality)

Funding sources Not reported Notes
No information about use of rescue medication.
Midway through the study, lorazepam was removed from the study due to treatment limiting adverse side effects in this group.
28 May 2019: E-mail sent to Dr Breitbart asking for additional information on risk of bias and outcomes. Reply was received. Additional data on delirium resolution was not received.

Randomised clinical trial
The trial was conducted in the Netherlands, in 8 ICUs Exclusion criteria for eligibility: Admission to the ICU because of a primary acute neurological condition; pregnancy or breast-feeding; known allergy to haloperidol; history of ventricular arrhythmia (including torsade de pointes); neuroleptic malignant syndrome; parkinsonism; schizophrenia or other psychotic disorder; dementia or an Informant Questionnaire on Cognitive Decline in the Elderly (IQCODE) score ≥ 4; expected duration of ICU admission < 24 hours; inability to speak the Dutch language or to undergo a valid delirium assessment (e.g. deafness or blindness); or participation in another interventional trial. Additional exclusion criteria after development of delirium: QTc prolongation (QTc>500ms), acute alcohol (or substance) withdrawal syndrome, an expected ICU stay <24 hours, or torsade de pointes, neuroleptic malignant syndrome, or parkinsonism since ICU admission.

Inclusion criteria for randomisation:
Delirium assessed with the Intensive Care Delirium Screening Checklist (ICDSC ≥4) or the Confusion Assessment Method for the ICU (positive CAM-ICU assessment), at the time of ICU admission or any ICU day after ICU admission. Written informed consent obtained from the patient or their legal representative. 3. All eligibility inclusion criteria (from above) are still met.
Exclusion criteria for randomisation: Prolonged QT-interval (QTc >500ms). (recent) Torsade de Pointes. (recent) Neuroleptic malignant syndrome or parkinsonism. Evidence of acute alcohol (or substance) withdrawal requiring pharmacological intervention (eg, benzodiazepines or alpha-2 agonist) to treat. Or if the patient is expected to die within 24hours or expected to leave the ICU within 24hours. No (previously) signed informed consent by patient or representative.
Delirium screening: CAM-ICU or ICDSC. Screening was performed by trained ICU nurses. Patients were evaluated 3 times daily. Patients were considered delirium free if delirium screenings were negative for 24 hours.

Interventions
Experimental intervention: 2.5mg haloperidol three times daily (patients aged ≥ 80 1mg), and increased up to 5mg three times daily (patients aged ≥ 80 2.5mg). Increased dose if delirium persisted in the next 8-hour shift. Decrease dose if negative 24 hours and stopped when another 24hours had elapsed. Trial drugs were re-initiated if delirium reoccurred.

Duration: 14 days
Open-label antipsychotics: Use of open-label haloperidol was strongly discouraged, but could be used for acute breakthrough delirium symptoms that could not be managed in the proposed protocol. Atypical antipsychotics could be administered to patients with hallucinations. A total 13 patients (9.8%) received open-label haloperidol and 55 (41.6%) received other antipsychotics (quetiapine or olanzapine) open-label.

Outcomes
Outcomes: Primary outcome -number of delirium-and coma-free days (DCFDs) to day 14 Secondary outcomes: -28-day mortality -duration of first delirium episode -number of days with coma, agitation or mechanical ventilation -mean daily RASS score -use of "escape medication" for agitation or psychotic symptoms -maximum mobility level -sleep quality -ICU length of stay -adverse events related to delirium -duration of the first delirium episode -days with agitation -serious adverse drug associated events -Time to 'readiness for discharge from the ICU'.
-Hospital discharge

Notes
The trial was pre-maturely stopped.

Funding
A grant has been provided by ZonMw -The Netherlands Organisation for Health Research and Organisation. ZonMw project number: 848041001 Notes Unpublished data of the EuRIDICE trial on mortality and SAEs/SARs were included in this review. Data on Days alive without delirium or coma and QTc prolongation were still privileged until publication of the trial results and thereby not included in this review. Email sent to Dr. van der Jagt and L. Smith on 24 February asking for additional data, reply was received.

Randomised clinical trial
The trial was conducted in India Exclusion criteria: Patients who at baseline had severe cognitive impairment, one who were at high risk for medication side-effects because of pregnancy and breast feeding. Patients with history of torsades de pointes, neuroleptic malignant syndrome, or allergy to haloperidol or quetiapine were not considered. Also those who had ongoing treatment with antipsychotics, having rapidly resolving organ failure, moribund patients, and those who were blind or unable to speak or understand were excluded Control intervention: Matching placebo or quetiapine (max 300mg)

Open-label antipsychotics: not reported
Duration: Unclear, but all outcomes are reported at 14 days. It is anticipated that this is the intervention period and patients received the intervention as long as they were delirious.
Inclusion criteria: adults (≥18 years old) with delirium admitted to medical and/or surgical ICUs in participating hospitals who were treated with mechanical ventilation, non-invasive positive pressure ventilation, vasopressor(s), or intraaortic balloon pump.
Exclusion criteria: patients who, at baseline, had severe cognitive impairment; were at high risk for medication side effects because of pregnancy, breast-feeding, a history of torsades de pointes, QT prolongation, a history of neuroleptic malignant syndrome, or allergy to haloperidol or ziprasidone; were receiving ongoing treatment with an antipsychotic medication; were in a moribund state; had rapidly resolving organ failure; were blind, deaf, or unable to speak or understand English; were incarcerated; or were enrolled in another study or trial that prohibited co-enrolment.
Delirium: Delirium was detected with the use of the Confusion Assessment Method for the ICU (CAM-ICU). Trained research personnel evaluated patients twice daily until delirium was present or until death, discharge from the ICU, development of an exclusion criterion, or a maximum of 5 days.

Interventions
Experimental intervention: patients younger than 70 years of age received 2. Patients younger than 70 years of age received 0.5 ml placebo (0.9% saline) and 0.25 ml of placebo when older than 70 years of age.
Volume and dose of a trial drug or placebo were halved if a patient did not have delirium (i.e., had a negative CAM-ICU assessment) for two consecutive assessments and was not yet receiving the minimum dose. Trial drug or placebo were temporarily withheld if a patient did not have delirium for four consecutive assessments or for safety reasons.
Trial drug or placebo were permanently discontinued when any of the following occurred: torsades de pointes, neuroleptic malignant syndrome, drug reaction with eosinophilia and systemic symptoms syndrome, new onset coma due to structural brain disease, or any life-threatening, serious adverse event that was related to the intervention, as determined by an independent data and safety monitoring board.
Timing: patients were randomised when delirium was present at the time of informed consent or during the 5 days after informed consent was obtained and the corrected QT interval was less than 550 msec on a 12-lead electrocardiogram. Immediately after the trial-group assignment, the first dose of trial drug or placebo was administered.
Duration: 14 days or at ICU discharge. The median duration of exposure to a trial drug or placebo was 4 days (interquartile range, 3 to 7).

Outcomes
Outcomes: -days alive without delirium or coma (defined as the number of days that a patient was alive and free from both delirium and coma during the 14-day intervention period) -duration of delirium -time to freedom from mechanical ventilation (defined as extubation that was followed by at least a 48-hour period during which the patient was alive and free from mechanical ventilation) -time to final successful ICU discharge (defined as the last ICU discharge during the index hospitalization that was followed by at least a 48-hour period during which the patient was alive and outside the ICU) -time to ICU readmission -time to successful hospital discharge (defined as discharge that was followed by at least a 48-hour period during which the patient was alive and outside the hospital) -30-day and 90-day survival -incidence of torsades de pointes -incidence of neuroleptic malignant syndrome -severity of extrapyramidal symptoms as measured on the modified Simpson-Angus Scale

Duration: 7 days Outcomes
Outcomes: -mean delirium Rating Scale scores -delirium severity assessed with Memorial Delirium Assessment Scale -delirium resolution -duration of delirium

Notes
The study included patients admitted to four medical wards, two ICUs and two oncology wards. As more than 90% of the included patients came from ICU, we have included this study in our review.
Memorial Delirium Assessment Scale scores (mean and standard deviation) were extracted from figure 1.

QTc-prolongation (follow-up: mean 14 days)
Electronic          Figure S5: TSA sensitivity analysis of all-cause mortality for haloperidol versus placebo (RR11%) Trial Sequential Analysis of all-cause mortality for trials investigating haloperidol versus placebo. In this sensitivity analysis we used a control event rate of 38.9%, alfa-level of 3.3%, beta 10% (90% power) and as indicated by the meta-analysis (Figure 1 manuscript) a relative risk reduction of 11% and diversity of 0%. TSA showed that 26% of the required information size was accrued and we are therefore unable to reject or accept a RR of 11%.  Figure S6: TSA of all-cause mortality for haloperidol versus other antipsychotics TSA of all-cause mortality for trials investigating haloperidol versus other Antipsychotics (Chlorpromazine, quetiapine and ziprasidone). We used a control event rate of 35% as indicated by the meta-analysis, a RRR/RRI of 20%, alfa-level of 3.3%, beta 10% (90% power) and diversity of 20%. TSA showed that 17% of the required information size was accrued and we are therefore unable to reject or accept a RRR/RRI of 20%. Test for subgroup differences (random effects): Chi 2 = 2.14, df = 4 (P = 0.71)        Figure S11: TSA of SAE/SAR highest proportion for haloperidol versus placebo (RR20%) TSA of SAE/SAR highest proportion for trials investigating haloperidol versus placebo. We used a control event rate of 43%, a RR of 20%, alfa-level of 3.3%, beta 10% (90% power) and diversity of 33% as indicated by the meta-analysis. TSA showed that the z-curve reached the futility area meaning that haloperidol does not cause a 20% relative change in experiencing SAE/SAR (highest proportion), randomising more patients will not change this result.  Figure S12: TSA sensitivity analysis of SAE/SAR highest proportion for haloperidol versus placebo (RR 6%) Trial Sequential Analysis of SAE highest proportion for trials investigating haloperidol versus placebo. In this sensitivity analysis we used a control event rate of 43.3%, alfa-level of 3.3%, beta 10% (90% power) and as indicated by the meta-analysis ( Figure S7) a relative risk reduction of 6% and diversity of 0%. TSA showed that 6% of the required information size was accrued and we are therefore unable to reject or accept a RR of 6%.  Figure S13: TSA of SAE/SAR for haloperidol versus other antipsychotics TSA of SAE/SAR highest proportion for trials investigating haloperidol versus other Antipsychotics (Chlorpromazine, quetiapine, ziprasidone, risperidone). We used a control event rate of 33% as indicated by the meta-analysis, a RRR/RRI of 20%, alfa-level of 3.3%, beta 10% (90% power) and diversity of 20%. TSA showed that 16% of the required information size was accrued and we are therefore unable to reject or confirm a RRR/RRI of 20%.     Test for subgroup differences (common effect): Chi 2 = 8.09, df = 4 (P = 0.09) Test for subgroup differences (random effects): Chi 2 = 6.85, df = 4 (P = 0.14)    Figure S18: TSA of cumulated SAEs/SARs for haloperidol versus placebo TSA of cumulated SAEs/SARs for trials investigating haloperidol versus placebo. We used a control event rate of 57.6%, a RRR/RRI of 20%, alfa-level of 3.3%, beta 10% (90% power) and diversity of 89% as indicated by the meta-analysis ( Figure S14). TSA showed that 18% of the required information size was accrued and we are therefore unable to reject or confirm a RRR/RRI of 20%.  Figure S19: TSA of cumulated SAEs/SARs for haloperidol versus other antipsychotics TSA of cumulated SAEs/SARs for trials investigating haloperidol versus other antipsychotics (chlorpromazine, quetiapine, ziprasidone, risperidone). We used a control event rate of 63.5%, a RR of 20%, alfa-level of 3.3%, beta 10% (90% power) and diversity of 20%. TSA showed that 49% of the required information size was accrued and we are therefore unable to reject or confirm a RRR/RRI of 20%.         Figure S24: TSA of Days alive without delirium or coma (14 days) for haloperidol versus placebo TSA of Days alive without delirium or coma (14 days) for trials investigating haloperidol versus placebo. We used a control MD 1 day, a RR of 20%, alfa-level of 2%, beta 10% (90% power) and diversity of 20%. TSA found that with 80% of the required information size, the cumulated z-curve crossed into the futility area, hence haloperidol does not cause a 20% relative change in days alive without delirium or coma compared with placebo.   Figure S25: TSA sensitivity analysis of Days alive without delirium or coma (14 days) for haloperidol versus placebo (MD=0.33) TSA sensitivity analysis of Days alive without delirium or coma (14 days) for trials investigating haloperidol versus placebo. We used an alfa-level of 2%, beta 10% (90% power), control MD 0.33 days and diversity of 0% as indicated by the meta-analysis. TSA showed that 11% of the required information size was accrued and we are therefore unable to reject or accept an absolute mean difference of 0.33 days in days alive without delirium or coma.   Figure S26: TSA of Days alive without delirium or coma (14 days) for haloperidol versus other antipsychotics TSA of Days alive without delirium or coma (14 days) for trials investigating haloperidol versus other antipsychotics (ziprasidone and quetiapine). We used a control MD 1.7 days, a RR of 20%, alfa-level of 2%, beta 10% (90% power) and diversity of 21%. TSA showed that 45% of the required information size was accrued and we are therefore unable to reject or accept a RRR/RRI of 20% in days alive without delirium or coma.