| Study | Reason for exclusion |
|---|---|
| Addington 1996 | Allocation: randomised. Participants: people with acute schizophrenia, but not specifically overtly aggressive. |
| Alexander 2004 | Allocation: randomised. Participants: acute agitation and dangerous behaviour. Intervention: haloperidol + promethazine vs olanzapine, no haloperidol alone group. |
| Allen 2007 | Allocation: not randomised, review. |
| Asadollahi 2015 | Allocation: randomised. Participants: psychotic disorders (N = 52), mood disorders (N = 23), cognitive impairment (N = 23), adjustment disorders (N = 17), others (infection, substance intoxication or withdrawal) (N = 35), and unknown aetiology (N = 10). Therefore, eligible patients are < 50% of the sample. An attempt to contact the authors for partial outcomes was made (3rd July 2016). |
| Beasley 1998 | Allocation: randomised. Participants: people with first episode psychosis, but not specifically aggressive. |
| Bieniek 1998 | Allocation: randomised. Participants: newly admitted patients to a psychiatric emergency service who were uncooperative, threatening, and unresponsive to verbal interventions. Intervention: haloperidol + lorazepam vs lorazepam, no haloperidol alone group. |
| Brook 1998 | Allocation: randomised. Participants: people with acute psychosis, but not specifically aggressive. |
| Campbell 1982 | Allocation: randomised. Participants: children with aggressive behaviour, but the number of those with suspected psychosis < 50% of the sample. |
| Chen 2004 | Allocation: quasi‐randomised. |
| Chen 2008 | Allocation: quasi‐randomised. |
| Chouinard 1991 | Allocation: randomised. Participants: people with acute psychosis, but not specifically aggressive. |
| Citrome 2001 | Allocation: randomised. Participants: people with schizophrenia and persistent aggressive behaviour, not acute exacerbation of symptoms. |
| Colonna 1998 | Allocation: randomised. Participants: people with acute schizophrenia, but not specifically aggressive. |
| Conde 2011 | Allocation: randomised. Participants: admitted patients to an emergency setting experiencing agitation associated with active psychosis; inpatients experiencing exacerbation or agitation with active psychosis. Intervention: haloperidol + clonazepam vs risperidone + clonazepam, no haloperidol alone group. |
| Crandall 2005 | Allocation: randomised. Participants: people with chronic schizophrenia, acute exacerbation of symptoms, not specifically aggressive. |
| Currier 2000 | Allocation: not randomised. |
| Daniel 2000 | Allocation: not randomised, review. |
| Daniel 2003 | Allocation: randomised. Participants: people with acute schizophrenia, not acutely agitated or aggressive. |
| Daniel 2004 | Allocation: randomised. Participants: people with psychosis, not acutely agitated or aggressive. |
| Fan 2012 | Quasi‐randomised controlled trial: "the participants were assigned by the order of admission". |
| Faretra 1970 | Allocation: randomised. Participants: children with suspected schizophrenia, not specifically aggressive. |
| Freeman 2009 | Allocation: not randomised, retrospective case note study. |
| Goldstein 1966 | Allocation: randomised. Participants: people with acute psychosis, not aggressive. |
| Harvey 2004 | Allocation: randomised. Participants: people with mild symptoms of schizophrenia, not aggressive. |
| Huang 2004 | Allocation: quasi‐randomised. |
| Jiang 2009 | Quasi‐randomised controlled trial: "the participants were assigned by the order of admission". |
| Jones 2001 | Allocation: not randomised, review. |
| Kane 2000 | Allocation: randomised. Participants: people with acute schizophrenia, not specifically aggressive. |
| Kinon 2001a | Allocation: randomised. Participants: people with schizophrenia, but not specifically overtly aggressive. |
| Kinon 2003 | Allocation: not randomised. |
| Krakowski 2006 | Allocation: randomised. Participants: people with schizophrenia and persistent aggressive behaviour, not an emergency situation. |
| Krakowski 2008 | Allocation: randomised. Participants: people with schizophrenia and persistent aggressive behaviour, not an emergency situation. |
| Lewis 2006 | Allocation: randomised. Participants: people with schizophrenia and persistent aggressive behaviour, not an emergency situation. |
| Li 2005 | Allocation: randomised. Participants: people with acute excitement in schizophrenia. Intervention: haloperidol vs tiapride. Outcome: there was no specific outcome for acute agitation (T1 at 7 days). |
| Li 2007 | Allocation: quasi‐randomised. |
| Liang 2003 | Quasi‐randomised controlled trials: "the participants were randomly assigned by the order of admission". |
| Liu 2012b | Allocation: randomised. Participants: people with schizophrenia and acute agitation. Intervention: haloperidol vs modified electroconvulsive therapy (MECT), no comparison drug. |
| Mandel 2008 | Allocation: randomised. Participants: people with schizophrenia, not specifically aggressive. |
| Mantua 2006 | Allocation: not randomised, naturalistic observational study. |
| McEvoy 1994 | Allocation: randomised. Participants: people with schizophrenia, not specifically aggressive. |
| Mendelowitz 2004 | Allocation: not randomised, review. |
| NCT00189995 2005 | Allocation: was to be randomised, but trial not conducted because ethics approval not granted. |
| NCT00631722 2008 | Allocation: randomised. Participants: people with acute schizophrenia, not specifically aggressive. |
| NCT00797277 2008 | Allocation: randomised. Participants: people with schizophrenia and agitation. Intervention: haloperidol plus lorazepam vs olanzapine ‐ no haloperidol alone group. |
| Pan 2005 | Allocation: quasi‐randomised. |
| Pathiraja 1995 | Allocation: randomised. Participants: people with schizophrenia, not specifically aggressive. |
| Pedros 2004 | Allocation: not randomised ‐ cohort study. |
| Pei 2009 | Quasi‐randomised controlled trial: "the participants were assigned by the order of admission". |
| Puech 1998 | Allocation: randomised. Participants: people with acute schizophrenia, not specifically aggressive. |
| Romeo 2009 | Allocation: randomised. Participants: people with intellectual disability and challenging behaviour, not psychosis‐induced aggression. |
| Simpson 2010 | Allocation: randomised. Participants: people with agitation in the emergency setting. Intervention: haloperidol plus benztropine mesylate vs lorazepam vs quetiapine ‐ no haloperidol alone group. |
| Singh 1981 | Allocation: not reported, described as double‐blind. Participants: people with schizophrenia, not acute exacerbation of symptoms. |
| Smith 1974 | Allocation: randomised. Participants: people with senile psychosis, not specifically aggressive. |
| Srinath 2010 | Allocation: randomised. Participants: "patients who presented with acute psychotic agitation". Intervention: haloperidol + promethazine vs lorazepam, no haloperidol alone group. |
| Teja 1975 | Allocation: randomised. Participants: people with chronic schizophrenia, not specifically aggressive. |
| Thomas 1992 | Allocation: randomised. Participants: violent and agitated people, mostly intoxicated, not necessarily psychosis‐induced aggression. |
| TREC CG | Allocation: randomised. Participants: acute agitation and dangerous behaviour. Intervention: haloperidol + promethazine vs midazolam (no haloperidol alone group). |
| Ungvari 1982 | Allocation: randomised. Participants: people with acute schizophrenia, not specifically aggressive. |
| Vaisanen 1981 | Allocation: randomised. Participants: people with intellectual disability and agitation, not psychosis‐induced aggression. |
| Veser 2006 | Allocation: randomised. Participants: people with psychosis‐induced agitation, not specifically aggressive (exclusion criteria included an inability to give informed consent). |
| Vives 2015 | Allocation: randomised. Participants: agitated schizophrenic patients in the emergency setting. Intervention: haloperidol intranasal vs haloperidol intramuscular, no other drugs except haloperidol. |
| Wan 2005 | Allocation: randomised. Participants: people with schizophrenia and agitation/aggression. Intervention: haloperidol vs chlorpromazine vs risperidone + clonazepam. Outcome: excitement and agitation scale (not validated), also the data for the haloperidol and chlorpromazine groups are combined. |
| Wang 2004 | Allocation: quasi‐randomised. |
| Wang 2005 | Allocation: quasi‐randomised. |
| Wyant 1990 | Allocation: randomised. Participants: people with schizophrenia and acute agitation. Intervention: haloperidol vs sodium amytal vs midazolam. Outcome: modified CGI ‐ no usable data, presented in graphs, no binary data. |
| Zapletalek 1986 | Allocation: not reported, described as double‐blind. Participants: people with schizophrenia, not specifically aggressive. |
CGI: Clinical Global Impression vs: versus