Haloperidol plus promethazine for psychosis-induced aggression

Abstract

Background: Health services often manage agitated or violent people, and such behaviour is particularly prevalent in emergency psychiatric services (10%). The drugs used in such situations should ensure that the person becomes calm swiftly and safely. Objectives: To examine whether haloperidol plus promethazine is an effective treatment for psychosis-induced aggression. Search methods: On 6 May 2015 we searched the Cochrane Schizophrenia Group's Register of Trials, which is compiled by systematic searches of major resources (including MEDLINE, EMBASE, AMED, BIOSIS, CINAHL, PsycINFO, PubMed, and registries of clinical trials) and their monthly updates, handsearches, grey literature, and conference proceedings. Selection criteria: All randomised clinical trials with useable data focusing on haloperidol plus promethazine for psychosis-induced aggression. Data collection and analysis: We independently extracted data. For binary outcomes, we calculated risk ratio (RR) and its 95% confidence interval (CI), on an intention-to-treat basis. For continuous data, we estimated the mean difference (MD) between groups and its 95% CI. We employed a fixed-effect model for analyses. We assessed risk of bias for included studies and created 'Summary of findings' tables using GRADE. Main results: We found two new randomised controlled trials (RCTs) from the 2015 update searching. The review now includes six studies, randomising 1367 participants and presenting data relevant to six comparisons. When haloperidol plus promethazine was compared with haloperidol alone for psychosis-induced aggression for the outcome not tranquil or asleep at 30 minutes, the combination treatment was clearly more effective (n=316, 1 RCT, RR 0.65, 95% CI 0.49 to 0.87, high-quality evidence). There were 10 occurrences of acute dystonia in the haloperidol alone arm and none in the combination group. The trial was stopped early as haloperidol alone was considered to be too toxic. When haloperidol plus promethazine was compared with olanzapine, high-quality data showed both approaches to be tranquillising. It was suggested that the combination of haloperidol plus promethazine was more effective, but the difference between the two approaches did not reach conventional levels of statistical significance (n=300, 1 RCT, RR 0.60, 95% CI 0.22 to 1.61, high-quality evidence). Lower-quality data suggested that the risk of unwanted excessive sedation was less with the combination approach (n=116, 2 RCTs, RR 0.67, 95% CI 0.12 to 3.84). When haloperidol plus promethazine was compared with ziprasidone all data were of lesser quality. We identified no binary data for the outcome tranquil or asleep. The average sedation score (Ramsay Sedation Scale) was lower for the combination approach but not to conventional levels of statistical significance (n=60, 1 RCT, MD -0.1, 95% CI - 0.58 to 0.38). These data were of low quality and it is unclear what they mean in clinical terms. The haloperidol plus promethazine combination appeared to cause less excessive sedation but again the difference did not reach conventional levels of statistical significance (n=111, 2 RCTs, RR 0.30, 95% CI 0.06 to 1.43). We found few data for the comparison of haloperidol plus promethazine versus haloperidol plus midazolam. Average Ramsay Sedation Scale scores suggest the combination of haloperidol plus midazolam to be the most sedating (n=60, 1 RCT, MD - 0.6, 95% CI -1.13 to -0.07, low-quality evidence). The risk of excessive sedation was considerably less with haloperidol plus promethazine (n=117, 2 RCTs, RR 0.12, 95% CI 0.03 to 0.49, low-quality evidence). Haloperidol plus promethazine seemed to decrease the risk of needing restraints by around 12 hours (n=60, 1 RCT, RR 0.24, 95% CI 0.10 to 0.55, low-quality evidence). It may be that use of midazolam with haloperidol sedates swiftly, but this effect does not last long. When haloperidol plus promethazine was compared with lorazepam, haloperidol plus promethazine seemed to more effectively cause sedation or tranquillisation by 30 minutes (n=200, 1 RCT, RR 0.26, 95% CI 0.10 to 0.68, high-quality evidence). The secondary outcome of needing restraints or seclusion by 12 hours was not clearly different between groups, with about 10% in each group needing this intrusive intervention (moderate-quality evidence). Sedation data were not reported, however, the combination group did have less 'any serious adverse event' in 24-hour follow-up, but there were not clear differences between the groups and we are unsure exactly what the adverse effect was. There were no deaths. When haloperidol plus promethazine was compared with midazolam, there was clear evidence that midazolam is more swiftly tranquillising of an aggressive situation than haloperidol plus promethazine (n=301, 1 RCT, RR 2.90, 95% CI 1.75 to 4.8, high-quality evidence). On its own, midazolam seems to be swift and effective in tranquillising people who are aggressive due to psychosis. There was no difference in risk of serious adverse event overall (n=301, 1 RCT, RR 1.01, 95% CI 0.06 to 15.95, high-quality evidence). However, 1 in 150 participants allocated haloperidol plus promethazine had a swiftly reversed seizure, and 1 in 151 given midazolam had swiftly reversed respiratory arrest. Authors' conclusions: Haloperidol plus promethazine is effective and safe, and its use is based on good evidence. Benzodiazepines work, with midazolam being particularly swift, but both midazolam and lorazepam cause respiratory depression. Olanzapine intramuscular and ziprasidone intramuscular do seem to be viable options and their action is swift, but resumption of aggression with subsequent need to re-inject was more likely than with haloperidol plus promethazine. Haloperidol used on its own without something to offset its frequent and serious adverse effects does seem difficult to justify.