Models of intervention and care for psychostimulant users, 2nd edition - monograph series no. 51

Acute behavioural disturbances and psychoses

Page last updated: April 2004



Urgent sedation in an emergency (sometimes referred to as chemical restraint) is a procedure for administering drug treatment to rapidly control extremely agitated, aggressive behaviour of an individual at risk of causing physical harm to themselves or others. The primary aim of emergency sedation is to attenuate specific symptoms of behavioural disturbance rather than as a treatment for an underlying cause or psychiatric condition. Nonetheless, effective sedation may provide a safe environment in which to determine and treat the cause of agitated behaviour.

Urgent sedation should be distinguished from procedures aimed at rapidly treating psychoses. Procedures such as 'rapid neuroleptisation' consist of giving high loading doses of antipsychotics to attenuate psychotic symptoms. These procedures are no longer recommended in an emergency setting where high or frequently repeated doses of antipsychotics may exacerbate the emergency situation through side-effects including dysphoria, akathisia or acute dystonia (Keckich, 1978; Siris, 1985).

Reliance on physical restraint alone is often not adequate for psychostimulant users experiencing acute behavioural disturbance and may actually cause harm if agitation increases. Stimulant use has been suggested as a possible risk factor for sudden death of individuals being physically restrained (Stratton, Rogers, Brickett & Gruzinski, 2001). Sedation using sedative drugs is acceptable to patients (Sheline & Nelson, 1993), provides a humane alternative to physical restraint (Richards, Derlet & Duncan, 1998) and ensures simpler and safer essential physiological monitoring than other types of restraint.


Use of psychostimulants has been associated with violent or agitated behaviour, which may lead to fatal outcomes (Dowling, McDonough & Bost, 1987). Behavioural effects are influenced by dose used, characteristics of the individual and the social context of the psychostimulant use (Miczek & Tidey, 1989). Violent behaviour is more common in chronic high-use psychostimulant users than occasional users. Other factors such as coincident opiate withdrawal may increase risk of aggressive behaviour (Miczek & Tidey, 1989).

Common acute effects of amphetamines include panic or motor agitation. With prolonged use, hypervigilance and euphoria gradually give way to auditory, visual and tactile illusions, hallucinations and paranoia. Delusions are common, as is the preoccupation with 'bugs' that are felt and seen on the skin, leading to picking and excoriation of the skin. Restless choreoathetoid and tic-like movements are often present. Experienced amphetamine users may describe the combination of paranoia and compulsive movements as 'tweaking'. Delirium may occur (Forster et al., 1999). Top of page


Conducting an exhaustive differential diagnosis is less important when sedating an acutely agitated patient than when formulating longer-term treatment in an in-patient unit (Citrome & Volavka, 1999). Indications for urgent sedation in suspected psychostimulant users include:
  1. failure of other attempts to control the patient such as de-escalation and other non-drug interventions;

  2. the patient is uncooperative;

  3. the patient is at known or imminent risk to themselves or others; and

  4. there is a perceived need for medical intervention (requirement of the Guardianship Act).
In some emergency situations, it may be difficult to differentiate between behavioural disturbance and potential drug-induced psychosis. Suspected drug-induced psychosis should not be considered a contraindication to urgent sedation. Rather, a period of sedation and behavioural control will allow clinicians to re-assess the patient after the acute effects of the drug have worn off, allowing for a more accurate differential diagnosis. In general, treatment of patients with psychostimulant-induced psychosis is similar to treatment of acute mania or schizophrenia (Forster et al., 1999) and establishing a 'safe' environment should be the first priority.


Non-specific sedation is frequently used in the management of acutely agitated or violent patients. The setting of clinical contact (emergency department versus ambulance attendance) may influence drug selection and route of administration.

Ideal medications for urgent sedation should possess rapid sedative action, providing quick control of dangerous behaviour. Sedation should generally be titrated to the point of rousable sleep, not unconsciousness. The aim of sedation is to control dangerous behaviour sufficiently to facilitate assessment and management.

Over-sedation in the form of loss of consciousness should be avoided. Health care providers who provide sedation, regardless of practice setting, should have access to advanced airway assessment and management skills so that successful 'rescue' of patients can be made should an adverse sedation event occur.


Forster and colleagues (Forster et al., 1999) suggest that benzodiazepines should be the agent of choice when there is unlikely to be an ongoing need for antipsychotic medication after acute treatment, warning that little data support frequent administration of 'as needed' antipsychotic medication. They suggest that benzodiazepines influence fewer neurotransmitter systems than antipsychotic agents and are thus a safer (pharmacologically 'cleaner') choice of drug. Consistent with other survey findings (Sheline & Nelson, 1993), their clinical experience predicts that most agitated patients are more willing to accept treatment with a benzodiazepine than with an antipsychotic and that following such treatment, patients tend to be calmer and better organised.

Secondary benefits of selecting a benzodiazepine are that they are also part of first line treatment for cardiac toxicity associated with psychostimulant use (Albertson, Dawson, de Latorre, Hoffman et al., 2001) and may exert some benefit in the agitation of serotonin toxicity (Graber, Hoehns & Perry, 1994). In cases of adverse events, a pharmacological antagonist (flumazenil) is available to reverse benzodiazepine effects (Hunkeler, Mohler, Pieri, Polc et al., 1981).
Top of page
Neave (Neave, 1994) suggests that parenteral midazolam may be effective in controlling agitated or aggressive patients. The dose administered should be based on the patient's general health, age, weight and level of agitation or aggression; usually 5, 10, or 15 milligrams are given incrementally at 15-minute intervals until the desired effect is achieved. The advantages of midazolam over other benzodiazepines include its rapid onset (onset of action: intravenous (IV) three to five minutes; intramuscular (IM) 15 minutes) (Nordt & Clark, 1997), shorter duration of action, less potential to cause hypotension and that prolonged administration results in more rapid awakening (Dundee, Halliday, Harper & Brogden, 1984; Simpson & Eltringham, 1981).

Although some commentators suggest that lorazepam is an "excellent choice" for sedating violent patients (Citrome & Volavka, 1999), few studies have demonstrated its superiority over other agents. It is considered to be at least as effective as haloperidol (Bick & Hannah, 1986) and demonstrates a better safety profile than typical antipsychotics (Lenox, Newhouse, Creelman & Whitaker, 1992), mainly attributed to its lack of extrapyramidal side-effects. However, lorazepam does not have a rapid onset of action (IV 15-20 minutes; IM two hours) (Dundee, Lilburn, Nair & George, 1977; Greenblatt, Ehrenberg, Gunderman, Scavone et al., 1989), which is also a slower onset of effect than that of droperidol (Richards et al., 1998). It must be noted that neither IM nor IV preparations of lorazepam are available in Australia.

Typical antipsychotics

It is sometimes considered that using antipsychotics for sedation purposes may also confer a benefit via their antipsychotic actions. However, this putative benefit would only be evident after the acute episode of agitation or violence has subsided (Citrome & Volavka, 1999).

Haloperidol, a butyrophenone antipsychotic, is frequently used for urgent sedation. Compared with other neuroleptics, haloperidol causes less hypotension, fewer anticholinergic side-effects and less decrease in the seizure threshold. Despite this, haloperidol is not the most sedative of neuroleptics and may thus be less appropriate than more sedating agents for emergency sedation purposes (Citrome, 2002; Citrome & Volavka, 1999).

In an uncontrolled study (Clinton, Sterner, Stelmachers & Ruiz, 1987), 136 patients were treated primarily with IM haloperidol although IV and oral routes were also utilised. Haloperidol alleviated the problem behaviour in 83% of patients; two patients experienced dystonic reactions.

Another butyrophenone used to treat acute agitation is droperidol. Droperidol is fast acting, rapidly eliminated from the body and may be administered IM or IV. Small, uncontrolled studies indicate that droperidol is useful in controlling methamphetamine intoxication (Gary & Saidi, 1978), severely agitated psychotic patients (Granacher & Ruth, 1979; Hooper & Minter, 1983), acute agitation related to traumatic brain injury (Stanislav & Childs, 2000), or agitated patients in a pre-hospital setting (Hick, Mahoney & Lappe, 2001). Chambers and Druss (1999) recommend that droperidol be considered a drug of choice in psychiatric emergencies due to its efficacy and rapidity of action.

Two placebo-controlled studies (Rosen, Ratliff, Wolfe, Branney et al., 1997; van Leeuwen, Molders, Sterkmans, Mielants et al., 1977) assessed the utility of droperidol in acute agitation and found that it provided greater sedative effect than placebo within 3-5 minutes of administration. Limitations of both studies include small sample size and no longer-term follow-up of participants.

Comparisons between lorazepam and droperidol have demonstrated that droperidol may have greater efficacy. One randomised study (Richards et al., 1998) compared the effectiveness of lorazepam versus droperidol in a heterogenous sample of agitated patients (N=202) in the emergency department. Agitation was attributed to methamphetamine toxicity in 72% of cases and cocaine toxicity in 14% of cases. Patients received either lorazepam 4 mg IV or droperidol 5 mg IV; sedation levels and vital signs were monitored for 60 minutes. Droperidol provided more rapid sedation than lorazepam and achieved higher levels of sedation that were maintained over the 60 minutes and required less frequent repeat dosing. No patients experienced adverse effects on their vital signs; no patients required airway intervention. There were no significant advantages to either drug with regard to net change in pulse, systolic blood pressure, respiratory rate or blood pressure.

One patient who received droperidol experienced an acute dystonic reaction. Results were similar in the subset of patients whose agitation was attributed to methamphetamine toxicity (Richards, Derlet & Duncan, 1997).

The main benefit demonstrated for droperidol in this study was its rapid onset of action. Given that lorazepam does not have a rapid onset of action (see above) it may have been more appropriate to compare droperidol with more rapidly acting benzodiazepines such as midazolam.
Top of page
Potential problems associated with use of droperidol include dystonia and akathisia, hypotension, prolongation of the QTc interval on ECG, lowering of the seizure threshold and respiratory depression (Chase & Biros, 2002; Granacher & Ruth, 1979; Heard, Daly, O'Malley & Rosen, 1999; Stanislav & Childs, 2000). However, some authors suggest that serious QTc interval prolongation associated with droperidol is uncommon in practice, where it is safe and effective for the treatment of violence and agitation (Shale, Shale & Mastin, 2003). These problems may be more of an issue in a psychostimulant-affected population although data supporting this contention are lacking.

The Cochrane Review of droperidol for acute psychosis (Cure & Carpenter, 2001) concludes that this area is under-researched and that use of droperidol in an emergency situation is currently based on experience rather than evidence from well-conducted clinical trials. In comparison studies (Resnick & Burton, 1984; Thomas, Schwartz & Petrilli, 1992), droperidol (5 mg IM) produced more effective sedation than haloperidol (5 mg IM), although there were no differences between the two drugs when given by the IV route (Thomas et al., 1992). Although some research has examined the role of atypical antipsychotics to treat agitation (e.g. Currier & Simpson, 2001), little evidence supports their use for urgent sedation.

Combination regimens

A combination of lorazepam (2 mg IM) and haloperidol (5 mg IM) was compared with 2 mg lorazepam alone in 20 agitated patients presenting to a psychiatric emergency service (Bieniek, Ownby, Penalver & Dominguez, 1998).The combination group exhibited greater improvements in some, but not all, outcome measures. Both groups improved over time.

Battaglia and colleagues (1997) compared three regimens: lorazepam 2 mg IM alone, haloperidol 5 mg IM alone and a combination of both drugs at the same doses. All medications led to significant improvements in aggression over 12 hours; the combination group demonstrated a greater improvement in some measures and experienced less extra-pyramidal symptoms than those receiving haloperidol alone. These three treatment groups were compared in a similar study (Garza-Trevino, Hollister, Overall & Alexander, 1989) although doses of lorazepam used were 4 mg instead of 2 mg. They report that the combination was superior to either of the single agents and that lorazepam alone was slightly superior to haloperidol alone.

These studies support the use of benzodiazepines and antipsychotics in combination as safe and effective options. Whilst these results also seem to indicate that combination therapies are superior to single agent regimes, it is important to note that doses in each arm are not necessarily equally effective and that superior efficacy of a combination regime may merely reflect that patients in the combination group received a greater total dose of drug than those in a single agent group.