Benzodiazepines not helpful for delirium.

Benzodiazepines are not the best choice for the treatment of delirium in mechanically ventilated inpatients, according to findings in a Cochrane review.

"Increased sedation and prolonged coma in lorazepamtreated patients suggest that such treatment may worsen the status of delirious patients, and clinicians should be guided accordingly," the authors cautioned.

Delirium, which affects up to 30% of hospitalized patients, has been associated with longer stays and increased morbidity and mortality. Benzodiazepines have been shown to be effective for delirium caused by alcohol withdrawal, and may have fewer adverse effects than do other agents that are used for withdrawal-associated delirium, according to the review.

However, in an extensive search of the literature in major medical databases and other sources, the researchers from the Cochrane Dementia and Cognitive Improvement Group found only one randomized, controlled, double-blind study of benzodiazepine for delirium not related to alcohol withdrawal (Cochrane Database Syst. Rev. 2009 [doi:10.1002/14651858.CD006379.pub2]).

In that trial's intention-to-treat analysis, the use of a selective [alpha.sub.2]-adrenergic receptor agonist (dexmedetomidine) resulted in better outcomes for delirious ICU patients who were on mechanical ventilation, compared with the use of a benzodiazepine (lorazepam).

Intravenous dexmedetomidine was administered in 52 patients (average age, 60 years; 30 men) and intravenous lorazepam in 51 patients (average age, 59 years; 23 men). There were no significant differences between the two groups in severity of illness, duration of ventilation, or admission diagnosis.

The lorazepam patients had an average of 7 delirium-free days and 8 coma-free days, compared with 9 days and 10 days in the dexmedetomidine patients.

The patients who received the benzodiazepine had an average of 3 days free of both delirium and coma, compared with 7 in the patients who received the [alpha.sub.2]-adrenergic receptor agonist. Coma prevalence was 92% in patients who received lorazepam, compared with 63% in those who received dexmedetomidine (JAMA 2007; 298:2644-53).

The 28-day mortality was not significantly different between the two groups.

The better outcomes with dexmedetomidine in patients on mechanical ventilation, in whom the incidence of delirium is very high, are not generalizable to the total population of palltients with delirium, the reviewers noted.

Two other studies comparing a benzodiazepine with neuroleptics for delirium were partially controlled and were not included in the analysis. One of these trials halted the benzodiazepine arm early because of unacceptable adverse effects, including oversedation.

In the prospective, randomized study that was halted early, 6 hospitalized AIDS patients received lorazepam, 13 received chlorpromazine, and 11 were treated with haloperidol. The patients on lorazepam showed no improvement in delirium-associated confusion and had more adverse effects than did the patients who received one of the neuroleptics (Am. J. Psychiatry 1996;153:231-7).

In the other randomized study, 48 nursing home patients received either a benzodiazepine (alprazolam) or a neuroleptic (haloperidol) for agitation associated with delirium; there was no significant difference in outcomes with the two drugs. This study was not adequately blinded and the patients' diagnoses were not clearly categorized, according to the reviewers (J. Am. Geriatr. Soc. 1998;46:620-5).

"At this time benzodiazepines cannot be recommended" for treatment of delirium in hospitalized patients, and "further research is required," the investigators concluded.

No conflicts of interest were stated.