Antipsychotics in the treatment of delirium in older hospitalized adults: a systematic review.

OBJECTIVES: To examine evidence of the efficacy of antipsychotics in the treatment of delirium in older hospitalized adults.

DESIGN: Systematic literature review.

SETTING: Hospital.

PARTICIPANTS: Older adults with delirium.

MEASUREMENTS: The MEDLINE (January 1980-December 2010) and Cochrane Databases were searched using the keywords "delirium" and "antipsychotics." References of review articles were reviewed to identify additional studies. Study selection criteria included prospective design, more than 10 participants (in treatment arms), mean age 60 and older, standardized criteria for diagnosing delirium, and validated delirium rating scales for reporting outcomes.

RESULTS: Thirteen articles met selection criteria: six single-agent and seven comparison studies. Of these, eight (62%) had fewer than 25 participants in treatment arms, 10 (77%) recruited participants from psychiatry referrals, and eight (62%) did not clearly describe their screening methods. All single-agent studies were open-label studies. Of the comparison studies, five (71%) used randomization, but only one of these (a placebo-controlled study) used adequate allocation concealment methods, and only one other study (comparing two antipsychotics) described a double-blind method in detail. In the only placebo-controlled study (which was stopped early), no statistically significant differences in mean delirium severity scores were found at individual time points (Days 2, 3, 4, 7, 10). The other 12 studies reported improvements in delirium severity or resolution of delirium based on cutoff scores of the scales, but it is not clear from any of these studies what the natural course of delirium would have been without use of antipsychotics.

CONCLUSION: Because of severe methodological limitations, the studies in this review do not support the use of antipsychotics in the treatment of delirium in older hospitalized adults. Additional well-designed randomized placebo-controlled trials are needed.