To evaluate the clinical effectiveness, safety, and cost of dexmedetomidine for the treatment of agitated delirium refractory to haloperidol in nonintubated critically ill patients.

Nonrandomized, controlled trial.

Intensive care department of a tertiary care nonprofit hospital.

All consecutive admissions to a medical-surgical ICU with a diagnosis of agitated delirium.

Initial haloperidol titration: all patients received IV bolus doses of haloperidol until agitation was controlled (Richmond Agitation Sedation Scale scoring range, 0 to -2) or reaching the maximum daily dose. Group comparison: patient responders to haloperidol (control group) were compared with nonresponders (dexmedetomidine group).

A total of 132 nonintubated patients were treated with haloperidol in the initial haloperidol titration phase. Forty-six patients (34.8%; 95% CI, 26.0-43.1%) did not respond to haloperidol, and 86 patients (65.2%; 95% CI, 56.3-73.0%) were responders. During the group comparison phase, dexmedetomidine achieved a higher percentage of time in satisfactory sedation levels than did haloperidol (92.7% [95% CI, 84.5-99.8%] vs 59.3% [95% CI, 48.6-69.3%], respectively; p = 0.0001). Haloperidol was associated with 10 cases (11.6% [95% CI, 6.5-21.2%]) of oversedation and two (2.0% [0.4-8%]) of corrected QT lengthening. Direct cost of dexmedetomidine was 17 times greater than haloperidol, but it achieved a mean savings of $4,370 per patient due to the reduction in length of ICU stay.

In the study conditions, dexmedetomidine shows to be useful as a rescue drug for treating agitation due to delirium in nonintubated patients in whom haloperidol has failed, and it seems to have a better effectiveness, safety, and cost-benefit profile than does haloperidol.