Prevalence and risk factors related to haloperidol use for delirium in adult intensive care patients: the multinational AID-ICU inception cohort study

Abstract

Purpose: We assessed the prevalence and variables associated with haloperidol use for delirium in ICU patients and explored any associations of haloperidol use with 90-day mortality. Methods: All acutely admitted, adult ICU patients were screened during a 2-week inception period. We followed the patient throughout their ICU stay and assessed 90-day mortality. We assessed patients and their variables in the first 24 and 72 h in ICU and studied their association together with that of ICU characteristics with haloperidol use. Results: We included 1260 patients from 99 ICUs in 13 countries. Delirium occurred in 314/1260 patients [25% (95% confidence interval 23–27)] of whom 145 received haloperidol [46% (41–52)]. Other interventions for delirium were benzodiazepines in 36% (31–42), dexmedetomidine in 21% (17–26), quetiapine in 19% (14–23) and olanzapine in 9% (6–12) of the patients with delirium. In the first 24 h in the ICU, all subtypes of delirium [hyperactive, adjusted odds ratio (aOR) 29.7 (12.9–74.5); mixed 10.0 (5.0–20.2); hypoactive 3.0 (1.2–6.7)] and circulatory support 2.7 (1.7–4.3) were associated with haloperidol use. At 72 h after ICU admission, circulatory support remained associated with subsequent use of haloperidol, aOR 2.6 (1.1–6.9). Haloperidol use within 0–24 h and within 0–72 h of ICU admission was not associated with 90-day mortality [aOR 1.2 (0.5–2.5); p = 0.66] and [aOR 1.9 (1.0–3.9); p = 0.07], respectively. Conclusions: In our study, haloperidol was the main pharmacological agent used for delirium in adult patients regardless of delirium subtype. Benzodiazepines, other anti-psychotics and dexmedetomidine were other frequently used agents. Haloperidol use was not statistically significantly associated with increased 90-day mortality.