Implementation of a delirium assessment tool in the ICU can influence haloperidol use

INTRODUCTION In critically ill patients, delirium is a serious and frequent disorder that is associated with a prolonged intensive care and hospital stay and an increased morbidity and mortality. Without the use of a delirium screening instrument, delirium is often missed by ICU nurses and physicians. The effects of implementation of a screening method on haloperidol use is not known. The purpose of this study was to evaluate the implementation of the confusion assessment method-ICU (CAM-ICU) and the effect of its use on frequency and duration of haloperidol use. METHODS We used a tailored implementation strategy focused on potential barriers. We measured CAM-ICU compliance, interrater reliability, and delirium knowledge, and compared the haloperidol use, as a proxy for delirium incidence, before and after the implementation of the CAM-ICU. RESULTS Compliance and delirium knowledge increased from 77% to 92% and from 6.2 to 7.4, respectively (both, P < 0.0001). The interrater reliability increased from 0.78 to 0.89. More patients were treated with haloperidol (9.9% to 14.8%, P < 0.001), however with a lower dose (18 to 6 mg, P = 0.01) and for a shorter time period (5 [IQR:2-9] to 3 [IQR:1-5] days, P = 0.02). CONCLUSIONS With a tailored implementation strategy, a delirium assessment tool was successfully introduced in the ICU with the main goals achieved within four months. Early detection of delirium in critically ill patients increases the number of patients that receive treatment with haloperidol, however with a lower dose and for a shorter time period.