In 2005, approximately 2.7 million Americans aged 65 and older spent at least 1 day in the intensive care unit (ICU), costing Medicare an estimated $27.5 billion. It is estimated that up to 80% of these older ICU patients had delirium, an often unrecognized acute brain failure that is an independent predictor of morbidity and mortality, during their hospital stay. Older adults with delirium are more prone to falls, injuries, pressure ulcers, and restraints, complications that may also contribute to prolonged ICU and hospital length of stay, higher mortality rates, poorer functional status, limited rehabilitation, increased institutionalization, and higher health care costs. The literature supports treatment with a reduction in the use of benzodiazepines and anticholinergics and the use of low-dose neuroleptics such as haloperidol. However, there have been no randomized controlled trials evaluating the efficacy of this approach on reducing delirium severity, duration, and complications.

The e-CHAMP study ("Enhancing Care for Hospitalized Older Adults With Memory Problems"; see NCT00182832) tested the effectiveness of cognitive screening coupled with computerized decision support in reducing delirium and other hospital-related complications among 424 older adults hospitalized on medical wards. It found that many of the older adults entering the study had already experienced delirium in the ICU prior to their transfer to the wards.

This study will test a pharmacologic intervention that allows a more targeted pharmacologic approach to the care of older adults with delirium, while still recognizing the clinicians' role in controlling symptoms and providing intensive care. The hypothesis is that patients in the intervention arm, compared to those receiving usual care, will have reduced delirium severity, as measured by the Delirium Rating Scale (DRS-R-98), at 1 week following randomization or hospital discharge; fewer hospital days with delirium or coma, as determined by the Confusion Assessment Method in the ICU (CAM-ICU); shorter hospital lengths of stay; and lower ICU, hospital, and 30-day mortality.