Defining 'recovery' for delirium research: a systematic review.

Background: Delirium is a common neuropsychiatric disorder. The natural course is of an acute, fluctuating and often transient condition; however, accumulating evidence suggests that delirium can be associated with incomplete recovery. Despite a growing body of research, a lack of clarity exists regarding definition and outcomes.

Objective: To clarify the definition of recovery of delirium used in the literature.

Methods: A Medline search was performed using relevant keywords. Studies were included if they were in English, provided any definition of recovery and were longitudinal. Excluded articles were duplicated studies, case studies, review articles or articles related to alcohol, children, subsyndromal delirium only or those investigating core symptoms such as function.

Results: Fifty-six studies met the inclusion criteria. Only two studies used clinical criteria alone for the diagnosis of delirium; most studies used at least one validated scale-either categorical or continuous severity scales. A variety of 16 different terms were used to define the 'recovery of delirium'. The definitions of each term also varied. Studies using severity scales used either cut-off points or percentage reduction between assessments, while others using dichotomous scales (yes/no) defined recovery as one or more days of negative delirium.

Conclusion: An agreed terminology to define recovery in delirium is required. A distinction should also be made between symptomatic and overall recovery, as well as between long- and short-term outcomes. It is proposed that cognitive recovery should be central to defining recovery in delirium.