Persistent delirium in older hospital patients: a systematic review of frequency and prognosis

Background: one explanation for the poor prognosis of delirium among older hospital patients may be that many of these patients do not recover from delirium. We sought to determine the frequency and prognosis of persistent delirium (PerD) in older hospital patients by systematically reviewing original research on this topic. Methods: MEDLINE, EMBASE, PsycINFO and the Cochrane Database of Systematic Reviews were searched for potentially relevant articles. The bibliographies of relevant articles were searched for additional references. Eighteen reports (involving 1,322 patients with delirium) met the following seven inclusion criteria: original research published in English or French, prospective study design, study population of at least 20 hospital patients, patients aged 50 years or more, follow-up of at least 1 week, acceptable definition of delirium at enrolment and included at least one assessment for PerD at discharge or later. The methods of each study were assessed according to the six criteria for prognostic studies described by the Evidence-Based Medicine Working Group. Information about the sample origin and size, age, proportion with dementia, criteria for delirium, timing of follow-up assessments, criteria for PerD, proportion with PerD and prognosis of PerD was systematically abstracted from each report, tabulated and combined using standard meta-analysis techniques. Results: the combined proportions with PerD at discharge, 1, 3 and 6 months were 44.7% (95% CI 26.8%, 63.7%), 32.8% (95% CI 18.4%, 47.2%), 25.6% (95% CI 7.9%, 43.4%) and 21% (95% CI 1.4%, 40.6%), respectively. The outcomes (mortality, nursing home placement, function, cognition) of patients with PerD were consistently worse than the outcomes of patients who had recovered from delirium. Conclusion: PerD in older hospital patients is frequent, appears to be associated with adverse outcomes and may account for the poor prognosis of delirium in this population. These findings have potentially important implications for clinical practice and research.