Incidence and Prevalence of Delirium Subtypes in an Adult ICU: A Systematic Review and Meta-Analysis*

Objectives: Use systematic review and meta-analytic methodology to estimate the pooled incidence, prevalence, and proportion of delirium cases for each delirium subtype (hypoactive, hyperactive, and mixed) in an adult ICU population. Data Sources: We conducted a search of the MEDLINE, EMBASE, CINAHL, SCOPUS, Web of Science, and PsycINFO databases following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses standards from database inception until October 22, 2017, with no restrictions. Study Selection: We included original research conducted in adults admitted to any medical, surgical, or speciality ICU that reported incidence or prevalence estimates of delirium according to delirium subtype. Data Extraction: Data were extracted on sample size, population demographics, condition information, and reported delirium estimates. Data Synthesis: Forty-eight studies (27,342 patients; 4,550 with delirium) with an overall pooled prevalence of 31% (95% CI, 24–41; I 2 = 99%) met inclusion criteria. The pooled incidence ( n = 18 studies) of delirium subtypes were hyperactive (4% [95% CI, 2–6]; I 2 = 92%]), hypoactive (11% [95% CI, 8–17; I 2 = 97%]), and mixed (7% [95% CI, 4–11; I 2 = 97%]). The pooled prevalence ( n = 31 studies) of delirium subtypes were hyperactive (4% [95% CI, 3–6; I 2 = 94%]), hypoactive (17% [95% CI, 13–22; I 2 = 97%]), and mixed (10% [95% CI, 6–16; I 2 = 99%]). The pooled prevalence of hypoactive delirium in study populations with a similarly high severity of illness or mechanically ventilated was higher (severity of illness: 29% [95% CI, 18–46%; I 2 = 95%], 100% mechanically ventilated: 35% [95% CI, 23–55%; I 2 = 93%]) compared with the pooled prevalence of hypoactive delirium. Conclusions: Despite significant heterogeneity between studies, these data show the majority of delirious ICU patients to have hypoactive delirium, a finding with potential monitoring, management, and prognostic implications. The prevalence of hypoactive delirium varies between-study populations and is higher in patients with greater severity of illness.