The consistent burden in published estimates of delirium occurrence in medical inpatients over four decades: a systematic review and meta-analysis study

Gibb, K.; Seeley, A.; Quinn, T.; Siddiqi, N.; Shenkin, S.; Rockwood, K.; Davis, D.

Introduction Delirium is associated with a wide range of adverse patient safety outcomes. We sought to identify if trends in healthcare complexity were associated with changes in reported delirium in adult medical patients in the general hospital over the last four decades. Methods We used identical criteria to a previous systematic review, including studies using DSM and ICD-10 criteria for delirium diagnosis. Random effects meta-analysis pooled estimates across studies, meta-regression estimated temporal changes, funnel plots assessed publication bias. Results Overall delirium occurrence was 23% (95% CI 19%-26%) (33 studies). There was no change between 1980-2019, nor was case-mix (average age of sample, proportion with dementia) different. There was evidence of increasing publication bias over time. Discussion The incidence and prevalence of delirium in hospitals appears to be stable, though publication bias may mask true changes. Nonetheless, delirium remains a challenging and urgent priority for clinical diagnosis and care pathways.