Association of postoperative delirium with cognitive outcomes: A meta-analysis.
Level 3 - non-randomized controlled study
Systematic review and meta-analysis of prospective observational cohort studies
PubMed 34482263 · doi:10.1016/j.jclinane.2021.110496
What was done
Authors searched PubMed, Cochrane, and EMBASE through November 2020 for prospective cohort studies assessing the relationship between postoperative delirium (POD) and cognitive outcomes at 1 or more months after surgery in patients aged 60 and older. Short-term outcomes were defined as up to 12 months, and long-term outcomes as beyond 12 months. Two reviewers conducted study screening, data extraction, and quality assessment. Pooled effect sizes (Hedges g and odds ratios) were calculated using random- and fixed-effects models, with meta-regression applied to explore heterogeneity.
What was found
Across 18 included studies, POD was significantly associated with worse cognitive outcomes at ≥1 month post-surgery (Hedges g = 0.61, 95% CI 0.43 to 0.79, I² = 65.1%). Cognitive impairment was significantly elevated in both short-term (g = 0.46, 95% CI 0.24 to 0.68, I² = 53.1%) and long-term follow-up (g = 0.82, 95% CI 0.57 to 1.06, I² = 57.1%). POD was associated with increased risk of dementia (OR = 6.08, 95% CI 3.80 to 9.72, I² = 0%), attention impairment (OR = 1.74, 95% CI 1.13 to 2.68, I² = 0%), executive dysfunction (OR = 1.33, 95% CI 1.00 to 1.80, I² = 0%), and memory impairment (OR = 1.59, 95% CI 1.20 to 2.10, I² = 43.0%). Multivariate meta-regression identified patient age and the specific delirium measurement method as significant sources of heterogeneity.
Why it matters
This meta-analysis quantifies the trajectory of post-surgical cognitive decline, showing that postoperative delirium is linked not only to transient cognitive issues but also to sustained, accelerating long-term cognitive impairment and dementia risk in older adults.
Limits
Total participant count was not reported in the abstract. As a meta-analysis of observational cohort studies, it cannot establish whether POD causally drives neurodegeneration or serves as a marker of pre-existing brain vulnerability. Substantial heterogeneity was present in the primary outcome (I² = 65.1%), and variability in delirium assessment tools across included studies contributed to this inconsistency.
Cited by
- supports Patients who develop delirium are three times more likely to subsequently develop dementia.