Impaired Cognitive Domains in Surgical Patients: A Systematic Review and Meta-Analysis.
Subin Park, Ellene Yan, Ray Martinez-Rodriguez et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis examined preoperative cognitive impairment, perioperative changes, and associated postoperative outcomes in adults undergoing surgery. It included 21 studies with 5,725 patients, spanning cardiac and non-cardiac procedures. Of six cognitive domains assessed before surgery, pooled impairment prevalence was highest for executive function at 18% (95% CI, 13%-24%), followed by visuospatial function at 16% (95% CI, 6%-26%) and attention/working memory/processing speed at 14% (95% CI, 9%-18%). Postoperative impairment in assessed domains was common at 1 week and lower within 3 months.
Key findings
- Preoperative impairment was most prevalent in executive function: 18% (95% CI, 13%-24%).
- Preoperative impairment prevalence was 16% (95% CI, 6%-26%) for visuospatial function and 14% (95% CI, 9%-18%) for attention/working memory/processing speed.
- At 1 week postoperatively, impairment prevalence was 35% (95% CI, 4%-66%) for attention/working memory/processing speed and 34% (95% CI, 16%-51%) for executive function.
- Within 3 months, pooled impairment prevalence was 16% (95% CI, 3%-35%) for attention/working memory/processing speed and 15% (95% CI, 6%-24%) for executive function.
Why it matters
Cognitive impairment before surgery was not evenly distributed across domains, and postoperative impairment was frequent in the early period assessed. These findings suggest that broad cognitive screening may miss domain-specific difficulties. Understanding which domains are commonly affected can inform the design of perioperative cognitive assessments, although the review does not establish that assessment improves outcomes.
What to keep in mind
The review included only studies with at least 100 patients and excluded cross-sectional, case-control, and case-series designs. Postoperative prevalence estimates had wide confidence intervals, and the authors state that further research is needed to establish the clinical utility of domain-specific assessment.