Effect of acute pain on the association between preoperative cognitive impairment and postoperative delirium: a secondary analysis of three trials

The association between preoperative cognitive impairment, postoperative pain, and postoperative delirium in older patients after noncardiac surgery is not known.This was a secondary analysis of datasets from three previous studies. Patients aged ≥55 yr who underwent elective noncardiac surgery were enrolled. Preoperative cognitive impairment was defined as Mini-Mental State Examination <27. Pain intensity with movement was assessed using an 11-point numeric rating scale at 12-h intervals during the first 72 h after surgery; time-weighted average (TWA) pain score was calculated. Primary outcome was the occurrence of delirium within the first 5 postoperative days. Mediation analysis was used to investigate the relationships between cognitive impairment, pain score, and delirium.A total of 1497 patients were included. Prevalence of preoperative cognitive impairment was 40.3% (603/1497). Patients with cognitive impairment suffered higher TWA pain score within 72 h (4 [3-5] vs 3 [2-5], P=0.004) and more delirium within 5 days (12.9% [78/603] vs 4.9% [44/894], P<0.001) when compared with those without. Total and direct associations between cognitive impairment and delirium were (adjusted β) 8.3% (95% confidence interval [CI], 4.8-12.0%; P<0.001) and 7.8% (95% CI, 4.4-12.0%; P<0.001), respectively. A significant indirect association with acute pain was observed between cognitive impairment and delirium (adjusted β=0.4%; 95% CI, 0.1-1.0%; P=0.006), accounting for 4.9% of the total effect.The association between preoperative cognitive impairment and delirium is significantly mediated by acute pain in patients after noncardiac surgery. Considering the small effect size, clinical significance of this mediation effect requires further investigation.