Postoperative delirium severity reflects baseline vulnerability more than measured postoperative exposures: A...

Postoperative delirium is commonly analyzed as a binary outcome or categorized by peak severity or duration. These approaches obscure the dynamic nature of delirium and limit insight into how perioperative physiologic stressors influence delirium severity over time.

We conducted a secondary analysis of a prospective postoperative delirium cohort with daily assessments using the Delirium Rating Scale (DRS). Time-varying postoperative opioid exposure (morphine milligram equivalents [MMEs]) and daily lowest oxygen saturation were examined in relation to daily delirium severity on postoperative days 0-4. Longitudinal associations were evaluated using regression models with cluster-robust standard errors at the subject level. Delirium burden was quantified using peak DRS scores and area under the DRS curve (AUC).

Among 206 patients (1660 subject-days), older age and higher ASA physical status were independently associated with greater daily delirium severity (β = 0.09/year; β = 0.88/class; both p < 0.01). In contrast, neither same-day nor lagged postoperative opioid exposure or oxygen saturation was independently associated with daily delirium severity after adjustment. Greater delirium burden was associated with age, ASA class, and intraoperative opioid exposure.

Postoperative delirium severity evolves dynamically and appears to be more strongly associated with baseline vulnerability than with the limited postoperative physiologic exposures measured in this study. These findings are consistent with vulnerability-stress models of delirium, additional perioperative and postoperative physiologic factors not captured in this dataset may also contribute substantially.