BACKGROUND/OBJECTIVES
Delirium is a common postoperative complication associated with prolonged length of stay, hospital readmission, and premature mortality. We explored the association between neighborhood‐level characteristics and delirium incidence and severity, and compared neighborhood‐ with individual‐level indicators of socioeconomic status in predicting delirium incidence.
DESIGN
A prospective observational cohort of patients enrolled between June 18, 2010, and August 8, 2013. Baseline interviews were conducted before surgery, and delirium/delirium severity was evaluated daily during hospitalization. Research staff evaluating delirium were blinded to baseline cognitive status.
SETTING
Two academic medical centers in Boston, MA.
PARTICIPANTS
A total of 560 older adults, aged 70 years or older, undergoing major noncardiac surgery.
INTERVENTION
The Area Deprivation Index (ADI) was used to characterize each neighborhood’s socioeconomic disadvantage.
MEASUREMENTS
Delirium was assessed using the Confusion Assessment Method (CAM) long form. Delirium severity was calculated using the highest value of CAM Severity score (CAM‐S) occurring during daily hospital assessments (CAM‐S Peak).
RESULTS
Residing in the most disadvantaged neighborhoods (ADI > 44) was associated with a higher risk of incident delirium (12/26; 46%), compared with the least disadvantaged neighborhoods (122/534; 23%) (risk ratio (RR) (95% confidence interval (CI)) = 2.0 (1.3–3.1). The CAM‐S Peak score was significantly associated with ADI (Spearman rank correlation, ρ = 0.11; P = .009). Mean CAM‐S Peak scores generally rose from 3.7 to 5.3 across levels of increasing neighborhood disadvantage. The RR (95% CI) values associated with individual‐level markers of socioeconomic status and cultural background were: 1.2 (0.9–1.7) for education of 12 years or less; 1.3 (0.8–2.1) for non‐White race; and 1.7 (1.1–2.6) for annual household income of less than $20,000. None of these individual‐level markers exceeded the ADI in terms of effect size or significance for prediction of delirium risk.
CONCLUSIONS
Neighborhood‐level makers of social disadvantage are associated with delirium incidence and severity, and demonstrated an exposure‐response relationship. Future studies should consider contextual‐level metrics, such as the ADI, as risk markers of social disadvantage that can help to guide delirium treatment and prevention.