The predictors and independent outcome association of delirium after cardiac surgery are important and yet poorly characterised.

Methods

We performed a retrospective observational study of cardiac surgery patients between January 2009 and March 2016. We defined delirium using ICD-10 diagnostic codes. Multivariable analysis was conducted to find independent associations between baseline variables, delirium, and key clinical outcomes.

Results

We studied 2,447 study patients (28.7% female, median age was 66 [IQR 57–74] years). Delirium was coded for in 12.9% of patients overall, and in 22.9% of those aged >75 years. Increasing age, Charlson co-morbidity index, admission not from home, peripheral vascular disease, respiratory disease, preoperative atrial fibrillation, duration of cardiopulmonary bypass and nature of surgery were all independent predictors of delirium. Delirium was independently and strongly associated with increased risk of reintubation (OR 8.18 [95% CI 5.24–12.78]), tracheostomy (OR 10.44 [95% CI 5.91–18.45]), and increased length of stay by 113.7 [95% CI 99.7–127.7] ICU hours and 6.95 [95% CI 5.94–7.95] hospital days, but not 30-day mortality (OR 0.78 [95% CI 0.38–1.59]; p = 0.5).

Conclusions

Delirium is common in cardiac surgery patients and increases with age. Delirium was the strongest predictor of reintubation, need for tracheostomy, and prolongation of intensive care unit (ICU) and hospital length of stay. Delirium prevention and attenuation are a priority in cardiac surgery patients.