Post-operative delirium: predictors and prognosis in elderly orthopedic patients

To compare the effect of post-operative analgesia using epidural versus intravenous infusions on the incidence of delirium after bilateral knee replacement surgery in elderly patients. Additional risk factors and impact on post-operative recovery were also assessed.

Prospective randomized controlled trial.

Urban referral hospital specializing in elective orthopedic surgery.

60 consecutive patients undergoing bilateral knee replacement surgery with epidural anesthesia were approached; 51 patients were eligible and consented. The mean age was 68, 55% were women, and there was a high prevalence of comorbid medical disease. No patient was demented pre-operatively.

Random allocation to either continuous epidural infusion of bupivacaine and fentanyl or continuous intravenous infusion of fentanyl. Infusions were initiated at the first complaint of pain and continued through the 36- to 48-hour stay in the recovery room.

Acute post-operative delirium defined using an algorithm based on DSM III criteria.

The overall incidence of acute delirium was 41%, with no difference between types of post-operative analgesia. Predictors of delirium were age, gender, and pre-operative alcohol use. All cases resolved within 1 week, and length of stay and achievement of physical therapy goals were the same for delirious and non-delirious patients.

There is a high incidence of post-operative delirium in elderly non-demented patients following bilateral knee replacement, regardless of whether post-operative analgesia is administered by the epidural or intravenous route.