Postoperative delirium is a common complication in elderly patients after surgery. Its occurrence is associated with worse outcomes. The pathophysiology of delirium remains poorly understood. However, an universal phenomenon is that delirium frequently occurs in elderly patients after major complicated surgery, but is rarely seen after minor ambulatory surgery (such as cataract surgery). This indicates that stress response produced by surgery might have an important role in the pathogenesis of delirium. It has been reported that, when compared with general anesthesia and postoperative intravenous analgesia, neuraxial anesthesia and analgesia reduced the occurrence of postoperative complications and mortality in high risk patients. Combined epidural-general anesthesia is frequently used in clinical practice. This anesthetic method provides advantages of both epidural and general anesthesia, i.e. it blocks the afferent pathway of nociceptive stimulus by neuraxial blockade during and after surgery, and allows patients to endure long-duration surgery without any awareness. The investigators hypothesize that combined epidural-general anesthesia and postoperative epidural analgesia can decrease the incidence of delirium in elderly patients after major surgery when compared with general anesthesia alone and postoperative intravenous analgesia.

Delirium is an acutely occurred and transient mental syndrome characterized by global impairment of cognitive functions, reduced level of consciousness, abnormalities of attention, increased or decreased psychomotor activity, and disordered sleep-wake cycle. Postoperative delirium is a common complication in elderly patients after surgery. Dyer et al reviewed 80 primary studies and found that the mean incidence of postoperative delirium is about 36.8% (range 0%-73.5%) after surgery. It occurs in up to 80% of patients in the intensive care unit (ICU). Our recent studies found that delirium occurred in 51.0% of patients after cardiac surgery and in 44.5% of patients after non-cardiac surgery.

The occurrence of postoperative delirium is associated with worse outcomes. Studies showed that delirious patients have prolonged ICU stay, increased incidence of complications, prolonged hospitalization, high mortality rate, and increased health care costs. Delirium is also associated with increased risk of long-term cognitive decline and poor quality of life. A recent follow-up study (mean follow-up time 27.9 ± 3.1 months) by our research group found that, after adjusting factors such as age, occurrence of postoperative complications, and stage of cancer, etc, the occurrence of postoperative delirium still remained an independent predictor of long-term mortality.

The pathophysiology of delirium remains poorly understood. An universal phenomenon is that delirium frequently occurs in elderly patients after major complicated surgery, but is rarely seen after minor ambulatory surgery (such as cataract surgery). Studies also found that postoperative pain is an independent risk factor of delirium, whereas effective pain relief may help to reduce the incidence of delirium. Our recent studies showed that high serum cortisol level is an independent risk factor of postoperative delirium. In addition, inflammatory response may also contribute to the pathogenesis of delirium. Trauma, pain, cortisol secretion and inflammation are all important components of surgical stress response. The above results indicated that stress response produced by surgery might have an important role in the pathogenesis of delirium.

Previous studies demonstrated that, when compared with general anesthesia, neuraxial anesthesia attenuates the hypersecretion of cortisol, and decreases the intensity of inflammatory response more effectively after surgery. And epidural analgesia provides better postoperative pain relief than intravenous analgesia. It was also reported that, when compared with general anesthesia and intravenous analgesia, neuraxial anesthesia and analgesia reduces the occurrence of postoperative complications and mortality in high risk patients. Combined epidural-general anesthesia is frequently used in clinical practice, and is performed in about 1/4 of patients undergoing surgery in the applicant's hospital. Theoretically, this anesthetic method provides advantages of both epidural and general anesthesia, i.e. it blocks the afferent pathway of nociceptive stimulus by neuraxial blockade during and after surgery, and allow patients to endure long-duration complicated surgeries without any awareness. However, there is no evidence whether combined epidural-general anesthesia/postoperative epidural analgesia can decrease the incidence of postoperative delirium in elderly patients undergoing major surgery.

The objective of the study is to compare the effects of combined epidural-general anesthesia/postoperative epidural analgesia and general anesthesia/postoperative intravenous analgesia on the incidence of postoperative delirium in elderly patients undergoing major noncardiac surgery.

Inclusion criteria:

1. elderly patients (age range 60-90 years);
2. scheduled to undergo noncardiac thoracic or abdominal surgery with an expected duration of 2 hours or longer. For those who undergo thoracoscopic or laparoscopic surgery, the expected length of incision must be 5 centimeters or more;
3. agree to receive patient-controlled postoperative analgesia.

Exclusion criteria (patients who meet any of the following criteria will be excluded):

1. previous history of schizophrenia, epilepsy or Parkinson disease, or unable to complete preoperative assessment due to severe dementia, language barrier or end-stage disease;
2. history of myocardial infarction within 3 months before surgery;
3. any contraindication to epidural anesthesia and analgesia, including abnormal vertebral anatomy, previous spinal trauma or surgery, severe chronic back pain, coagulation disorder (prothrombin time or activated partial prothrombin time longer than 1.5 times of the upper limit of normal, or platelet count of less than 80 × 10^9/L), local infection near the site of puncture, and severe sepsis;
4. severe heart dysfunction (New York Heart Association functional classification 3 or above), hepatic insufficiency (Child-Pugh grades C), or renal insufficiency (serum creatinine of 442 μmol/L or above, with or without serum potassium of 6.5 mmol/L or above, or requirement of renal replacement therapy); or
5. any other conditions that were considered unsuitable for study participation.

Data will be provided on request.