Preoperative Beck Depression Inventory data were available from 320 of the original 460 patients. Of these patients, complete cognitive domain endpoints were available for 88–98% at baseline and 69–79% at both baseline and the one-month follow-up testing session. This range in end-points data was due to variability in the availability of specific neuropsychological test results between patients. A patient flow diagram of the study is shown in Figure 1. Depression was present at the time of surgery in 50 (15.6%) of the patients of whom 32 (10%) had mild depression, 8 (2.5%) moderate depression, and 10 (3.1%) severe depression. The characteristics of the entire cohort and for those with a Beck Depression Inventory score ≤ 13 and > 13 are listed in Table 1. Of note, there were no differences in patient characteristics for those with Beck Depression Inventory data compared to those with missing data (supplemental Table 1). Patients with preoperative depression were more likely to have had a history of prior stroke, chronic obstructive lung disease, and a lower level of attained education compared with patients without depression. There were no other differences between depressed and non-depressed patients for other variables including operative data or the cerebral autoregulation metrics during cardiopulmonary bypass. Depression was present in 43 (13.4%) of the 32 patients with Beck Depression Inventory data at the one month postoperative testing period with 34 (79.1%), 8 (18.6%), and 1 (2.3%) demonstrating mild, moderate, or severe depression, respectively. Of the 50 depressed patients before surgery, 34 had postoperative Beck Depression Inventory data, and 17 (50%) of those individuals remained depressed one month after surgery. Fourteen (6.6%) patients not demonstrating depression before surgery were now demonstrating depression at this latter time point.
The relationship between depression before surgery and postoperative neuropsychological test results by specific cognitive domain are listed in Table 2. There was no relationship between having evidence of depression before surgery and change in postoperative neuropsychological test results adjusted for sex, age, level of education, preoperative cognitive testing results, parent study randomized blood treatment group, and chronic medical conditions (Table 2). These findings were generally consistent in sensitivity analyses with multiple imputations to account for missing data (supplemental Table 2). In contrast to preoperative depression, there was a significant relationship between being classified as having depression one month after surgery with several domain-specific neuropsychological test results after surgery (Table 3). Patients with depression one month after surgery had significantly poorer performance on tests of attention (p = 0.030), verbal fluency (p = 0.011), processing speed (p = 0.044), and fine motor speed (p = 0.030). Additionally, when considered as a composite cognitive outcome, there was no difference in the risk of neurocognitive dysfunction one month after surgery from baseline between patients with and without preoperative depression (Figure 2). Neurocognitive dysfunction one month after surgery was observed in 16.2% of patients with preoperative depression and 15.5% of patients without preoperative depression (p = 0.741). In contrast, there was a significant association between postoperative depression and neurocognitive dysfunction one month after surgery (Figure 2) occurring in 33.3% of patients with postoperative depression compared to 14.5% of patients without postoperative depression (p = 0.025).
Percentage of patients with postoperative neurocognitive dysfunction one month after surgery based on preoperative or postoperative depression. The number above each column represents the number of patients with or without depression at each respective perioperative testing period. Beck Depression Inventory data were available from 320 patients prior to surgery of which 50 (15.6%) demonstrated depression. After surgery, Beck Depression Inventory data were available from322 patients of which 43 (13.4%) demonstrated depression. Postoperative neurocognitive dysfunction occurred in 16.2% of patients with preoperative depression and 15.5% without preoperative depression (p = 0.777). Postoperative neurocognitive dysfunction occurred in 33.3% and 14.5% of patients with and without postoperative depression (p = 0.040)
Quality of life measures one month after surgery for patients with and without preoperative depression are shown in Table 4A. Compared with patients without depression, patients with preoperative depression had lower self-ratings of energy/fatigue, emotional well-being, and general perception of their health. Patients with preoperative depression reported higher levels of state and trait anxiety one month after surgery compared with non-depressed patients. Postoperative quality of life measures were adversely affected by depression detected in the one month testing period Table 4B. All but physical health limits were statistically negatively associated with postoperative depression. There was no statistically significant difference in the frequency of clinically detected delirium in patients with and without preoperative depression (16% versus 7.4%, p = 0.158). No other significant associations were detected with clinical outcomes (results not shown).
BMC Anesthesiol. 2022;22(157) © 2022 BioMed Central, Ltd.