Table 4. Risk factors for incident delirium in patients with IQCODE <3.44, n = 200.
| Characteristic | No incident delirium | Incident delirium |
Univariate models | Intermediate multivariate models | Final multivariate model | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| N = 166 (83%) | N = 34 (17%) | OR | 95% CI | P-valuea | OR | 95% CI | OR | 95% CI | P-valuea | |
| Background data | ||||||||||
| Age > 82 years, n/N (%) | 75/166 (45) | 21/34 (62) | 2.0 | 0.9–4.2 | 0.08 | |||||
| NEADL < 45 points, n/N (%) | 33/162 (20) | 14/33 (42) | 2.9 | 1.3–6.3 | 0.010 | 2.9 | 1.2–6.9 | 2.6 | 1.0–6.7 | 0.045 |
| BMI < 20.0 kg/m2, n/N (%) | 20/145 (14) | 13/33 (39) | 4.1 | 1.7–9.4 | 0.002 | 4.5 | 1.8–11.1 | 6.5 | 2.3–18.7 | 0.001 |
| Time admission to surgery > 24 hours, n/N (%) | 77/164 (47) | 11/33 (33) | 0.6 | 0.3–1.2 | 0.15 | |||||
| Admission and preoperative data | ||||||||||
| ASA group III, IV or V, n/N (%) | 54/158 (43) | 21/32 (66) | 3.7 | 1.7–8.2 | 0.001 | 3.5 | 1.6–7.8 | 3.0 | 1.2–7.9 | 0.023 |
| CRP > 10 mg/L, n/N (%) | 50/165 (30) | 16/34 (47) | 2.0 | 1.0–4.3 | 0.07 | |||||
| MAP at admission, mmHg, median (IQR) | 106 (97–115) | 111 (100–123) | 0.18 | |||||||
| HR at admission, bpm, median (IQR) | 81 (69–92) | 82 (68–93) | 0.34 | |||||||
| Per- and postoperative data | ||||||||||
| Type of anaesthesia | ||||||||||
| General anaesthesia | 15/163 (9) | 2/33 (6) | 0.6 | 0.1–2.9 | 0.56 | |||||
| Spinal or epidural anaesthesia | 148/163 (91) | 31/33 (94) | ||||||||
| Duration of anaesthesia, in quartiles | 0.74 | |||||||||
| Q1 | 39/160 (24) | 6/33 (18) | ||||||||
| Q2 | 45/160 (28) | 8/33 (24) | 1.2 | |||||||
| Q3 | 39/160 (24) | 9/33 (27) | 1.5 | |||||||
| Q4 | 37/160 (23) | 10/33 (30) | 1.8 | |||||||
| Being in Q4 vs Q1-Q3 | 1.4 | 0.6–3.3 | 0.38 | |||||||
| Received benzodiazepine iv perioperatively, n/N (%) | 72/166 (43) | 12/34 (35) | 0.45 | |||||||
| Lowest MAP during anaesthesia, mmHg, median (IQR) | 67 (57–78) | 63 (55–72) | 0.24 | |||||||
| Difference in MAP from admission to lowest value during anaesthesia | 0.44 | |||||||||
| ≤ 30 mmHg, n/N (%) | 58/162 (36) | 7/33 (21) | ||||||||
| 31–40 mmHg, n/N (%) | 27/162 (17) | 6/33 (18) | 1.8 | |||||||
| 41–50 mmHg, n/N (%) | 25/162 (15) | 6/33 (18) | 2.0 | |||||||
| > 50 mmHg, n/N (%) | 52/162 (32) | 14/33 (42) | 2.2 | |||||||
| Pressor used during surgery, n/N (%) | 72/164 (44) | 20/34 (59) | 1.8 | 0.9–3.9 | 0.13 | |||||
| MAP postoperatively, mmHg, median (IQR) | 83 (75–95) | 81 (71–87) | 0.43 | |||||||
| HR postoperatively, beats per minute | 0.51 | |||||||||
| < 66 bpm, n/N (%) | 33/162 (20) | 5/32 (16) | ||||||||
| 66–75 bpm, n/N (%) | 39/162 (24) | 7/32 (22) | 1.2 | |||||||
| 76–85 bpm, n/N (%) | 48/162 (30) | 8/32 (25) | 1.1 | |||||||
| 86–95 bpm, n/N (%) | 24/162 (15) | 9/32 (28) | 2.5 | |||||||
| > 95 bpm, n/N (%) | 18/162 (11) | 3/32 (9) | 1.1 | |||||||
| SpO2% postoperatively, median (IQR) | 96 (95–98) | 96 (93–97) | 0.061 | 0.8 | 0.7–1.0 | |||||
| Received ≥ 2 blood transfusions (SAG units), n/N (%) | 46/161 (29) | 20/34 (59) | 3.6 | 1.7–7.7 | 0.001 | 3.4 | 1.5–7.6 | 3.2 | 1.3–7.8 | 0.012 |
The intermediate multivariate models consist of variables from the same main category (Background, Admission, Per- and Postoperative), and variables that turned out to be significant in these analyses were candidates for the final model.
IQCODE = Informant Questionnaire in Cognitive Decline in the Elderly; OR = Odds Ratio; CI = Confidence Interval; NEADL = Nottingham Extended Activities of Daily Living; BMI = Body Mass Index; ASA = American Society of Anaesthesiologists; CRP = C -reactive protein; MAP = Mean Arterial Pressure; IQR = Interquartile range; HR = Heart Rate; bpm = beats per minute
Missing data (N) for several variables: NEADL: N = 5, BMI: N = 22, Time to surgery: N = 3, ASA group: N = 10, CRP: N = 1, Type of anaesthesia: N = 4, Duration of anaesthesia: N = 7, Difference in MAP from admission to lowest value during anaesthesia: N = 5, Pressor used during surgery: N = 2, HR postoperatively: N = 6, Blood transfusions: N = 5.
a Mann-Whitney tests for skewed continuous variables, p-values are 2-tailed. Logistic regression for categorical and ordinal variables.