Risk Factors for Postoperative Cognitive Dysfunction Following Multilevel Lumbar Spinal Fusion

STUDY DESIGN: Retrospective observational cohort. OBJECTIVES: To determine the incidence of and risk factors for postoperative cognitive dysfunction (POCD) following multilevel spine fusion. SUMMARY OF BACKGROUND DATA: POCD is a serious, common and under-recognized complication in elderly patients undergoing surgery. Risk factors for POCD vary in the literature and include preoperative as well as intraoperative factors. MATERIALS AND METHODS: A retrospective cohort of 566 thoracolumbar fusion cases with a minimum of 4 surgical levels were identified. A chart review was performed for the occurrence of POCD and known risk factors for POCD. Anesthetic and surgical data included operative time, fluid volume, blood loss, blood product replacement, and use of vasopressors. Arterial line-based mean arterial pressure (MAP) data were collected at 1-minute intervals, and the cumulative duration of MAP <65 mm Hg was recorded. Univariate and multivariate statistical analysis was used to investigate the relationship between demographic, preoperative and intraoperative risk factors, and the occurrence of POCD. RESULTS: Overall, 70 out of 566 patients (12.4%) experienced POCD described most commonly as encephalopathy (57, 81%), delirium (8, 11%), hallucinations (3, 4%), and altered mental status (2, 3%). Patients who developed POCD were older (68.7 yr vs . 59.6 yr, P <0.001), had a worse ASA scores (2.9 vs . 2.7, P =0.004), higher BMI (32.0 vs . 30.0, P =0.007), a higher incidence of diabetes (31% vs . 16%, P =.002), and sleep apnea (47% vs . 28%, P =0.002). Intraoperatively, patients who experienced POCD had greater fluid shifts and hemodynamic instability in terms of blood loss (800 mL vs . 660 mL, P =0.047), blood transfusion (350 mL vs . 201 mL, P =0.014), minutes of intraoperative hypotension (11.6 vs . 6.4, P =0.043) and vasopressor use (10604 mcg vs . 6823 mcg, P =0.029). CONCLUSIONS: Postoperative cognitive dysfunction is associated with age, preoperative comorbidities, and greater intraoperative hemodynamic instability. These factors present targets for optimization before surgery to lower the incidence of postoperative cognitive dysfunction.