Background: Postoperative complications are integral to cardiac surgery. The most serious ones are stroke, which develops in about 7.5% of the patients, and postoperative encephalopathy, which affects 10–30% of the patients. According to bibliographical data, the number of complications is increasing.

Aim: To analyse the risk factors and the types of neurological complications in patients undergoing heart surgery.

Methods: We assessed retrospectively 323 consecutive patients undergoing surgery at the Department of Cardiac Surgery, University Teaching Hospital, Medical University of Bialystok, Poland, between July 2007 and June 2008. Group 1 comprised patients without neurological complications (n = 287; 89%) and Group 2 consisted of patients with neurological complications (n = 36; 11%). Our analysis included the following: preoperative status (age, sex, co-morbidities), intraoperative course (surgery type, duration of cardiopulmonary bypass [CPB], duration of aortic cross-clamping, types of medications administered, necessity of reinfusion from the cardiotomy reservoir and the necessity of tranexamic acid infusion) and the postoperative course (time to regaining consciousness, duration of mechanical ventilation, development of complications, types of complications). The results were then analysed statistically: arithmetic means and standard deviations were calculated for quantitative variables and the quantitative and percenage distributions were calculated for qualitative variables. The between- group comparisons of the quantitative variables were carried out using the t-Student test, while the qualitative variables were compared using the χ² test. The variables that proved significant in the univariate comparisons were included in the multivariate model. Regression analysis was the final step of the analysis of the risk factors for neurological complications. Based on the analysis of the ROC curve we calculated the cutoff values for the continuous variables. We calculated odds ratios with their 95% confidence intervals. P values of less than 0.05 were considered statistically significant.

Results: Among the 36 patients in Group 2, postoperative encephalopathy developed in 22 patients, transient ischaemic attacks in 7 patients, ischaemic stroke in 6 patients (associated with right hemisphere damage in 3 patients and with left hemisphere damage in 3 patients) and haemorrhagic stroke in 1 patient (right hemisphere). Early mortality was 5% with 2 (0.69%) patients dying in Group 1 and 14 (38.9%) in Group 2. Univariate analysis revealed that the preoperative risk factors of neurological complications were: age > 68 years (with a cutoff value of 58.5 years), a history of stroke with paresis, atrial fibrillation (AF) and a euroSCORE of > 6 (with a cutoff value of 4.5). The peri- and postoperative risk factors included: surgery type (complex coronary and valvular surgeries aortic valve surgeries), duration of CPB of > 142 min, duration of aortic crossclamping of > 88 min, mean perfusion pressure during CPB of < 70 mm Hg, haemodilution manifested by a haematocrit (HCT) of < 28%, perfusate supply, time to regaining consciousness of > 14.5 h and duration of artificial ventilation of > 30.5 h. Multivariate analysis revealed the following factors to increase the risk of neurological complications: long duration of ventilation, a history of stroke with paresis, AF, low HCT values and long duration of aortic cross-clamping. The Nagelkerke R2 coefficient of determination was 0.636, the sensitivity was 74.36%, the specificity was 97.545% and the accuracy was 94.74%.

Conclusions: In patients undergoing heart surgery, the independent risk factors of neurological complications in the first 30 days include: long duration of ventilation, a history of stroke with paresis, AF, haemodilution manifested by an HCT of < 28% and long duration of aortic cross-clamping. Neurological complications are associated with high postoperative mortality.