Incidence and prediction of permanent neurological deficits after cardiac surgery — are the existing models of prediction truly global?

OBJECTIVE: Permanent neurological deficit (PND) is a relatively rare but serious complication of cardiac surgery, associated with a high mortality and a poor prognosis for an acceptable quality of life. A few predictive models of PND have been developed; however, it is not certain whether they may be extrapolated to any cardiac surgical population. We aimed to assess the epidemiology and identify predictors of PND on the basis of a single, prospective hospital database from Eastern Europe. METHODS: We performed a retrospective review of 6016 consecutive adult patients (coronary revascularisation with or without cardiopulmonary bypass - 3,613 patients; isolated single-, double- or triple-valve repair or replacement - 1,221 patients; CABG+valve repair or replacement - 563 patients; aortic aneurysm surgery - 228 patients; and other procedures - 391 patients). PND was defined as a new focal or global disorder of cerebral function lasting longer than 24h and still present at the time of hospital discharge or the patient's death. Thirty independent preoperative, intra-operative and postoperative variables that might influence PND were selected and analysed. RESULTS: In total, PND was identified in 2.5% of patients (coronary surgery - 1.7%, isolated valve surgery - 2.9%, combined procedures - 5.3%, aortic aneurysm surgery - 7.5% and others - 2.2%). The overall mortality among patients with PND was very high in comparison to the remaining patients (40.4% vs 2.2%, p<0.001). In a multivariate analysis, PND was associated with five variables: cardiopulmonary bypass time >2h (odds ratio (OR) 3.35), emergency surgery (OR 3.34), early rethoracotomy (OR 3.17), age >65 years (OR 1.70) and unstable course of cardiac disease (OR 1.60). CONCLUSION: PND after cardiac operation is associated with a high mortality and poor prognosis. The incidence of PND varies depending on the procedure. Predictive models of neurological injury post-cardiac surgery should be more centre-specific.