LAUSR

absolute decline from baseline scores greater than the defined proportion of the SD or a percentage change from the baseline. On the other hand, the reliable change index is an alternative and increasingly popular statistical method, which makes POCD investigations even more confusing. Furthermore, a substantial number of experts advocate including a nonsurgical or noncardiac surgical control group, but the inclusion of control groups also has some issues in light of the complicated physiological processes attributed to cardiac surgery and cardiopulmonary bypass. Following the suggestion of our colleagues, we conducted a new statistical analysis of our data using a difference of 1 SD from the baseline scores as a definition of POCD. In this newly performed analysis, 44 of 80 (55%) patients in the dexamethasone group and 54 of 81 (66.7%) patients in the placebo group fulfilled the diagnostic criteria for POCD [relative risk (RR), 0.83; 95% confidence interval (CI), 0.64 to 1.06; P1⁄4 0.13]. In the previous analysis, nine of 80 (11.3%) patients in the dexamethasone group and 21 of 81 (25.9%) patients in the placebo group fulfilled the diagnostic criteria for POCD (RR, 0.43; 95% CI, 0.21 to 0.89; P1⁄4 0.02). Therefore, we confirmed previous findings that a more liberal POCD definition results in a significantly higher reported incidence, but this also may produce skewed results, particularly in the early postoperative period when the patients are under the effect of sedatives, postsurgical pain, sleep deprivation and the hospital environment. Thus, applying a strict POCD definition reflects the higher cognitive deterioration required for fulfilling the diagnostic criteria, and truly represents a #safety& switch for early postoperative testing.