Carolien Zeetsen

119 CHAPTER Cognitive i mpairments in patients with GUD predict relapse 6 T2 Non–relapse Relapse p –value ( n = 29) ( n = 33) 4.80 (1.08) 4.69 (1.26) .750 5.27 (1.07) 5.00 (1.09) .319 4.72 (0.59) 4.61 (0.66) .463 1.72 (0.53) 1.89 (0.33) .167 3.83 (1.47) 3.24(1.58) .138 5.83 (0.47) 5.76 (0.50) .574 26.17 (2.24) 25.21 (2.91) .163 27.6% 42.4% .171 Given that only cognitive performance on T1 was related to treatment outcome, these scores were used in the backward logistic regression analyses to explore the predictive value of the MoCA for relapse. The logistic regression model was statistically significant, χ 2 (1) = 8.62, p < .003, with only MoCA–DS memory as a significant predictor in the final model. The model explained between 10.2% and 14.1% (Nagelkerke R 2 , depending on which MoCA–DS were included/excluded in the model) of the variance in relapse and correctly classified 68.8% of the cases. Each point scored on MoCA–DS memory at T1 increases the odds of non–relapse with 1.64.