LAUSR

Abstract Background Cognitive remediation has been associated with enhanced cognition and psychosocial functioning in schizophrenia (SCZ). We present the preliminary results of a naturalistic study using a Cognitive Rehabilitation Programme (CRP) in Athens, Greece. The programme includes 40 individual hourly sessions, having a frequency of at least 2 sessions per week. It has a cognitive exercises and a social cognition module and aims at improving social functioning through cognitive enhancement. The CRP is implemented at a specialized Unit of the Greek National Health Service which is located in the centre of Athens for patients with SCZ living in the community. Methods We analyzed the CRP effects on verbal learning and memory, for the first 47 patients with SCZ referred to the Unit who took part in a naturalistic study. We compared these effects with those from a control group of 8 patients with SCZ undergoing occupational therapy. We conducted seven linear regression analyses investigating the effects of the CRP on verbal learning parameters using the Hopkins Verbal Learning Test (HVLT) (immediate total recall-ITR, immediate recall at the three learning trials (IRLT1, IRLT2, IRLT3), delayed recall (DR), retention (RT), recognition (RC) controlling for the duration of illness (DOI). In these analyses the relevant HVLT post-treatment scores were the dependent variables and CRP (receipt or not) and DOI were the independent variables controlling for the effects of baseline HVLT scores. Results 47 patients with SCZ completed the CRP (15 women and 32 men, mean age=42.34 years-standard deviation/SD=11.69, mean DOI=18.27-SD=10.88). 8 patients (5 women and 3 men, mean age=54.5 years, SD=8.14, mean DOI=26.25, SD=10.63) participated in occupational therapy sessions of the same duration. Baseline HVLT scores strongly correlated with post-treatment scores in all analyses. CRP was associated with increased post-treatment HVLT total immediate recall scores (B=3.35, 95% Confidence Interval-CI=0.66, 6, t=2.5, df=49, p=0.016). DOI was associated with decreased post-treatment HVLT total immediate recall scores (B=-0.11, 95%CI=-0.2, -0.02, t=-0.25, df=49, p=0.015). Further analysis of the subjects’ performance in the three IRLTs revealed that the effect of CRP approached statistical significance in the IRLT1 (B=1.21, 95% CI=-0.04, 2.46, t=1.95, df=47, p=0.058). DOI was negatively associated with IRLT1 scores (B=-0.05, 95%CI= -0.09, -0.004, t=-2.23, df=49, p=0.023) and IRLT3 scores (B=-0.05, 95%CI=-0.1, -0.007, t=-2.3, df=48, p=0.026). The association of CRP and DOI with IRLT2 scores was not significant. DOI was negatively correlated with DR scores (B=-0.11, 95%CI=-0.16, -0.06, t=-4.38, df=49, p<0.001). Similarly, DOI was negatively associated with RT scores (B=-1, 95%CI=-1.67, -0.35, df=49, p=0.004). We failed to find any effect of CRP on DR and RT scores. The ANOVA Model for RC scores was not significant. Discussion We presented preliminary results of an ongoing naturalistic study. CRP was associated with improved immediate recall, after controlling for the effects of the DOI. However, we failed to find any association of CRP with other verbal learning and memory measures. DOI negatively affected immediate and delayed verbal memory and learning. Although the naturalistic design of our study supports its external validity, it also limits the interpretation of our findings. Due to its preliminary character, our study was underpowered. A future blind randomized trial recruiting more subjects could shed more light onto the effect of CRP on cognition and functioning in schizophrenia.