This study aimed to assess patient's capacity to perform a patient-reported outcome (PRO) measure (i.e., a self-rating scale) and examine its relationship with clinical characteristics including cognition. Fifty patients with schizophrenia were asked to rate the Subjective Well-being under Neuroleptics scale – Short form (SWNS) twice; the second rating was started immediately after they completed the first to minimize the gap between ratings. At the same time, the Positive and Negative Symptoms Scale (PANSS) and Brief Neurocognitive Assessment (BNA) were administered. The correlations between the two ratings for the SWNS total and each item scores were high ( r _s =0.94 and r _s=0.60–0.84, respectively); however, for 16 (80%) of 20 items, 5 or more patients (i.e., ≥10%) demonstrated a>1 point score difference. There was no significant correlation between the SWNS total score difference and any clinical characteristics including age, education duration, illness duration, antipsychotic dose, psychopathology, and cognition. In contrast, the number of items with a>1 point score difference was significantly correlated with disorganized symptoms and overall severity ( r _s=0.29 for both), as well as working memory and global cognition ( r _s=−0.41 and r _s=−0.40, respectively). These findings suggest that PROs should be interpreted with caution in patients with schizophrenia with prominent disorganization and cognitive impairment.

Patients with schizophrenia completed a self-rating scale back to back.

Significant differences in the two self-ratings were observed in many patients.

The differences were correlated with disorganization and cognitive impairment.