< Back to previous page

Publication

A comparison between self-report and interviewer-rated retrospective reports of childhood abuse among individuals with first-episode psychosis and population-based controls

Journal Contribution - Journal Article

The typical reliance on self-report questionnaires in retrospective case-control studies of childhood abuse and psychotic disorders has been criticised, due to the potential for recall bias associated with, amongst other factors, cognitive impairments and detachment from reality, among individuals with psychosis. One way to establish if any substantial bias may exist is to examine whether the concordance of reports of childhood abuse established from retrospective self-report methods versus more comprehensive interviewer-rated assessments differ between individuals with psychosis and controls. Data from the Childhood Adversity and Psychosis (CAPsy) study were used to examine the accuracy, strength of agreement, and convergent validity of two distinct retrospective measures of childhood abuse: a self-report questionnaire (the Childhood Trauma Questionnaire; CTQ) and a comprehensive interview (the Childhood Experiences of Care and Abuse schedule; CECA). In a sample of 234 cases with first-episode psychosis and 293 controls, we found no strong evidence that the validity of the two measures differed between cases and controls. For reports of sexual and emotional abuse, we found fair levels of agreement between CECA and CTQ ratings in both groups (kappa coefficients 0.43-0.53), moderate to high sensitivity and specificity, and reasonably high convergent validity (tetrachoric correlations of 0.78-0.80). For physical abuse, convergent validity was slightly lower in cases compared with controls. Both measures can be used in future studies to retrospectively assess associations between childhood abuse and psychotic phenomena, but time-permitting, the CECA is preferable as it provides additional important contextual details of abuse exposure.
Journal: Journal of Psychiatric Research
ISSN: 0022-3956
Volume: 123
Pages: 145 - 150
Number of pages: 6
Publication year:2020