Agreement between self-reported and registered psychiatric conditions in recently bereaved adults: The Aarhus...

There is concern for overestimating psychiatric conditions after bereavement. Our aim is to examine agreement between self-reported psychiatric symptoms over clinical cutoff scores and register-based psychiatric conditions among recently bereaved adults. We used multimodal data from The Aarhus Bereavement Study Cohort, combining longitudinal surveys (n = 1,227) with Danish register data covering the bereaved cohort (N = 7,017). Agreement for loss characteristics and psychiatric conditions was assessed by prevalence ratios, percentage agreement, and multinomial logistic regression to identify predictors of disagreement. Loss characteristics showed mid-range agreement between data sources. Psychiatric conditions were more frequently identified in self-reports than in register data. Agreement improved when psychiatric symptoms were self-reported multiple times, with the strongest alignment observed for a heterogenous grouping of complicated grief reactions. Integrating multiple data sources can provide more comprehensive interpretation, however self-reported symptom endorsement and registered diagnoses do not represent perfectly interchangeable indicators of psychiatric morbidity.