Diagnostic consistency of major depression with psychosis across 10 years.

OBJECTIVE

The present study investigates how consistently DSM-IV major depressive disorder (MDD) with psychosis was diagnosed by research consensus across 10 years and the association of clinical characteristics with diagnostic consistency.

METHOD

The sample included 146 participants, part of a larger first-admission cohort (N = 628) presenting to a psychiatric inpatient facility with psychosis, who were diagnosed with psychotic depression at least once across 4 assessments spanning 10 years (after first admission and at 6-month, 24-month, and 10-year follow-ups). The primary outcome of this prospective epidemiologic study was retention of the best-estimate consensus diagnosis at each assessment. Diagnoses at each assessment were determined from semistructured interviews, medical records, and informant reports. The participants were recruited from 1989 to 1995.

RESULTS

Fifty-five of the 146 participants (37.7%) were diagnosed with psychotic depression at each available assessment; 13 (8.9%) switched from MDD to bipolar disorder, 24 (16.4%) switched from MDD to schizophrenia or schizoaffective disorder, and the remaining 54 (37.0%) had other patterns of diagnostic change. Only 47 of 80 participants (58.8%) diagnosed with MDD at baseline retained a mood disorder diagnosis 10 years later (36 [45.0%] had MDD and 11 [13.8%] had bipolar disorder), while 16 of 52 participants (30.8%) who ended the study with MDD were initially misdiagnosed. Compared to participants who were consistently diagnosed with MDD, those switching from MDD to bipolar disorder had better premorbid adjustment, more first-degree relatives with MDD, better functioning, and fewer negative symptoms at baseline, whereas those shifting to the schizophrenia spectrum had a more insidious onset, longer initial hospital stays, worse functioning, and more negative symptoms (all P values < .05).

CONCLUSIONS

The diagnosis of MDD with psychosis among inpatients showed poor long-term consistency. For clinicians, results indicate that the diagnosis of MDD with psychosis based on a single assessment should be considered provisional.