Administration and Policy in Mental Health and Mental Health Services Research

, Volume 33, Issue 5, pp 568–577

Comparison of Self-report and Clinician-rated Measures of Psychiatric Symptoms and Functioning in Predicting 1-year Hospital Readmission


    • Department of EpidemiologyHarvard School of Public Health
    • Massachusetts Department of Public HealthApplied Statistics, Evaluation and Technical Services
  • Jane M. Murphy
    • Department of EpidemiologyHarvard School of Public Health
    • Department of PsychiatryHarvard Medical School
    • Department of PsychiatryMass General Hospital-East
  • Susan V. Eisen
    • Center for Health Quality, Outcomes & Economic ResearchEN Rogers Memorial Veterans Hospital
    • Health Services DepartmentBoston University School of Public Health
  • Sharon-Lise T. Normand
    • Department of BiostatisticsHarvard School of Public Health
    • Department of Health Care PolicyHarvard Medical School
Original paper

DOI: 10.1007/s10488-006-0066-y

Cite this article as:
Clements, K.M., Murphy, J.M., Eisen, S.V. et al. Adm Policy Ment Health (2006) 33: 568. doi:10.1007/s10488-006-0066-y


This study compared the self-report Behavior and Symptom Identification Scale (BASIS-32) and clinician-rated Global Assessment of Functioning (GAF) in their ability to predict a measure of psychiatric outcome, 1-year psychiatric hospital readmission. BASIS-32 and GAF were completed at admission for 1034 patients in an inpatient psychiatric facility. Multiple informants analysis was used to determine the difference between the two in predicting readmission within 1 year. We also examined sensitivity, specificity, and predictive value positive of the two measures, and whether self-report added information above clinician rating in predicting outcome. While the odds of 1-year readmission decreased with increasing BASIS-32 score, there was no association between change in GAF score and 1-year readmission. Although neither measure used alone demonstrated high predictive value, using both scores improved predictive ability over using clinician rating alone. In this setting, self-report was better than clinician rating in predicting psychiatric outcome. Differences between the two in relation to other outcomes need to be examined.


Psychiatric rehospitalization self-reportClinician ratingBASIS-32GAFMultiple informant analysis

Copyright information

© Springer Science+Business Media, Inc. 2006