Seal P (Center for Health Quality Outcomes & Economic Research), Glickman M
(Center for Health Quality Outcomes & Economic Research), Spiro AS
(Boston VAMC), Rosen A
(Center for Health Quality Outcomes & Economic Research), Herz L
(Center for Health Quality Outcomes & Economic Research), Rosenheck R
(Northeast Program Evaluation Center)
The Veterans Health Administration (VHA) currently uses the Global Assessment of Functioning Scale (GAF) to monitor outcomes of mental health services. However, recent research and interest in patient-centered care suggest that including the patient’s perspective is important for assessing course and outcomes of treatment. The objective of this study was to determine whether self-report measures of mental outcome provide additional explanatory power relative to the GAF in predicting functional outcomes including psychiatric hospitalization and employment status.
The sample included 457 veterans receiving mental health services from one of two VA health centers between 2004 and 2006. Participants completed the Behavior and Symptom Identification Scale (BASIS-24) at enrollment in the study (T1) and three months later (T2). BASIS-24 assesses six symptom/functioning domains: depression/functioning, interpersonal relationships, self-harm, emotional lability, psychotic symptoms and substance abuse. The GAF is a unidimensional 100-point clinician-rated measure of psychiatric impairment. GAF ratings at time points corresponding to the self-report assessments, prior and post-enrollment mental health utilization and ICD-9-CM psychiatric diagnoses were obtained from administrative data files. Logistic regressions were done on two T2 functional outcome measures, psychiatric hospitalization and paid employment, to compare the predictive value of improvement in GAF ratings made by clinicians with improvement in BASIS-24 self-report scores.
Controlling for age, gender, race/ethnicity, education, level of care at T1 and T2 (inpatient vs. outpatient), and primary psychiatric diagnosis, results indicated that improvement in GAF ratings from T1 to T2 did not significantly predict post-enrollment hospitalization or employment status (odds ratios = 1.01 and 1.02, respectively, ), whereas improvement on the depression/functioning subscale of BASIS-24 significantly predicted both post-enrollment hospitalization and having a paid job at T2 (odds ratios =.59, and 1.67, respectively, p’s < .01). Greater improvement in depression/functioning decreased the odds of inpatient hospitalization and increased the odds of having a paid job.
Improvement in self-reported mental health status was a better predictor than clinician-rated improvement in GAF scores for predicting hospitalization and employment status in a sample of veterans receiving mental health treatment.
Efforts to further develop and implement quality and outcome measures for use in the VHA should consider inclusion of the patient’s perspective.