Global Assessment of Functioning scores are not sensitive to variations in social cognition among patients with schizophrenia, a major part of "real world outcomes" for this population, a study has shown.
Indeed, "although the GAF scale is reported to be a valid measure of function and, as such, a useful tool for clinicians to assess patients’ progress in rehabilitation, it is a less sensitive tool for establishing and monitoring improvements in function associated with social aspects of cognition," wrote Deirdre A. Robertson in an article published online in Schizophrenia Research (2013 [http://dx.doi.org/10.1016/j.schres.2013.01.016]).
According to Ms. Robertson, a PhD candidate at Trinity College Dublin, and her colleagues, although the GAF is widely used, "it is unclear if the GAF scale is a reliable tool for assessing the functional difficulties associated with more subtle impairments in social cognition" in schizophrenia, once investigators control for the outward presentation of positive and negative symptoms.
To test this hypothesis, Ms. Robertson enrolled 216 patients from the Resource for Psychosis Genomics, Ireland, initiative with schizophrenia or schizoaffective diagnoses recruited from sites across Ireland.
Social cognition was measured using two "theory of mind" tasks. The first was the reading the mind in the eyes test, in which subjects rate 36 photographs of eyes portraying different expressions and correctly identify the thought or feeling as one of four descriptors printed around each photograph.
The second was the hinting task, in which participants listen to 10 vignettes describing social interactions and infer the intent behind the hint after each vignette.
Patients also were scored according to a neurocognition composite score sensitive to deficits typically seen in schizophrenia – intelligence quotient, episodic memory, working memory, and attentional control.
Finally, the researchers used linear regression analyses to assess the degree of variance in global functioning explained by symptom severity, medication, neurocognition as represented on the composite score, and social cognition.
First, the authors found that scores on the GAF scale were significantly positively correlated with the hinting task, but not with the eyes task, so that the latter was excluded from further analysis.
Then, the authors performed a regression analysis with symptom severity entered on the first step, followed by neurocognitive and hinting task performance on the second step and GAF scores entered as the dependent variable.
"After the 25.2% of variation in GAF scores explained by symptom severity, neurocognitive performance explained an additional 4.7% of variation, while the hinting task failed to explain any further variation," they wrote.
According to the researchers, one limitation of their study was the use of only one social cognitive domain, theory of mind. Indeed, "it is possible that the GAF scale may have been more sensitive to social cognition deficits in emotion perception/recognition or social knowledge," they wrote, which would have been better perceived on some other domain.
However, the investigators chose to focus solely on theory of mind for two reasons: A 2011 meta-analysis [(Neurosci. Biobehav. Rev. 2011;35:573-88)] reported that theory of mind was the social cognitive domain that explained the largest amount of variance in functioning, and "a previous study from our group demonstrated that theory of mind mediated the relationship between basic neuropsychological function and functional outcome using another measure of functional outcome, the Independent Living Scale."
They concluded that, given their findings, "it is likely that as our understanding of social cognition increases and as remediation program[s] become more frequently incorporated into treatment plans for schizophrenia, a new scale will be needed for clinical use which can more sensitively monitor progress in these domains than the largely symptomatically based GAF scale."
Individual researchers were supported by the Wellcome Trust, Science Foundation Ireland, and Ireland’s Health Research Board. The authors stated they had no conflicts of interest or other funding.