Reliability of global assessment of functioning ratings made by clinical psychiatric staff

In the Swedish psychiatric care system, systematic follow-up of clinical work with patients is becoming a part of regular service, and a number of care providers are using the Global Assessment of Functioning (GAF) to measure outcomes. This study investigated the reliability of the GAF and analyzed certain factors that affect measurement errors when the scale is used by regular psychiatric staff.

Eighty-one raters from various psychiatric outpatient clinics rated eight case vignettes. Interrater reliability was assessed by using intraclass correlation coefficients (ICCs), and factors associated with reliability were analyzed by using raters' unique residual values.

The results showed that staff who are responsible for assessing first-time patients at outpatient psychiatric clinics and making diagnoses are using the GAF with satisfactory reliability (ICC(1,1)=.81). The factors associated with reliability were raters' subjective attitude toward the GAF and motivation to use the scale and other measurement instruments in psychiatry.

GAF ratings made by an individual rater can be used to measure changes and outcomes at the group level. However, the measurement error is too large for assessment of change for an individual patient, in which case it might be necessary to use several raters. If raters are positively inclined to use rating instruments, measurement errors are minimized and reliability is maximized.