Multidimensional assessment of psychosis: a factor-analytic validation study of the Routine Assessment of Patient...

The areas of function affected by major mental disorders are more diverse than the list of core symptoms assessed by many psychiatric rating scales, and the cross-sectional picture obtained in mental status interviews often fails to capture important data. Information on patient function can be obtained from measures that are based on extended observation and complement symptom-focused assessments. The Routine Assessment of Patient Progress (RAPP) is a 21-item rating scale that assesses both functional and psychiatric symptoms. It is usually completed by nursing staff who have observed patients over a 1-week period. Previous research has shown it to be reliable, valid, simple to complete, and of substantial value for patient care and diagnosis. The present study sought to examine the psychometric structure of the RAPP to define what domains of symptoms and behavior it measures. RAPP scores obtained from 165 psychotic inpatients were submitted to a factor analysis. A five-factor solution was derived in which 18 of 21 RAPP items were assigned to factors. The factors were labeled aggression, positive symptoms, negative symptoms, somatization/anxiety, and organic/ disorganization. The RAPP factors were moderately correlated with conceptually similar factor scores derived from the Positive and Negative Syndrome Scale (PANSS). RAPP aggression scores were validated with an independent clinical measure of aggression. Patients who were independently rated as improved over their hospital stay showed significant improvement on all RAPP factors, and unimproved patients showed stability or deterioration on RAPP measures. The data indicate that RAPP factors assess domains of psychopathology that are moderately correlated with both global ratings and symptom-focused scales. The RAPP's sensitivity to change suggests it is a valid measure of treatment outcome that could be used in controlled trials, as well as standard care outcome evaluation.