Validating the measurement of real-world functional outcomes: phase I results of the VALERO study

Abstract

Objective: Treatment of cognitive impairment has been proposed as an intervention to reduce disability in people with schizophrenia. The Validation of Everyday Real-World Outcomes (VALERO) study was conducted to evaluate functional rating scales and to identify the rating scale or scales most robustly related to performance-based measures of cognition and everyday living skills.

Method: Adults with schizophrenia (N=198) were tested with the neurocognitive measures from the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Consensus Cognitive Battery, the UCSD Performance-Based Skills Assessment-Brief Version, and the advanced finances subscale of the Everyday Functioning Battery. They and an informant (a friend, relative, clinician, or case manager) also reported their everyday functioning on six rating scales. Best judgment ratings were generated by an interviewer who administered the rating scales to patients and informants.

Results: Statistical analyses developed an ability latent trait that reflected scores on the three performance-based (i.e., ability) measures, and canonical correlation analysis related interviewer ratings to the latent trait. The overall fit of the model with all six rating scales was good. Individual rating scales that did not improve the fit of the model were systematically deleted, and a final model with two rating scales was fitted to the data. A regression analysis found that the Specific Levels of Functioning Assessment was a superior predictor of the three performance-based ability measures.

Conclusions: Systematic assessments of real-world functioning were related to performance on neurocognitive and functional capacity measures. Of the six rating scales evaluated in this study, the Specific Levels of Functioning Assessment was the best. Use of a single rating scale provides an efficient assessment of real-world functioning that accounts for considerable variance in performance-based scores.

Figure 1. Model Fits During Sequential Deletion of Real-world Functioning Scales^a

^a RMSEA = root mean square error of approximation ^b Best fitting model: smallest chi-square value, largest p-value, and smallest RMSEA; the remaining scales were the Life Skills Profile and the Specific Levels of Functioning Assessment.