Are indirect utility measures reliable and responsive in rheumatoid arthritis patients?