The ability to produce high quality instruments for the assessment of quality of life has advanced considerably in recent years. As the science progresses it has become clear that certain standards must be met if outcome measures are to be capable of providing useful, reliable, and valid information within the context of clinical studies and trials. This paper specifies what these standards are with particular reference to theoretical basis, practicality, acceptability to respondents, unidimensionality, scaling and psychometric properties, and cultural validity and equivalence. The paper also indicates how failure to achieve such standards results in measures that are inaccurate and insensitive to true changes in outcome.
