Study objective: To clarify concepts and methodological problems in existing multidimensional health status measures for children.
Design: Thematic review of instruments found by computerised and manual searches, 1979-95.
Subjects: Nine health status instruments.
Main results: Many instruments did not satisfy criteria of being child centered or family focussed; few had sufficient psychometric properties for research or clinical use; underlying conceptual assumptions were rarely explicit.
Conclusions: Quality of life measures should be viewed cautiously. Interdisciplinary discussion is required, as well as discussion with children and parents, to establish constructs that are truly useful.