Ebrahim S
Department of Public Health, Royal Free Hospital School of Medicine, London, England.
Soc Sci Med. 1995 Nov;41(10):1383-94. doi: 10.1016/0277-9536(95)00116-o.
Health-related quality of life (HRQL) measures have been developed from the utilitarian ethical perspective of public health medicine which may be contrasted with individual-centered indicators; these emphasise the unique experience of a patient. The impairment-disability-handicap framework provides a more complete description of disease consequences and health outcomes than a generic HRQL indicator. The assessment of reliability and validity of HRQL indicators is often carried out inappropriately: population repeatability is measured when an indicator will be used to examine changes in individuals; between observer variation may be large compared with variation between populations. Content and construct validity are usually measured but the more important predictive validity is neglected. Effect sizes of HRQL indicators are likely to be inflated by use of between subject estimates of variance but the more appropriate within subject variance is seldom reported. HRQL indicators are of very limited value for many clinical and public health tasks: monitoring health of individuals and populations; evaluating the effects of health and social policy; allocating resources; evaluating the effects of treatment. Alternative methods of assessment derived from the impairment-disability-handicap conceptual framework are preferable.
