Quality of life: status and change (QLsc) reliability, validity and sensitivity of a generic assessment approach tailored for diabetes.

ABSTRACT

The aim was to review the psychometric properties of a generic strategy for assessing status and change in quality of life (QLsc) partly tailored for patients with insulin-dependent diabetes mellitus. Quality of life was defined as perceived well-being and life satisfaction, globally as well as within key domains and functions. The strategy has been developed iteratively and, thereby, proven psychometrically satisfactory across groups and settings. Seventy-three consecutive outpatients were examined in 1988, after a period on multiple injection treatment by insulin pen; 66 were re-examined in 1990. In 1988 and 1990, perceived well-being and life domain status were self-rated. Life domain changes attributed to pen treatment were rated retrospectively in 1988 and subsequent general changes over 2 years rated similarly in 1990. Shortened parallel ratings were performed by significant others. Series of cross-sectional and longitudinal analyses yielded convincing reliability, validity and sensitivity evidence of various kinds for all ratings, irrespective of rater-self or significant other. As expected, a consistent correlational pattern emerged between status self-ratings, suggesting 'the better/worse off' in one sense, the 'better/worse off' in other senses assessed. Moreover, congruent change linkages, suggesting 'the greater change for the better/worse' according to the one type, the 'greater change for the better/worse' according to the other one, were also salient. Fourteen persons with less congruence in the disparate, but parallel change rating parameters appeared to be in a critical disease phase, applying reaction formation or denial like defence mechanisms. In conclusion, the entire cohesive strategy also functioned well in the current setting. It may probably be applied in search for other vulnerable sub-groups and in analyses of coping strategies. It offers rich analytic options especially in evaluational contexts. The study outcome may stimulate development of methodology within this field.