The Quality of Life Research Journal, A-21, Abstract #1198.

Abstract presented as a poster at 14th annual ISOQOL conference in Toronto 10-13.10.2007

AIMS To develop the individualised RetDQoL, which includes 2 overview items (present QoL and retinopathy-specific QoL), 26 domain-specific items and an open-ended question about any other effects on QoL. The design was based on the widely used ADDQoL with specific item content determined by 44 in-depth interviews in the UK and Germany.

METHODS 207 German patients with diabetic retinopathy (DR) were included. Principal components analysis identified factor structure; Cronbach’s alpha assessed internal consistency. To obtain an average weighted impact (AWI) score, impact and importance ratings for each applicable domain-specific item were multiplied, summed and the sum divided by the N of applicable domains. Construct validity was examined by testing correlations between overview items and AWI and expected relationships of overview items and AWI with visual acuity (VA), stage of DR (ranging from non-proliferative to proliferative), additional impact of macular oedema and SF-12 subscales. Content validity was explored using the open-ended question.

RESULTS A forced one-factor solution without ‘working life’ (applicable to n=55) showed high reliability (alpha=0.96). Worse VA was associated with worse present QoL and retinopathy-specific QoL and more negative AWI (all p<0.001). Proliferative DR was associated with worse retinopathy-specific QoL and more negative AWI (both p<0.001) than non-proliferative. Macular oedema was associated with worse present QoL (p<0.01), worse retinopathy-specific QoL (p<0.05) and more negative AWI (p<0.01). SF-12 subscales correlated with overview items and AWI (r:0.22-0.54, all p<0.001). As expected, AWI correlated more strongly with retinopathy-specific QoL (r=0.71) than with present QoL (r=0.28, both p<0.001). Employed people had less negative AWI (p<0.05) regardless of VA. For 6 domains, >60% of patients reported no impact. Open-ended question responses (n=12) mainly stressed existing items; no additions were needed.

CONCLUSIONS The RetDQoL is a reliable and valid measure of the impact of DR on individual’s QoL. It may usefully be shortened if findings are confirmed cross-culturally.