Assessing disutility associated with diabetic retinopathy, diabetic macular oedema and associated visual impairment using the Vision and Quality of Life Index

Background: Use of generic multi-attribute utility instruments (MAUI) to assess the impact of diabetic retinopathy (DR) on health-related quality of life (HRQoL) has produced inconsistent findings. Therefore, we assessed the impact of DR, diabetic macular oedema (DME) and associated visual impairment on vision-related QoL (VRQoL) using a vision-specific MAUI. Methods: In this cross-sectional study, 203 diabetic patients were recruited from specialised eye clinics in a Melbourne tertiary eye hospital. Severity of combined DR/DME was categorised as: no DR/no DME, mild non-proliferative DR (NPDR) and/or mild DME; moderate NPDR and/or moderate DME and vision-threatening DR (severe NPDR or proliferative DR (PDR) and/or severe DME) in the worse eye. Visual impairment was categorised as: none (up to 0.18logMAR); mild (from 0.18 to 0.3logMAR); moderate (from 0.3 to 0.48logMAR); severe (from 0.48 to 0.78logMAR); and profound (worse than 0.78logMAR). The Vision and Quality of Life Index (VisQoL) vision-specific MAUI was the main outcome measure. As the distribution of the utilities was skewed, independent associations with covariates were explored using multivariable quantile regression models (five groups: 15 ^th, 30 ^th, 45 ^th, 60 ^th and 75 ^th percentiles) ranging from poorest to highest VRQoL. Results: Participants' median age was 65years (range: 27 to 90years). Of the 203 participants, 50 (24.6 per cent) had no DR/DME, 24 (11.8 per cent) had mild NPDR/DME, 47 (23.2 per cent) had moderate NPDR/DME and 82 (40.4 per cent) had vision-threatening DR. After adjusting for relevant covariables, only profound visual impairment was independently associated with VisQoL utilities (β= -0.297 ± 0.098 p < 0.01). Severity of DR/DME was not significantly associated with any group of VisQoL utilities. Conclusion: The variation in VisQoL utilities was attributed to profound visual impairment but not mild, moderate or severe visual impairment or DR/DME severity. These findings support the use of vision-specific MAUI to capture the impact of profound visual impairment associated with DR and DME. A DR-specific MAUI might be required to assess the specific utility deficits associated with DR/DME across the spectrum of the condition.