Utility by BCVA in treated eye
Health state defined by BCVA category (letters; treated eye) | RESTORE* | Lloyd et al12 † | Brown et al13 § |
Mean utility (SE) | |||
1: 86–100 | 0.860 (0.034)‡ | 0.830 | 0.839 |
2: 76–85 | 0.860 (0.014) | 0.750 | 0.839 |
3: 66–75 | 0.813 (0.012) | 0.750 | 0.783 |
4: 56–65 | 0.802 (0.014) | 0.715 | 0.783 |
5: 46–55 | 0.770 (0.018) | 0.680 | 0.732 |
6: 36–45 | 0.760 (0.027) | 0.680 | 0.681 |
7: 26–35 | 0.681 (0.053) | 0.530 | 0.630 |
8: 0–25 | 0.547 (0.083) | 0.340 | 0.579 |
↵* Utility scores were calculated based on EQ-5D scores in RESTORE; EQ-5D scores were converted to utilities using social tariffs measured in a UK population.14 Mean utility for each BCVA state was calculated using a regression technique for repeated measurements at baseline, month 3, month 6 and month 12. Data from several measurement points were pooled to cover all possible health state transitions with a sufficient sample size. A possible trend effect in the pooled data was rejected (p<0.05).
↵† Patients underwent a Snellen visual acuity (VA) assessment and were categorised based on the better-seeing eye. Some adjustments were made to published values in order to convert VA ranges in Lloyd et al (obtained in a population of patients with diabetic retinopathy) to health states as defined in the current model.
↵§ Utilities were elicited from patients with diabetic retinopathy. Patients underwent a Snellen VA assessment and were categorised based on the better-seeing eye. Some adjustments were made to published values in order to convert VA ranges to health states as defined in the current model.
↵‡ Restricted to being greater than or equal to the utility in health state 2.
BCVA, best corrected visual acuity.