Table 3. Secondary Outcomes on Masked UWF-Color Images by Baseline Overall NPI.
| Outcome | No. of eyes | Kaplan-Meier estimate (95% CI) | Hazard ratio (95% CI) | P value |
|---|---|---|---|---|
| Development of PDR within the ETDRS fields on masked UWF-color images or receipt of DR treatmenta | ||||
| No nonperfusion | 47 | 7.8 (2.6-22) | 1.16 (1.04-1.29)b | .008 |
| Low nonperfusion | 153 | 22 (15-30) | ||
| Medium nonperfusion | 154 | 18 (12-26) | ||
| High nonperfusion | 154 | 32 (25-41) | ||
| Development of vitreous hemorrhage within the ETDRS fields on masked UWF-color images or at clinical examination or receipt of DR treatmenta | ||||
| No nonperfusion | 47 | 2.5 (0.4-16) | 1.30 (1.14-1.48)b | <.001 |
| Low nonperfusion | 153 | 7.2 (3.8-13) | ||
| Medium nonperfusion | 154 | 11 (6.2-18) | ||
| High nonperfusion | 154 | 21 (15-29) | ||
| DRSS improvement of ≥2 steps within the ETDRS fields on masked UWF-color imagesc | ||||
| No nonperfusion | 47 | 10 (4.0-25) | NA | NA |
| Low nonperfusion | 153 | 19 (13-28) | ||
| Medium nonperfusion | 154 | 18 (12-26) | ||
| High nonperfusion | 154 | 18 (12-26) |
Abbreviations: DME, diabetic macular edema; DR, diabetic retinopathy; DRSS, Diabetic Retinopathy Severity Scale; ETDRS, Early Treatment Diabetic Retinopathy Study; FA, fluorescein angiography; NA, not applicable; NPI, nonperfusion index; PDR, proliferative diabetic retinopathy; PPL, predominantly peripheral lesions; UWF, ultra-widefield; VEGF, vascular endothelial growth factor.
Eyes receiving treatment with anti-VEGF, steroid, or vitrectomy for conditions other than DR were censored after treatment initiation. The Cox proportional hazards model adjusted for baseline DRSS score within the ETDRS fields on UWF-color images, percentage of total imaged area on FA, baseline overall NPI, and the correlation between the 2 study eyes from the same participant.
Per 0.1-unit increase.
Eyes receiving treatment with anti-VEGF, steroid, or vitrectomy for any conditions (ie, DR, DME, or others) were censored after treatment initiation. Because there was a significant interaction between DRSS score and NPI (P < .001), and the outcome rates in the mild and moderate NPDR subgroups were low, Cox proportional hazards analysis was performed including only the moderately severe, severe, and very severe subgroups (eTable 10 in Supplement 1). The Kaplan-Meier estimates for the cumulative proportion of DRSS improvement of ≥2 steps included eyes that improved at any time during the study (before initiating any treatment); therefore, some eyes may have subsequently worsened. At 4 years, only 8% of eyes (22/292) experienced ≥2-step DRSS improvement from baseline without initiating any treatment.