Figure 3. Development of Glaucoma-Related Adverse Events (AEs) According to Characteristics at Time of Lensectomy Among Children With Pseudophakic Eyes.

A Cox proportional hazards regression model was used to estimate the 5-year cumulative incidence of glaucoma-related AEs. The robust sandwich variance estimator was used to adjust for the correlation between eyes of participants who contributed data from both eyes to the analysis. The date of the glaucoma-related AE was not collected, so the date of the most recent office visit at which the complication was reported was used to estimate the cumulative incidence. Office visits were included in the analysis if reported up to 5.5 years following lensectomy to estimate the cumulative incidence (referred to as 5-year incidence). Hazard ratios (HRs) are adjusted estimates from a Cox proportional hazards regression model. The model was limited to 3 factors as categorical (laterality, anterior vitrectomy, and both). If the factor of interest was laterality or anterior vitrectomy, only 2 factors were included in the model. P ≤ .01 was considered significant. Observations with missing or unknown values in any of the subgroup factors were not included in the model.