Purpose: To determine the long-term reproducibility of axial length measurements and mean postoperative prediction errors after combined phacovitrectomy in patients with idiopathic epiretinal membranes.
Design: Retrospective cohort study.
Methods: The study included 43 patients who underwent combined phacovitrectomy and 30 patients who underwent only phacoemulsification. To determine the effect of vitrectomy, we compared patients treated with phacoemulsification only versus those treated with combined phacovitrectomy. Axial lengths were measured three times with a one-year interval, and the intraclass correlation coefficient (ICC), coefficient of variation (CV), and test-retest standard deviation (TRTSD) were assessed.
Results: There was no significant change in axial length, and axial length measurements showed high reproducibility in all groups. ICC, CV, and TRTSD values were 0.997, 0.24%, and 0.056, respectively, for the vitrectomized eyes. The mean postoperative prediction error was -0.37 diopters(D) in vitrectomized eyes (p < 0.001), while it was +0.11 D in patients with phacoemulsification (p = 0.531). The myopic shift was more obvious in eyes with a shallower anterior chamber (p = 0.008) and a thicker lens (p = 0.025).
Conclusions: Axial length measurements showed excellent long-term reproducibility at 2 years after combined phacovitrectomy. Myopic shifts were observed after combined phacovitrectomy, which was probably due to changes in the effective lens position after combined phacovitrectomy, rather than to changes in the axial length.
Keywords: axial length; combined phacovitrectomy; postoperative prediction errors; reproducibility.