RESUMO
BACKGROUND: To investigate the accuracy of preoperative biometry in eyes undergoing combined phacovitrectomy and to compare it with eyes having cataract surgery at a later point in time following vitrectomy. METHODS: Patients with epiretinal membrane or macular hole who underwent combined phacovitrectomy (group 1) or phacoemulsification following pars plana vitrectomy (PPV) (group 2) were included in this retrospective, comparative, interventional case series. The primary outcome measures were the intraocular lens power prediction error (PE) and the percentage of eyes with PE > ± 0.5D in the two groups. Secondary outcome measures included the correlation between epidemiological, clinical, or surgical factors and dioptric shift. In addition, the influence of optical coherence tomography characteristics to the PE was investigated. RESULTS: Group 1 and 2 consisted of 55 and 54 eyes, respectively, for a total of 109 eyes included in the study. The mean absolute PE was 0.59 D (range + 1.4 to - 2.5D) in group 1 and 0.35 (range + 1.0 to - 1.45D) in group 2 (p = 0.01). PE greater than 0.5D was observed in 47% of eyes in group 1 as opposed to 16.6% of eyes in group 2 (p = 0.027). The PE was associated with shallower anterior chamber depth (ACD), increased central macular thickness (> 300 µ), and worse baseline best-corrected visual acuity. Photoreceptor ellipsoid zone or external limiting membrane disruption was not associated with significantly greater postoperative refractive deviations. CONCLUSION: Combined phacovitrectomy may result in greater postoperative refractive prediction error compared to phacoemulsification alone following vitrectomy. Patients with worse vision, greater central macular thickness, and shallow anterior chambers require more caution since they are prone to inaccurate preoperative biometry.