To evaluate the accuracy of predictive refraction and the factors influencing the predictability in combined vitrectomy and cataract surgery.

We retrospectively investigated patients who received combined vitrectomy and cataract surgery for idiopathic epiretinal membrane (ERM) and macular hole (MH), followed up for a minimum of 6 months. Preoperative refraction, target refraction, postoperative refraction, predictive refraction error (target refraction - postoperative refraction), accuracy of predictive refraction error (predictive refraction error was within ±0.5 diopter), intraocular pressure, axial length, central macular thickness, and tools that were used for intraocular lens power calculation (A-scan and IOL master) were assessed by analyzing medical records.

A total of 176 eyes (including 132 idiopathic ERM cases and 44 MH cases) were included in this study. The accuracy of predictive refraction error was 60.8% at 6 months and there was no difference between the idiopathic epiretinal membrane group (59.8%) and the macular hole group (63.6%). There was no significant difference in predictive refraction error according to axial length and tools (IOL master vs A-scan). Predictive refraction error correlated positively with preoperative refraction (r = 0.227; p = 0.002). In the ERM group, predictive refraction error correlated negatively with both preoperative central macular thickness and the change in central macular thickness between, before, and 6 months after surgery (r = -0.211; p = 0.015 and r = -0.241; p = 0.005).

The accuracy of predictive refraction error was approximately 60% in combined vitrectomy and cataract surgery. Postoperative refraction appeared to be myopia relative to target refraction with higher preoperative myopia and thicker preoperative central macular thickness. Hence, the intraocular lens power should be determined considering the above factors.