METHODS: Thirty-eight patients(40 eyes)with cataract who had undergone vitrectomy were divided into two groups: vitrectomy alone as group A(22 eyes), vitrectomy combined with inert gas(C₃F₈)tamponade as group B(18 eyes). In addition, normal eyes(20 eyes)without vitrectomy history who required cataract surgery were included as a control group C. Axial length(AL), keratometric value(K), anterior chamber depth(ACD), and white-to-white(W-W)were measured with IOL Master 500 before the surgery. The power of IOL was calculated with the online Barrett Universal Ⅱ formula. Phacoemulsification combined with IOL implantation by an iris hook assistance technique was used to treat groups A and B, group C was treated only with phacoemulsification and IOL implantation. There were no complications, such as posterior capsule rupture. The postoperative refraction was obtained three months after surgery. The prediction error(PE), absolute prediction error(AE), and median absolute error(MedAE)among the three groups were compared.

RESULTS: There was no significant difference among the three groups in AL, K value, and IOL power. A significantly deep ACD was noted in groups A and B compared with group C(P<0.01). There was a significant difference in the prediction error among the three groups(P=0.042). Group B showed a significant hyperopic shift compared with group C. AE and MedAE among groups showed no significant difference.

CONCLUSION: There was no significant difference in the refractive results of patients with phacoemulsification sequential to vitrectomy alone compared with the normal control group. However, a hyperopic shift was found in patients with phacoemulsification sequential to vitrectomy with inert gas tamponade comparing with the normal eyes. Undercorrection should be avoided when selecting IOL power.