Background To evaluate the effects of pretreatment for dry eye disease (DED) on the accuracy of intraocular lens (IOL) power calculation. Methods Patients who underwent uneventful cataract surgery were included in the study. IOL power was determined using the SRK/T and Barrett Universal II (Barrett) formulas. The patients were divided into non-pretreatment and pretreatment groups, and those in the pretreatment group were treated with topical 0.5% loteprednol etabonate and 0.05% cyclosporin A for 2 weeks prior to cataract surgery. Ocular biometry was performed in all groups within 2 days before surgery. The mean prediction error, mean absolute error (MAE), and proportions of refractive surprise were compared between the non-pretreatment and pretreatment groups at 1 month postoperatively. Refractive surprise was defined as MAE ≥ 0.75D. Results In a total of 105 patients, 52 (52 eyes) were in the non-pretreatment group and 53 (53 eyes) in the pretreatment group. The MAE was 0.42 ± 0.33, 0.38 ± 0.34 (SRK/T, Barrett) and 0.23 ± 0.19, 0.24 ± 0.19 in the non-pretreatment and pretreatment groups, respectively (p < 0.001/=0.008). The number of refractive surprises was also significantly lower in the pretreatment group. [non-pretreatment/pretreatment: 9/2 (SRK/T); 8/1 (Barrett); p = 0.024/0.016]. Pretreatment of DED was related to a reduction in postoperative refractive surprise. [SRK/T/Barrett: OR = 0.18/0.17 (95% CI: 0.05–0.71/0.05–0.60), p = 0.014/0.006]. Conclusions The accuracy of IOL power prediction can be increased by actively treating DED prior to cataract surgery.

No abstract