The comparison of the intraocular lens power calculation formulas in long eyes

Objective: To investigate the accurate intraocular lens power formula in eyes with axial length over 26mm by using optical biometry. Methods: The retrospective case series was conducted at Medicana International Avcilar Hospital, Istanbul, Turkey from 2016 to 2018, and comprised data related to years between 2014 and 2018. Optical low-coherence reflectometry had been used in measurements. The calculation methods were included Hill-radial basis function, Sanders-Retzlaff-Kraff-Theoretical, Hoffer Q, Barrett Universal II, Olsen, Haigis and Holladay-2. The spherical equivalent values that were obtained with automated refractometer at the first postoperative month were recorded. Mean absolute error for each formula was estimated. SPSS 16 was used for data analysis. Results: There were 56 eyes of as many patients with a mean axial length of 26.70 +/- 0.88mm. The postoperative numerical error was within +/- 0.50 diopters in 46(82%) patients, and between the -0.50 and -1.00 diopters in 10(18%). The mean absolute error values were significantly lower in Haigis, Sanders-Retzlaff-Kraff-Theoretical, HofferQ and Holladay-2 formulas compared to the Barrett, Hill, and Olsen formulas (p<0.05). There was no statistically significant intra-group difference for Haigis, Sanders-Retzlaff-Kraff-Theoretical, Hoffer Q, and Holladay-2 formulas (p>0.05). Conclusion: In long eyes with an axial length over 26mm, Haigis, Sanders-Retzlaff-Kraff-Theoretical, Hoffer Q and Holladay-2 formulas yielded better postoperative refractive outcomes compared to Hill-radial basis function, Barrett Universal lland Olsen formulas. The Haigis formula had the lowest refractive numerical error.