Accuracy of IOL Power Calculation Using Various Formulas and Partial Coherence Interferometry in Eyes With Axial Length Less Than 22.0 mm

Monday, April 20, 2015: 8:06 AM
Room 5A (San Diego Convention Center)
Maithil M. Thakkar, MD
Adrian Braganza, MBBS, MS
Somshekar Nagappa, MD
Mikeen Shah, MD
Rohit Shetty, DNB, FRCS
Danish Siddiquei, BOpto

To assess the predictability of Intraocular Lens power calculations using IOL Master and four different IOL power calculation formulae (Haigis, Holladay 1, Hoffer Q and SRK/T) in eyes with axial length of < 22.00 mm.

This retrospective comparative study included 41 eyes of patients undergoing routine phacoemulsification with in-the-bag IOL implantation with axial length < 22 mm, from July 2012 to June 2014. Preoperative Axial Length and keratometric power were measured by IOL Master and power of the implanted IOL was determined using Haigis, Holladay 1, Hoffer Q and SRK/T formulae. Postoperative refractive errors 1 month after surgery were measured and were compared with the predicted postoperative power. Mean absolute error (MAE) was defined as the average of the absolute value of the difference between actual and predicted spherical equivalents of postoperative refractive error.

Mean Absolute Error for Haigis, Holladay 1, Hoffer Q and SRK/T formulae were 1.14, 0.42, 0.496 and 0.47 respectively. The proportion with an absolute error (AE) of less than 1 D was greatest in the SRK-T formula (97.56%), followed by those in the Holladay 1 (95.12%), Hoffer Q (90.24%) and Haigis formula (36.58%). Using unpaired Student's t-test, there was no statistically significant difference (p>0.05) with the use of SRK-T formula, whereas there was a statistically significant difference (p<0.05) for the remaining 3 formulae which means that SRK-T formula predicted post-operative outcomes better as compared to the other 3 formulae.

Mean Absolute Error was less than 0.5 in all formulae except Haigis formula. SRK-T formula was most predictive for post-operative refractive status in eyes with short axial length.