Accuracy of Holladay 2 formula using IOLMaster parameters in the absence of lens thickness value.

Background: The accuracy of the Holladay 2 (H2) formula is well-documented. This formula requires seven variables to estimate effective lens position (ELP) for the IOL power calculation. The lens thickness (LT) value is one of the required variables. Interestingly, the IOLMaster, which is one of the most commonly used optical biometers, can provide all the required ocular variables except LT value. It has become a pertinent issue to evaluate the accuracy of theH2 formula when it is used without the LT value. The purpose of this study was to evaluate the results when using the H2 formula, without the LT value, and compare such results to those obtained using the Haigis formula and the Hoffer Q formula.

Methods: The Institutional review board (IRB) gave their approval for the conduct of this prospective comparative study. One hundred and sixty-three eyes of 143 cataract patients from the Ophthalmology Department, Siriraj Hospital, Thailand were recruited. All eyes were measured using the IOLMaster (Carl Zeiss Meditec, Jena, Germany) for keratometry (K), axial length (AL), anterior chamber depth (ACD), and horizontal white-to-white (WTW) corneal diameter. Then, the LT measurement was obtained by A-scan ultrasonography (Quantel Axis-II, Quantel Medical, USA). Every patient underwent uncomplicated phacoemulsification by a single surgeon (NC) with a single technique using a single IOL model. Post-operative refraction was obtained at 3 months. The mean absolute errors (MAEs), median absolute errors (MedAEs) and percentage of the eyes within ±0.25, ±0.50, and ±1.00 D of predicted refraction was calculated for H2 formula both with and without LT input, Haigis, and Hoffer Q formula. The results were also classified into a group of short AL (<22.0 mm), average AL (22.0 to 24.5 mm) and long AL (>24.5 mm).

Results: There was no statistically significant difference in either MAEs or MedAEs of all formulas in all AL groups including the H2 with and without LT. There was a trend toward lower MAEs and MedAEs for H2 in the long AL group. Percentage of the eyes within ±0.25, ±0.50, and ±1.00 D of predicted refraction were similar in all AL groups.

Conclusion: The preliminary results of this study showed that the H2 formula performed well even without the LT value. It was comparable to the Haigis and Hoffer Q formulas.