Accuracy and reproducibility of biometry using partial coherence interferometry

Purpose: To determine whether partial coherence interferometry is comparable to standard contact ultrasonic biometry in accuracy, reproducibility, and ease of use. Setting: Academic university practice. Methods: In this prospective study, simultaneous biometry was performed in 91 consecutive patients (111 eyes) using partial coherence interferometry (IOLMaster, Zeiss Humphrey Systems) and ultrasonic biometry (I3, Innovative Imaging, Inc.). The differences between the predicted spherical equivalent in diopters (D) and the actual results with both techniques were compared. Results: The IOLMaster was significantly better in the mean absolute error (0.533 D ± 0.589 [SD] versus 0.757 ± 0.723 D; P = .012) and the percentage of eyes within ±0.5 D (61.2% versus 42.3%; P = .003) and ±1.0 D (87.4% versus 77.5%; P = .05) of the predicted refraction. Conclusion: The IOLMaster was more accurate and reproducible than the contact ultrasonographic technique.