B-mode-guided vector-A-mode versus A-mode biometry to determine axial length and intraocular lens power

Purpose: To compare prospectively the reproducibility and accuracy of B-mode-guided biometry with those of A-scan biometry using a conventional A-mode probe to calculate intraocular lens (IOL) power. Setting: Department of Ophthalmology, Hotel-Dieu de Paris, France. Methods: The axial length (AL) in 87 eyes of 72 candidates for cataract surgery was determined by B-mode-guided vector-A-mode and A-mode biometry using an Ophthascan S Ultrasound imager. Patients were assigned to one of two groups based on the B-mode biometry: nonmyopic (AL < 24.5 mm; n = 54) or myopic (AL > 24.5 mm; n = 33). Postoperative refractive results were compared with attempted values. Results: Mean AL variance was significantly greater when using the A-mode than the B-mode: 0.157 mm +/- 0.260 (SD) versus 0.015 +/- 0.018 mm in the myopic group (P<.0001) and 0.024 +/- 0.045 versus 0.009 +/- 0.011 mm in the nonmyopic group (P<.0001). More eyes having B-mode biometry achieved a final refraction within +/-0.50 diopter (D) of the attempted refraction (63 and 43%, respectively; P<.05). No deviation greater than 1.60 D was observed with the B-mode in the myopic or nonmyopic group. Three cases with a such a deviation Cup to 2.24 D) would have been observed had A-mode-based biometry been chosen for the IOL power calculation. In the myopic group, attempted postoperative refraction was within +/-0.50 D in 78% of eyes having B-mode biometry compared with 65% having A-mode. This difference was not statistically significant. Conclusion: These results suggest that the reproducibility and accuracy of AL measurements are significantly better with B-mode-guided A-mode biometry than with A-mode biometry in myopic and nonmyopic eyes.