Purpose: To examine the refractive results and limitations of current intraocular lens power formulas when implanting two posterior chamber lenses in-the-bag to achieve emmetropia in extremely short eyes.
Methods: Preoperative measurements (corneal diameter, axial length, keratometry, anterior chamber depth, and lens thickness) and postoperative measurements (refraction, corneal vertex to iris depth, and iris to front anterior lens surface) were taken in six eyes from three patients, with axial lengths ranging from 15.09 to 19.95 mm. These data were used to calculate the prediction error for three current third-generation formulas (Holladay, Hoffer Q, SRK/T) and two older formulas (SRK2 and SRK1).
Results: None of the formulas accurately predicted the refractions using the optimized lens constants for normal eyes. The third-generation formulas were not different (P > or = 0.602) and averaged 5 diopters (D) of absolute error (Hoffer Q = 4.64 +/- 1.57 D; Holladay = 5.07 +/- 1.28 D; SRK/T = 5.12 +/- 1.43 D). The older formulas were significantly worse (P = 0.0006), with average mean absolute errors of 10.93 +/- 5.09 D for the SRK2 and 13.33 +/- 5.09 D for the SRK1. When the formulas were optimized for these six eyes, the mean absolute errors were Holladay = 1.33 +/- 1.25 D; SRK/T = 2.10 +/- 1.31 D; Hoffer Q = 4.54 +/- 2.00 D; SRK2 = 4.71 +/- 1.94 D; and SRK1 = 4.71 +/- 1.94 D. The Holladay and SRK/T formulas were statistically better (P = 0.0068) than the Hoffer Q and the two older formulas.
Conclusion: Current third-generation formulas are better than older formulas for extremely short eyes, but still are not acceptable for the desired clinical accuracy. Newer formulas that will use additional anterior segment measurements (corneal diameter, anterior chamber depth, and lens thickness) will be required for improved accuracy, because the anterior segment often is not proportional to the axial length.