We were pleased to read that our article stimulated comments and we are glad to provide a reply. First, we ought to point out how unfortunate the reader was in overlooking key methodological points of 3 of the 5 articles he cited: (1) we adopted as “target refraction” the refractive error predicted with each of the adopted formulas, as clearly stated (Methods: “the predicted target with the actual IOL power that was implanted in the studied eyes was noted for each of the tested formulas and used for the calculations.”) ; (2) in the study by Aristodemou and associates, the Holladay 2 formula was not tested and eyes with axial length below 20.0 mm were not included (the median axial length in our series was 19.94 mm), hence it cannot be used as a guidance in this specific group of eyes; (3) in the study by Hoffer and associates, only implants with a dioptric power up to 30.0 D (diopter) were investigated (the dioptric range was 35.0-39.0 D in our series), hence it is not relevant to our study.
We concluded that “The Hoffer Q formula led to good or fair refractive outcomes in less than two thirds of the cases. With Holladay 1 and 2 and Haigis formulas, outcomes would have not been significantly different.” In small eyes, any statement that might favor the Hoffer Q or any other of the latest-generation formulas over the others clearly remains unsupported, pending further evidence.
Are the formulas inadequate in small eyes? Perhaps such large variability in refractive outcomes might be related to the large lens power tolerance allowed for highly powerful implants rather than to the inaccuracy of the lens prediction formulas: it is a fact that a lens labeled as 35 D could have an effective dioptric power ranging anywhere from 34.01 D to 35.99 D. This undoubtedly creates problems for surgeons in trying to understand their refractive outcomes and to optimize the formula used, and for researchers who try to develop formulas that would work accurately in these eyes. We discussed this logical point in the original manuscript, and it was surprising to note that it had been questioned.
Besides the obvious limitations related to the small sample size, clearly highlighted in the original manuscript, the methodology adopted in our study was certainly more robust than previously available; and certainly our study provided a further insight, offering findings and comments that could be of help to colleagues and valuable feedback to lens manufacturers.
Lastly, we understood that the reader was not aware of the usefulness of negative controls. The SRK-II formula is obsolete and inaccurate and is commonly not in use for calculation in any eye, regardless of the axial length, because of poor power prediction. We expected it would have performed the worst by far among the other formulas tested, and this was eventually confirmed in the study.