We were glad to see that our article stimulated comments, and we praise constructive criticism. Our study showed better results than previously reported in this specific subgroup of very rare patients, almost certainly owing to the application of rigid selection criteria; however, a main limitation was related to the small sample size, as clearly pointed out in our manuscript. Given the comments, it seems necessary that we reiterate the conclusions of our study: regardless of the adopted formula for the prediction of the lens power “our study undeniably demonstrated that refractive outcomes achieved by patients with small eyes are definitely worse than those achievable, and routinely achieved, by ‘‘normal” cataract populations,” and “statistical analyses […] failed to show a superiority of any of the latest-generation formulas over the others.” These conclusions hold true irrespective of whether the median or mean absolute errors are chosen to compare the formulas.
In our specific group of eyes implanted with the same intraocular lens, 11 of 28 eyes achieved a final postoperative refraction within 0.5 diopter (D) of target (with the adopted Hoffer-Q formula), whereas 12 of 28 eyes would have achieved the same refractive accuracy with the Holladay-2 and Haigis formulas. Besides the clinically negligible difference between the formulas, these results are poor (only circa 40% of eyes within 0.5 D of target). Clearly, it seems clinically irrelevant to be discussing the better accuracy of one formula over another if we talk of minimal differences in very large median absolute errors (for example, 0.76 D vs 0.80 D for the Hoffer-Q and Holladay-2, respectively), particularly given the limited sample size of the considered study.
We hope we were constructive when we suggested that a main problem could be the large lens power tolerance from nominal power that is allowed for the highly powerful implants often required in these eyes: hopefully, the manufacturers might help future analyses, and perhaps the development of specific algorithms to be used in small eyes, by reducing the power tolerance.
Given the above considerations, for the time being these rare patients need to be properly counseled before cataract surgery and must be told not to have high expectations in terms of refractive outcomes. In contrast, those patients should expect a safe surgery if they did not undergo previous or combined intraocular surgical procedures.
Lastly, we respectfully disagree that the large and well-conducted study from Aristodemou and associates might offer any guidance in these eyes: the Holladay-2 formula was not tested, and they did not include any eyes with axial length below 20.0 mm (in our study the median axial length was 19.94 mm).