Sir,

Dr Haigis' comment about the nomenclature is very justified. About ACD=AtoACD(ELPtoA(ELPconst)), this function converts ELP constant to ACD constant for the particular IOL. As can be seen in the definition of the HofferQ function in the appendix, ACD is first defined as a variable of the double type and is assigned the value of ACD constant (personalised ACD constant or manufacturer's ACD constant, whatever) in the first step; then this value is modified as we go down the next steps inside the function. Unfortunately, the whole subject of biometry is replete with overlapping and sometimes confusing terms, for example, ACD (is it ac depth or ELP?), corneal height (from corneal vertex or from secondary principal plane or posterior surface?), K value (optical or keratometric?), axial length (ultrasonic or optical?), etc.

About the computation of the predicted ACD according to Dr Hoffer's 1993 paper, in my article I did mention that I already experimented with all these data and it gave identical values of the predicted ACD (4.40887, 3.14482, and 6.06225 in examples 1, 2, and 3, respectively). I think our confusion is from the supposition of ELP constant being equal to ACD constant, which it is not.