How to improve refractive prediction

The long and the short of IOL power calculations

Dermot McGrath

Posted: Thursday, April 1, 2021

Thomas Olsen MD, PhD

Identifying the individual sources of error in biometry will enable clinicians to improve the refractive prediction in all types of eyes, and particularly those that fall outside the “normal” range of measurements, according to Thomas Olsen MD, PhD.
Speaking at the 25th ESCRS Winter Meeting, Dr Olsen, Aarhus University, Denmark, discussed some of the issues that occur when performing biometry in long and short eyes as well as those with flat and steep corneas.
“Be careful about those outliers that we might encounter. The newer multiple-variable statistical formulas tend to break down when dealing with a disproportionate anterior-posterior segment. It is important to identify those cases and go for ray-tracing formulas or other optical formulas that can deal with the unusual anterior or posterior segment,” he said.
Issues to watch for with long eyes include measurement of the true axial length, effective lens plane (ELP), keratometry readings and the IOL optic configuration, said Dr Olsen.
Measurements of the true axial length are often confounded by the fact that the IOLMaster (Carl Zeiss Meditec) was originally calibrated against immersion ultrasound by Wolfgang Haigis in 2001.
“In a way, it is immersion ultrasound values that we get out of the IOLMaster readings. The IOLMaster does not give the optical axial length directly but delivers in fact a re-transformed measurement, which can be a source of error,” he said.
Another anomaly is the difference seen between preoperative and postoperative axial length measurements with the IOLMaster, said Dr Olsen.
“I think the source of error here is the crystalline lens because we do not know the true refractive index of that lens as compared to the calibrated index given by the machine. Increasing the group refractive index of the lens in IOLMaster results in a slightly higher overall refractive index of the phakic eye and no difference between the preoperative and postoperative measurements,” he said.
Refractive accuracy may also be improved by using segmented axial length calculations for such unusual eyes, said Dr Olsen.
Keratometry is another potential source of error, as the assumed index of 1.3375 on the IOLMaster gives too high a reading of the corneal power, and may therefore resulting in a hyperopic error. The IOL optic configuration, which determines the principal planes of the lens, will also have significant influence on the refractive effect of the IOL, noted Dr Olsen.
Sources of error to watch for in short eyes include ELP prediction and IOL optic configuration, said Dr Olsen. One strategy to improve accuracy is to use a fellow eye outcome to fine-tune results in the second eye. For steep and flat corneas, tomography and ray tracing can be useful in addition to keratometry to determine the true cornea power of these eyes, added Dr Olsen.