Lens power calculations should really be referred to as lens power estimations due to the fluctuations that we see in the postoperative refractive results. Even the most meticulous ophthalmologists will find that it is difficult to have more than 80% of their patients within 0.5 D of intended target after cataract surgery. While part of this variability is due to our preoperative biometry, a more significant factor is the difficulty in predicting the exact resting position of the IOL within the eye. Choosing the best formula for lens power estimation requires us to look at the characteristics of each individual eye.
First- through fourth-generation formulae
Decades ago, the initial lens power estimation method was to use the patient’s preoperative refraction, but this soon evolved into using more specific biometric parameters. While the refraction can change based upon the nuclear sclerotic progression of the cataract, the measurement of the corneal power and axial length do not. First-generation formulas, such as the Sanders-Retzlaff-Kraff (SRK), used the power of the cornea and the axial length, along with the A-constant of the specific IOL, to come up with power estimation for the IOL. This was based on a regression study of many eyes, and for patients with average eyes, it was reasonable. For eyes with shorter-than-average or longer-than-average axial lengths, this formula was not as accurate, so a modification was made to produce the SRK-II, a second-generation formula. This modification was to alter the A-constant based upon the axial length, and the results were better but still less than ideal.
Moving from regression-based formulae to theoretical formulae helped increase accuracy because these third-generation formulae now used biometric data to estimate the effective lens position within the eye. The commonly used third-generation formulae are the Holladay 1, the SRK-T (T for theoretical) and the Hoffer Q. Each of these formulae estimates the position of the IOL within the eye based on the keratometry and/or axial length, and the results are more accurate. For this reason, it is recommended that the older regression formulae should not be used in clinical practice. These third-generation formulae have proven to be very popular because they balance good results with simplicity because the only biometric data points required are keratometry and axial length.
The latest formulae, the fourth generation, use additional biometric parameters: the Haigis requires the anterior chamber depth, while the Holladay 2 needs that as well as the white-to-white, the lens thickness, the refraction and the age. There are other new formulae such as the Olsen, which uses a new C-constant that describes the IOL position as a constant fraction of the lens thickness with good results. Although the third- and fourth-generation formulae agree for most eyes, in unusual circumstances there can be greater accuracy with the newer formulae.
Which formula for which eye?
With many different formulae, how do you choose which one to use for a specific patient? Certainly, avoid using the older regression formulae (SRK-I and SRK-II, among others) and instead opt for a third- or fourth-generation calculation. Of the third-generation formulae, I tend to prefer the Holladay 1 in most eyes, reserving the Hoffer Q for short eyes with an axial length less than 22 mm and the SRK-T for long eyes with an axial length greater than 26 mm. The Holladay 2 is a great choice for nearly every eye because it is a full software package that allows many types of calculations in unusual eyes and even allows for honing of individual results by personalizing the A-constants.
To come to a decision, I look at the three factors that determine the lens power estimation: the axial length, the keratometry and the estimated effective lens position, which is often reflected in the anterior chamber depth.
The axial length measurement is critical in determining the ideal IOL power. A small 1 mm change in the axial length measurement can change the IOL power by 2.5 D in average eyes and even more in shorter eyes. With the advent of optical coherence biometry, we have accurate measurements of the axial length, so the previous issues of inadvertently measuring a staphyloma with manual A-scan are largely gone.
When I see an eye with a very long axial length greater than 26 mm, I know that most calculations will end up giving me an IOL power that is too low and likely to produce unintended hyperopia postop. A simple correction would be to aim for a more myopic postop goal, such as –0.5 D to –1 D using the Holladay 1 or SRK-T, in order to be closer to the desired plano result. A more accurate result can be achieved by using the Wang-Koch axial length (AL) modification in which the modified AL = 0.83*measured AL +4.27, and then the Holladay 1 is used with this newly modified AL.
In an eye with a very short axial length less than 22 mm, among the third-generation formulae, the Hoffer Q tends to produce good results. A more accurate choice may be using the Holladay 2 software because it incorporates the additional biometric data to calculate the IOL power.
We are looking for unusual cases in which the keratometry is very low (less than 40 D), very high (greater than 46 D), irregular or post-refractive surgery. Because the third-generation formulae use the keratometry value to determine the effective lens position in the eye, there is an assumption that a low keratometry value corresponds to a shallower anterior chamber and an IOL position that is more anterior and thus lower in power. Of the third-generation formulae, I find that the Holladay 1 tends to perform the best with most eyes whereas the SRK-T may be more thrown off by unusual keratometry readings. The fourth-generation formulae, such as the Haigis and the Holladay 2, do quite well in these eyes with flat or steep corneas because they incorporate the measured anterior chamber depth in the calculation.
For irregular corneas, the goal is to avoid leaving the patient hyperopic postop, which is what happens if too high of a keratometry value is entered. For these eyes, look at the topography and choose the lowest mean keratometry value in the central pupillary zone. This will tend to produce patients with a more myopic postop result, which is certainly preferable to a hyperopic outcome.
In a post-surgical cornea, such as one that has undergone LASIK or RK, the anterior chamber may be deep and the effective lens position more posterior even though the keratometry value is quite low. And because we typically measure the paracentral cornea and then extrapolate the value of the central cornea, post-surgical eyes can produce incorrect keratometry readings. In eyes with prior myopic corneal treatment, the central cornea is flatter than we are measuring, and in eyes with prior hyperopic corneal treatment, the central cornea is steeper than we are measuring. In eyes with these post-surgical corneas, we need to use the post-refractive surgery corrections before lens power calculations.
Effective lens position
Determining exactly where in the eye the IOL will end up is a challenge because the anterior chamber depth tends to increase with pseudophakia. Our manmade IOLs are also much thinner than the cataract that they are replacing, and when we factor in the shrink-wrap effect of capsular bag contraction, it becomes difficult to accurately predict the effective lens position. The presence of an unusually shallow anterior chamber would lead me to choose a lower IOL power because it will be sitting more anterior. Conversely, a very deep anterior chamber would mean that the IOL will sit more posterior, and thus a higher IOL power would be needed. The fourth-generation formulae incorporate the anterior chamber depth measurement into the calculation.
We have come a long way with our lens power estimation formulae, and the future may bring even more changes, such as accurate intraoperative aberrometry that could do away with preoperative calculations in many eyes. For now, the best advice is to use the third- and fourth-generation formulae for patients undergoing cataract surgery. The third-generation formulae have the benefit of simplicity and accuracy in most eyes, while the fourth-generation ones have more accuracy in unusual eyes but they require more input variables for calculation. In my clinic, my go-to third-generation formula is the Holladay 1, and my preferred fourth-generation is the Holladay 2. Finally, while plano can be a desirable outcome for many patients, there is still a benefit to being mildly myopic and no benefit to being hyperopic. Aiming for a pinch of postoperative myopia, or choosing the higher-power IOL when there is a discrepancy between formulae, can be beneficial.
- Uday Devgan, MD, is in private practice at Devgan Eye Surgery and Chief of Ophthalmology at Olive View UCLA Medical Center. He can be reached at 11600 Wilshire Blvd #200, Los Angeles, CA 90025; 800-337-1969; fax: 310-388-3028; email: email@example.com; website: www.DevganEye.com.
- Disclosure: Devgan has no relevant financial disclosures.