Accurate axial length measurements and calculations are essential in cataract surgery. Paying attention to the details and eliminating common errors are necessary for excellent intraocular lens calculation. This article discusses current and future biometry methods, offers tips to achieve consistent measurements, and suggests steps to address challenging ocular situations.

The Future of Biometry
Biometry is headed in the direction of the Zeiss Humphrey IOLMaster and more sophisticated laser interferometry equipment (See the feature on the IOLMaster this month). This change is being spurred on by the rapid evolution of refractive lens surgery and what it can provide our patients. Immersion biometry is also a technique that will be a part of our immediate future.

In addition, we have  an eye on Ultralink’s 3-D Artemis Ultrasound, a new device presented at the winter European Society of Cataract and Refractive Surgeons meeting in Barcelona. Dan Reinstein, MD, assistant clinical professor of ophthalmology at the Weill Medical College of Cornell University in New York, led the development of the device. It’s a 3-D, very high-frequency ultrasound system that provides thickness measurements of individual corneal layers with 1-µm precision. It’s an immersion B-scan that implements a very high frequency, so it’s much more accurate. You don’t have to lay the patient back and put him in a water bath. It’s actually a little cup. The patient sits up and places his face on the cup, creating a fluid chamber that the operator can lean over and view.

The digital high-frequency ultrasound improves the resolution and allows A- and B-scans to be combined. Not only do you obtain the axial length measurements that you want, you may be able to predict the position of the lens after surgery better. That’s the imprecise variable we have today: The one distance that you can’t measure preoperatively is the length from the corneal vertex to the IOL. This device allows you to map the entire anterior segment and see where the sulcus and the bag are. This will move us another step forward toward reducing the variability of where the lens finally rests.

Biometry Staples
No matter the device, you need to be certain that the data are reliable. Here’s a checklist to assure that.
 • Personalize your A-constants. Personalization is a vital step in improving accuracy in IOL calculations. The best way to achieve it is to assemble data on at least 25-30 of the same lens, implanted by the same surgeon  with the same method of biometry and the same technician performing these measurements.
Enter those 25-30 cases into a calculation program that will average the constants. The program will tell you what you actually achieved in terms of the final refraction, and you will know the lens power and the measurements of the eye. From that, you will see the constant that would have been right for that patient if everything had been perfect.

Keep in mind, though, that the lens, the surgeon and the patient all affect the constant. Things drift over time. For example, your A-scan changes, your surgical technique changes and you use different postop drops.
Continue to record and monitor postop data regularly, even after you’ve established a personalized constant.

The current  IOLMaster and Holladay IOL Consultant software with the Holladay II Formula allow for the personalization of lens constants on an ongoing basis. If your clinic has done thousands of cat-aracts a year, it’s difficult to change a constant until you’ve done another thousand. The Holladay IOL Consultant implements a “trailing window” that looks backward and weighs the most recent cases first, before continuing further back in the past. As soon as it gathers enough statistical significance—200-300 patients—it stops and gives you the most recent constant that’s appropriate. What you were using before that may have been dragging the constant down.

If you don’t have any experience with a new IOL, you have to refer back to the manufacturer’s constant. The “lens constant” listed on the IOL label is typically inaccurate, though. If you’ve learned that you are usually on the low side of the manufacturer’s constant, you can use that information to make a better guess the first time out. That is risky, however, because you don’t know if the manufacturer’s constant is going to be like yours. After those first 25-30 cases with the new lens, plan to go back and personalize that constant. 

 • Avoid simple errors. Common errors in biometry prior to cataract surgery are typically found in axial length measurements. Simple transfer data errors are also possible and can lead to refractive surprises. For this reason, have three members of your office team examine each chart prior to cataract surgery, looking specifically for any inadvertent errors or any need to repeat biometry preoperatively. If it’s the technician’s fault, such as for not aiming the probe correctly, it’s a good time to retrain all involved with the biometry. 
 
• Measure both eyes. Always do biometry on both eyes, especially if you already have an implant in one eye. Symmetry in the body is one check that every physician has used for years to see whether something is normal or not. The variation from one person to the next is much greater than from one side to the other. If you achieve the same measurement in both eyes and you are very close, then you are assured you are on target. But, if you get a measurement between the two eyes more than about 0.3 mm different, something is wrong.

Biometry Methods
There are pluses, minuses and limitations to methods of evaluating axial length, and there are reasons for learning and implementing all three in your practice. 
 
• Immersion A-scan. This continues to be an accurate and reliable method of evaluating axial length measurements with ultrasonic equipment. You simply place a Hansen and/or Prager shell between the patient’s lids, fill it with saline or methylcellulose and immerse the probe into the fluid, avoiding contact with the cornea. The disadvantage is that it requires more effort by the technician

Immersion A-scan is more consistent than applanation because there is no corneal indentation. It’s an accurate and reliable method of evaluating axial length measurements with ultrasonic equipment.

and the patient. But, the fact that there is no corneal contact reduces the chance of corneal indentation, leading to more accurate axial length measurements. An excellent resource for information concerning the technique of immersion A-scan is available through the video produced by Tarpon Springs, Fla., surgeon James Gills (Call (727) 938-2020).
 
• Contact biometry. In the theoretical world, there is no question that immersion A-scan should be superior to contact biometry, simply because it is easier for the machine to recognize where the vertex of the cornea is. With applanation, it’s often difficult to recognize that first “bang,” because it’s right at the vertex of the cornea. An imprecise measurement comes about, because it’s difficult to measure that first surface that you’re actually on.

An experienced technician, using a spring-loaded applanator, usually can obtain accurate measurements, though. If the technician uses too much pressure and causes significant corneal compression, the probe will spring back toward him. On the average, axial length measurements are about 200 µm shorter than those done with immersion. That’s the result of not recognizing that first bang, however, and not necessarily due to compression of the cornea.
Immersion is definitely better if you’re working with inexperienced technicians and you don’t have a spring-loaded probe handy. Immersion makes it easier for a technician to avoid a mistake. Some studies have found no difference between immersion and contact, while other studies point out that there is one. The difference usually is the technician.

The accuracy of the immersion and applanation ultrasounds can reach a tenth of a millimeter in good hands. 

 • Optical coherence. There are many obstacles to accurate axial length measurement, including patient cooperation, technician expertise and the type of equipment being used. The IOLMaster helps to level the playing field here, assuming that the nucleus is not too dense for the light pathway to reach the retina. This is also a problem with dense posterior subcapsular cataracts.

A technician enters data for an IOL calculation. Have three members of your office team examine each chart prior to cataract surgery, looking specifically for any inadvertent errors or any need to repeat biometry preoperatively.

We’ve had the IOLMaster for two years and have been impressed with its ability to provide accurate scans with no contact to the eye and still be very patient- and technician-friendly. Its wavelength of light, much smaller than the wavelength of ultrasound, allows you to achieve tolerances eight to 10 times better (0.04 mm).

In order to propagate a good wavefront into the eye and reflect it back, however, you must have fairly clear media. You can’t usually obtain a good reading on a patient with a severe cataract. If you have a posterior subcapsular cataract, where the light can’t penetrate the center very well, it gives you very unusual measurements, because it is measuring through the periphery of the lens.

The IOLMaster doesn’t measure the thickness of the lens, nor does it measure anterior chamber depth using the coherent light interference. Instead, it uses the Scheimpflug principle, a technique that doesn’t correlate very well with the measurements obtained with ultrasound. As a result, if you’re using the Holladay II, you have to use the ACD and lens thickness measurements from the A-scan to get measurements consistent with previous data.
For refractive lensectomies, mild cataracts and non-dense posterior subcapsular cataracts, you can achieve very good measurements.

Challenging Situations
No matter the method, be prepared for curve balls with IOL calculations.

• Piggybacking IOLs. Use the software of the Holladay IOL Consultant and the Holladay II Formula to calculate the intended piggyback powers. Place the higher powered lens posteriorly. If you’re performing Array (Allergan) implantation, consider placing the Array anteriorly in the ciliary sulcus and the stronger monofocal lens in the bag. This arrangement assures the multifocal Array will see the normal incident rays emerging from the cornea.

With the Array, aim for a little bit on the hyperopic side to produce the best outcomes. You end up shooting  for about +0.12 D or +0.25 D, or even +0.3 D, but you always want to err on the hyperopic side. If you err on the myopic side, up to -0.25 D or -0.50 D, the near vision may be too close. The patient ends up with pretty good distance and near vision, but he may have a defect in the middle.

• Excessive axial length. This problem occurs primarily in high myopes. Because of an off-axis posterior staphyloma, the macula or fovea is often not the most posterior fundus structure. A B-scan is necessary in these cases. If you can’t find the macula, move the B-scan 4.5 mm temporally to the expected site of the macula. Measure from this location on the retina to the vertex of the cornea for the true optical axial length.

The IOLMaster does a much better job here, because the patient is fixating on a target and you measure a true optical axial length.

Following Refractive Surgery
Keratometry can be unreliable in patients who had a previous refractive procedure. Determining IOL power in these cases is best achieved with the following methods.

• The contact lens method. Use a trial set of contact lenses to determine the actual power of the cornea in the central plane. Place a rigid contact lens with a known base curve and plano refractive power on the patient and determine the change in the spheroequivalent refraction with and without the contact lens. If the base curve of the plano contact lens is equal to the average power of the cornea, there will be no known change in the spheroequivalence of the refraction. If the patient is 1 D more myopic with the contact lens than without, the lens is 1 D stronger than the cornea.

For example, if a patient’s spheroequivalent refraction were -2.0 D with no contact lens, and became -3.0 D with a contact lens with a base curve of 40.0 D and plano power, the cornea would have to be 1 D weaker than the contact lens (39.0 D in this case). Refractions greater than 4 D must be vertexed back to the corneal plane to avoid error due to vertex distances.

• The historical method. You have the preoperative keratometry readings, and you subtract the refractive change that took place with the refractive procedure and use that new K reading for the IOL calculation. This is usually the most accurate method, but it has its limitations, because it is often hard to obtain the data on a patient who was operated on by another surgeon.
Patients who underwent radial keratotomy can be even more difficult to measure. For instance, take a patient who was -5.0 D from an RK and came out plano in one year. He ended up farsighted, at +1.0 D to +3.0 D after 10 years, and then he developed a cataract. Now, that patient has gone from -5.0 D to +3.0 D, and you don’t know whether that last 3.0 D is compounded by a hyperopic or myopic shift from the cataract.

• Corneal topography. Average the central 3-mm measurements on the numerical map, part of the software on the corneal topographic equipment.
In the next year, we will have devices that will accurately measure the posterior surface of the cornea as well as the anterior surface. With those instruments, we should achieve a much better estimate of the true corneal power. 

Dr. Holladay, who is in private practice, is a clinical professor of ophthalmology at Baylor College of Medicine in Houston. He may be reached at (713) 668-7337, fax: (713) 668-7336, or via e-mail at docholladay@docholladay.com. His website is www.docholladay.com.
Dr. Wallace, who is in private practice, is an assistant clinical professor of ophthalmology at Tulane School of Medicine in New Orleans and is past president of the Society for Excellence in Eyecare and the American College of Eye Surgeons.