THE WORLD OF COMPUTER PROgramming gave us a great acronym, GIGO--garbage in, garbage out--to describe why computer programs sometimes fail to provide the results their programmers intended. As it turns out, GIGO has applications in many other fields, including wavefront sensing. Occasionally, a patient's wavefront output doesn't look right, or, more often, his wavefront refraction doesn't jibe with his manifest, and the ophthalmologist is left to decide whether or not a clean image went into the device in the first place, or if the manifest deserves more weight than the wavefront refraction. In this article, surgeons share their strategies for handling situations where some aspect of the wavefront doesn't seem quite right.

Getting a Sharp Wavefront

Greensboro, N.C., surgeon Karl Stonecipher recommends making sure the wavefront image is valid before moving on to calculating a possible refractive target. Here, he uses the WaveLight Allegretto machine he's currently studying for an FDA trial to explain the steps he takes.

"First, once the wavefront has been taken, we look at the iris image to make sure the display is centered along the pupil margin," he says. "Second, we look to make sure that the eyelid isn't in the way, and there's not a dry spot on the cornea. We want to make sure that we've got a good image, even before the machine has read it, and also to make sure that the image is the proper 7 mm in diameter."

Next, he moves onto an analyzer screen where he can look at the actual pattern of dots that comprise the wavefront. "We look at the red central dots, the green ones in the mid-periphery, and the blue ones in the outermost area to make sure there's a good pattern," he explains. "They'll be either solid, hollow, crowded or just not there. Hollow spots tell you that there's some information there but not enough to glean the data you need to treat. Absent spots are usually related to something like a dry spot on the cornea, a cataractous change or something in the vitreous. "You may still get a wavefront refraction from the exam," says Dr. Stonecipher, "but if the dot pattern's bad, it's noise, not valid data. I had a patient just the other day who had a posterior vitreous detachment and so many vitreous opacities that we couldn't get a good dot pattern. So, since we couldn't get a good wavefront, we did a conventional LASIK procedure on him."

Louis Probst, MD

Vitreous changes, lens clouding or corneal opacities such as that shown here can result in bad data on the wavefront map, and the need to switch to a non-custom LASIK.

Louis Probst, MD, who practices in Chicago and Madison, Wis., is also mindful of issues that might skew a wavefront refraction. "It can be an opacity on the cornea, an irregular pupil, a pupillary membrane, a very small pupil, or any opacity in the visual system," he says. "In these cases, we'd just perform a conventional treatment because of how the wavefront works, it would be incorrect if it were being blocked or altered in any way." He advises Visx users to be suspicious of wavefront refractions gleaned from pupils that are smaller than 6 mm. "Especially when they get to 5.5 or 5 mm, the results of the wavefront scan don't seem to be as accurate," he says.

Manifest vs. Wavefront Refractions

In addition to obvious problems with a wavefront capture that will appear on the wavefront image itself, there is also the occasional issue that crops up when the aberrometer's refraction doesn't match the manifest refraction.

• Spherical aberration and accommodation. "The most common reason the manifest and wavefront refractions don't match up is accommodative," says Minneapolis surgeon David Hardten, speaking from his use of the Visx WaveScan system. "So, they're accommodating on the manifest or on the wavefront."
  
  Accommodation isn't the only reason for a discrepancy, however.
  
  "If the manifest and wavefront refractions don't match up, I go back and compare my manifest refraction to a cycloplegic refraction. If those are very similar, then on the wavefront refraction there are two possibilities: The patient could be more myopic on it or more hyperopic on it."
  
  If the patient is more hyperopic on the wavefront refraction, Dr. Hardten says the first possibility to explore is whether the discrepancy could be the result of spherical aberration. To determine if this is the case, compare the wavefront refractions at different pupil sizes that the software allows on the wavefront analyzer. If they change as you change the pupil size, it's probably spherical aberration, says Dr. Hardten.
  
  As an example, Dr. Hardten offers a hypothetical case of someone who has had LASIK surgery previously, and whose cornea is very flat in the middle but steeper in the mid-periphery. The patient may have a wavefront refraction at a 4-mm pupil size that's less myopic than the manifest refraction because, with the latter, the patient is being refracted at a 5 or 6 mm pupil. So, at the 3-mm pupil size, the refraction might be plano, at 4 mm it might be -0.75 D, at 5 mm it might be -1.25 D and at 6 mm it might be -1.5 D.
  
  "With this type of patient, as long as you feel comfortable that he isn't accommodating on the wavefront because you determine that at least one of those refractions is similar to the manifest refraction, then I think that's a situation where you trust the machine and proceed with the case," explains Dr. Hardten. "What the machine is going to do is correct for the spherical aberration component of it and, hopefully, reduce the variance of the refractive error between the different methods."
  
  Dr. Hardten says that the other reason someone may refract as more hyperopic on the wavefront is a "pretty rare one," but was discussed at a 2004 meeting by London surgeon Julian Stevens. "In some very young patients with an extremely shiny retinal nerve fiber layer, sometimes the wavefront will reflect off of a retinal area that's more superficial," Dr. Hardten says. "So, the patient may have a more hyperopic refraction by 0.5 D or so. I haven't seen this personally, but since it was something Dr. Stevens described, I'm on the lookout for it."
  
  If the patient is more myopic on the wavefront refraction than on his manifest, surgeons say this is where accommodation issues appear.
  
  "When you have a patient, usually someone who's younger, who tends to accommodate during the wavefront test, the wavefront will tend to measure a much more minus prescription than he actually has, and more myopic than the manifest refraction," says Dr. Probst.
  
  

  David Hardten, MD

  Surgeons say in many cases accommodation can cause the wavefront refraction to appear more myopic than the manifest refraction. Here, the patient was accommodating at first (top). After relaxing his accommodation, a new refraction was done (bottom).

  
  "When the patient is accommodating during the wavefront, a few things happen," explains Dr. Hardten. "Not only does the sphere change, but, sometimes, because the lens is accommodating and is a different shape, the cylinder changes, as does the coma or higher-order aberrations. So, this isn't the ideal situation in which to best adjust the sphere measurement using the system's surgeon adjustment factor [which allows the surgeon to adjust the spherical correction by a limited amount]. What you want to do, in general, is to try to get the patient to relax his accommodation for the wavefront."
  
  Surgeons use a number of different methods to try to defeat accommodation and improve the validity of a wavefront refraction.
  
  "We may give them an Ativan or a Valium and have them calm down for 20 to 30 minutes," says Dr. Probst.
  
  "While the patient is still in my office for the preoperative evaluation, I'll get a wavefront analysis that's dilated or cyclopleged," says Dr. Hardten. "If the cyclopleged wavefront analysis shows very similar refractive information to what it showed without dilation, then I know it's probably spherical aberration, or higher-order aberration issues in general, that's causing the difference between the refractive numbers. If, however, on the cyclopleged one, they move back more toward their manifest refraction, then I know they were just accommodating on the machine."
  
  If the patient was accommodating during the wavefront analysis, Dr. Hardten recommends a number of techniques to try to get him to relax for another wavefront, either on the day of surgery or at a separate exam preop:
  
• Have the patient gaze past the target in the wavefront machine;
  
• Distract the patient by having him tap his feet, fidget or talk to the technician;
  
• Take away his glasses (if a myope) and don't let him read for 15 minutes before the exam to relax accommodation; if it's a hyperope, put +3-D "cheaters" over his glasses and have him walk around the office, looking at distance targets;
  
• Seat the patient in a dark room, since this can sometimes relax his accommodation.
  
  "If, after all of this, you can't get him to relax his accommodation, a standard LASIK treatment is typically better," says Dr. Hardten.
  
  If a patient accommodates on the second wavefront exam, Dr. Probst will take him from his Visx machine, which he uses most of the time, and try him, cyclopleged and dilated, on the Alcon CustomCornea device that he also has access to. "Then, they won't be accommodating at all, and you can see if you have a good match between the manifest and the CustomCornea refractions," he says.
  
  Ronald Krueger, MD, director of the Cleveland Clinic's refractive surgery department, says, "In some cases, I'll actually see that the patient's refraction on the day of surgery is just slightly more myopic than when I captured it before [on the Alcon CustomCornea]. That helps me to know that, yes, there may be a real, slight difference that it picked up and that's where I should use the surgeon offset feature. In that case, I'm less worried that it's different from the manifest and more interested in factoring in that it's different from the previous wavefront. I trust the one that's less myopic because I want to avoid an overcorrection, if possible."
  
• Using the surgeon adjustment factor. When surgeons feel the wavefront refraction's values are within the acceptable range as defined by the laser's manufacturer, the custom systems allow their users to make small adjustments to the wavefront refraction's spherical measurement to get the final correction as sharp as possible. At least in the United States, users are limited to small adjustments, usually around .75 D.
  
  "I usually won't adjust the sphere more than ±0.5 D," says Dr. Stonecipher. "I won't adjust for more because I think your accuracy diminishes dramatically above those levels."
  
  "About 80 percent of the CustomVue patients need some type of adjustment in their sphere to get fine tuned," says Dr. Probst, who uses up to 0.75 D for adjustments to help the wavefront refraction match the manifest more closely. "And that has worked very well for us."
  
  He will also sometimes make this type of adjustment based on the patient's age.
  
  "Sometimes, for patients who are younger, like 20 years old, I'll increase their myopic correction by about
  0.5 D or so," says Dr. Probst, "because they'll likely regress a little more, being so young. And, if you overcorrect them by a small amount, they'll actually prefer it because it will give them sharper vision."
  
  Dr. Krueger also uses the adjustment factor, and will vary it based on how he made the flap. In a flap-only study he did at his practice, he found that using mechanical microkeratomes (specifically the now-defunct Alcon SKBM and the Moria M2) not only induces some aberrations, but also causes a small refractive change, shifting the eye slightly toward hyperopia. "The reason a flap alone causes a change in refraction is because of biomechanical changes in the cornea coming from a meniscus-shaped flap," he says. "When using a mechanical microkeratome like the Moria on a myope, I'll generally use no offset for patients in their 20s, maybe +0.25-D offset [to decrease the correction] for people in their 30s and 40s, and maybe up to +0.5-D offset for patients in their 50s.
  
  "I'm using fewer offsets when I perfrom CustomCornea LASIK with IntraLase because of an initial greater incidence of undercorrection with the previous settings. I'm in the midst of doing a similar flap-only study of the IntraLase right now and I believe it won't show that shift toward hyperopia because the planar-shaped flap it creates induces fewer biomechnical shifts in refraction."
  
  In most close-call cases, Dr. Hardten trusts his CustomVue machine. "If the manifest and wavefront refractions are close, I trust it," he says. "That's why the FDA study parameters of 0.5 D difference on the myopic side and +0.75 D on the hyperopic side were designed. If the discrepancy is in that range, the correction is probably accurate most of the time. If it's outside that, you have to determine why and how to manage it. If it's only slightly outside that range due to accommodation, I will often still go ahead and use the wavefront refraction, even though I know they were accommodating a tiny bit on the machine. I'd use the surgeon adjustment factor to account for it."
  
  Even though aberrometers are capable of very sensitive mapping of the visual system, Dr. Probst says the old methods of refraction will be around awhile. "When wavefront first came out, some people said, 'The days of manifest refractions are over! Phoropters will be obsolete!' But it's become clear we can't turn over all our refraction responsibilities to the wavefront machine completely," he says. "We still really need to do a manifest refraction with the phoropter, because that's our baseline. It's what we know works, and we need to have a subjective test we know the patients can see well with."