 So my name is David DeMille, I'm a fourth year medical student here at the University of Utah, and when I was trying to pick a topic to present today, my first thought was to pick a topic that was very straightforward and did not have a lot of strong opinions about it, and I don't know how I decided on RK after that, but it's a very interesting topic. So the first thing I want to do is talk about the history of intraocular lens calculations just before I start talking about RK. And so the first generation formulas that came out were theoretical and regression formulas, and one of those, the first one in 1967 was by Dr. Fyodorov, who was a famous Russian ophthalmologist who made a lot of great contributions to the field of ophthalmology, and he was also the father of RK. He died 11 years ago, but some of the issues with these first generation formulas is there was no means of predicting the IOL final resting position in the eye, so a constant value is used for this, for the effective lens position, and these also had issues with extremes of axial length. The second generation formulas, like the SRK2, they made adjustments to try to account for extremes in axial length, and they also attempted to predict the effective lens position based on axial length measurements. And the third generation formulas came out later, and they were created from mergers of the linear regression models with theoretical eye models, and I think the Hoffer-Q of the Holiday 1 and the SRK2 are very common formulas that we're familiar with, and these all can be optimized using either the surgeon factor in a constant or the anterior chamber depth, and some of these are known to be slightly better at different axial lengths mentioned there. The Hoffer-Q at smaller axial lengths, the SRKT at a little higher axial lengths. These formulas, one thing that they do is they assume that the anterior and posterior segments of the eye are mostly proportional, and they use a combination of axial length and keratometry to estimate the effective lens position, whereas the second generation formulas only use the axial length. So what this does do, if the central cornea power is very low, as in somebody that's had LASIK or RK, the formula assumes that the anterior chamber is shallow, and so it'll call for a lens with less power. So the fourth generation formulas came about and tried to solve a little bit of that issue amongst other things. The two formulas are the highest on the Holiday 2, and I'll just talk about the Holiday 2. It's a formula that uses these seven characteristics to help estimate the effective lens position, and these are very good for extremes of axial length. And the highest does an excellent job as well. So now moving on to RK, I wanted to just kind of talk about the history of intraocular lens calculations because it does relate to RK and some of the difficulties that are found in these. So RK involves corneal incisions in a radial or spoke-like pattern that cause the outer cornea to relax and the central cornea to flatten. And RK is rarely done nowadays, but there were an estimated 1.2 million people who underwent RK in the 80s, according to the Perk study that was done about RK. And refractive outcomes after cataract surgery and such patients are very difficult to predict. And what I want to talk about today, I know that after years of experience, many have come to find what works for them, what works best. So I just want to kind of paint a broad overview of the topic and give some recommendations of something I found in some research I did with Dr. Moshevar. So among the challenges in IOL calculation after RK, two of the main ones are determining the central corneal power and determining the effective lens position. There are other ones, patients get progressively hyper-opic with time. The number of RK incisions can affect outcomes. But we're just going to focus on these main two complications today. So after RK, RK does flatten the anterior and posterior corneal surfaces in a small central optical zone. And standard caretometry measures an intermediate area, oftentimes kind of on the outside there. And it extrapolates this central power based on broad assumptions. And this typically causes an overestimation of the central corneal power because it's steeper out here and flatter here. So if these measurements are taking it in the outer zone, it's not going to give a correct reading. So using the central corneal power rather than topography derived caretometry or standard caretometry improves IOL calculations after corneal refractive surgery. There are several machines that can measure central corneal power. I just listed a couple of them. There's more like the OrbScan. The OCT can do that too. The reason I mentioned these, there are two studies that talked about the ISIS-3000 and the topographic modeling system. And then in a study that we did, we looked at the Humphrey Atlas and the Pentacam. So these are important in determining the central corneal power. And these are better than using just standard caretometry measurements. The other major issue is the effective lens position. As I mentioned, the third generation formulas estimates effective lens position based on Ks and axial length. But in RK, I mentioned the central corneal power flattens, but the effective lens position is only mildly affected. And so you can see here, if someone has a short axial length and they have a flat cornea, the effective lens position is shallow. What you do if you make the cornea even flatter, then the effective lens position will be assumed to be shallower with low K values. So to solve this problem, the Holiday 2 formula does account for this error, so that's a great formula for that. And in 2003, Dr. Aaron Berry came out with this double K method, which can be applied to third generation formulas in post-refractive surgery eyes. And what this does is it uses pre-op refractive surgery Ks to determine the effective lens position, and it uses post-op Ks to calculate the refractive power of the eye. And apparently the mean K values for the population are 43.86. So people do not have their K values before they got RK for the whole population. And that's a good question. I don't know why the Holiday 2 formula uses that as well. I don't know why Dr. Warren Hill, his article mentioned that this should be used if pre-op Ks are unknown. That's a good point. I don't know why. He just said that was the mean for the population and that's what... Sorry, I don't have a better answer for that. So there are two main studies that use correct formulas for effective lens position, and use correct measurements for good machines that can measure central corneal power. Two big ones that I'm aware of. There was one in 2004 by Dr. Packer where they did a retrospective study. All these studies are with a very small amount of eyes. They used 14 eyes with RK, 3 eyes with LTK, and 3 eyes with HK. And they used...to determine the corneal power, they used the Isis 3000, which takes the corneal power from a 3 millimeter optical zone, and they give a little more...the formula actually counts the central part of the cornea a little stronger than the outer part of the 3 millimeters. And they also used the Holiday 2 formula for the effective lens position. And they claimed that targeting ametropia, all eyes had a post-op refaction within plus or minus one and a half diopters, and 80 percent were within half a diopter, which was pretty impressive. There was another study done in 2007 by Dr. Awad, and they looked at 16 eyes as well. They used the topography modeling system to measure the central corneal power, and they used the double K adjusted Holiday 1 IOL formula. And they actually looked at 1, 2, 3, and 4 millimeter optical zones, and they found that the measurements using up to 3 millimeters were the best. And they found that 87.5 percent of eyes were within half a diopter of ametropia, and 100 percent were within one diopter. So I thought those were really good results. And when we did our retrospective study, I thought we were going to get something similar. We didn't get quite those results. So what we wanted to do in a study I did with Dr. Mosfar is to assess the Ascaris post-RK IOL calculator from the American Society of Cat Art and Refractive Surgeons. There's a nice article about the LASIK calculator that many are familiar with, and there hasn't been an article about this yet. So what this does, you input your data, the centricornial power measurements can go in here. You can use the corneal isis, the Humphrey Atlas, or you can put in whatever other machine you use. We chose to use the Pentacam equivalent K reading, which Dr. Warren Hill wrote an article about how that can be used for the centricornial power reading. So our goal was to evaluate this calculator. We did not, since this was a retrospective study, we have not used the isis corneal topographer in our measurements, so we did not evaluate that. So we looked at the Humphrey Atlas one through four measurements, where it measures out to the four millimeter edge of the optical zone. We used the average central power method with the Pentacam, and we used an average of these two methods here. So we also had 16 Is. What we did, when we used the IOL calculator in this study, the actual refraction after cataract surgery was targeted, and the IOL prediction error was obtained by subtracting the predicted IOL power from the power of the IOL implanted. That took me a few times to read and understand that when we were making this paper. So, for example, if we used a 20-diopter lens in a patient, and they came out minus half a diopter, we plugged that half a diopter back into the IOL calculator and saw what came out. So, for example, in the Humphrey Atlas method, putting in half a diopter as the target refraction, the Humphrey Atlas suggested a 19.61-diopter lens. We know that a 20-diopter lens came out as minus a half, so if 19.61 is what came out, we know the IOL prediction error was plus 0.39, which would have resulted in a hyper-opic outcome. So what we are trying to do here is measure the IOL prediction errors of the various of these methods. So this one here is the Pentacam. This is the Humphrey Atlas. This is the ASCRS average. These are the mean IOL prediction errors. The positive values mean they come out with a hyper-opic outcome. So even using the central coronary power measurements and using the ELP estimation, we still came out with hyper-opic outcomes. And the patients themselves didn't end up hyper-opic. This was just after we calculated these IOL prediction errors. I shouldn't say not any of them ended up hyper-opic. Some of them may have, but... So if we use an assumption that was mentioned by Dr. Hill that one diopter of IOL prediction error produces 0.7 diopters of refractive error, we can deduce that the average central power leaves a patient about plus 0.38 diopters after cataract surgery, the Humphrey Atlas plus 0.75, and the ASCRS average about 0.57. So patients are still coming out hyper-opic in our study. This is a box plot that just shows, if you're familiar with box plot, this line here is the median, and this is the upper quartile, lower quartile, minimum, maximum, and these are outliers. So you can see that the patients are still ending up positive, which is hyper-opic. So the percentage of I's using these methods that ended up within these categories are found here. And so we found the average central power using the pentacam came out with these values, the Humphrey Atlas. It was surprisingly that nobody ended up within half a diopter after using that. It was pretty consistently ending up with hyper-opic results, as you can see there, and using a combination of the two, so these two averaged as this, came out with those values. So further studies are obviously needed to, you know, use multiple machines to measure the central corneal power on the same patients, stratify outcomes based on the number of RK incisions. We tried to do that. We didn't have the number of patients necessary to do that. Different IOLs are used that could possibly affect outcomes, and we need to confirm findings with larger numbers of patients. So in summary, central corneal power measurements and ELP estimation are very important in IOL calculations and post-arcade patients. The methods outlined by Packer and Awad produced excellent results, as I mentioned. But the one thing we like about the ASCRS calculator is that it's accessible online. It's very simple to use. And to use this, we do suggest, if you use the pentacam or the Humphrey Atlas, we do suggest adding about 0.5 to 1 diopters to the IOL prediction, or 1 to 1.5 diopters to those two predictions. And that's to try to best approximate ametropia. You know, it's always safe to hedge your bets a little bit and go a little more myopic than that. So I think it's appropriate to use this calculator if we keep these thoughts in mind. And that's it. Sorry, I know it was kind of fast, but there's a lot to cover. Any questions? Yeah, thanks. And I agree. So aiming for about minus 0.75 from our data would show that's the best chance of getting ametropia in the near term. But like you said, I would even suggest aiming for more myopia for that. Yeah. There was an article last year about the cataract and refractive surgery journal in July 2010 where Dr. Maske actually talks about that. You know, the measurements can actually change during the day. So the time of day you take the measurements can make a difference as well. Thanks for bringing up the contact lens method. I forgot to mention there was an article by Dr. Chen and Dr. Mark Manis from UC Davis in 2004, I believe, where they used the contact lens method and the historical method and they used the flattest case from those two and they published their outcomes and still they wanted to aim for myopia. But it's a good thing. The more measurements the better. And to average those out it gets rid of the outliers and any other thoughts? Yeah.