 So I have to admit, last night at about 11.30 at night, when I was going over and over this presentation, I was secretly hoping to myself there might be some massive eye emergency affecting everyone in Salt Lake City and the turnout would be pretty poor. Fortunately, it looks like for the people of Salt Lake, that was not the case. All of you were able to join us this morning. So as I was introduced earlier, my name is Rachel Simpson, I'm a fourth year medical student at the University of Arizona. I'm very happy to be joining you here at Moran for the month of July doing research with Dr. Moschvar. And I'll be talking to you today about an issue that I think you'll find is pretty interesting and also pretty important to discuss, medical errors and refractive surgery. So I would imagine most, if not all of you were, pretty familiar with this book, Waking Up Blinds. And if you're not, you probably should be. But it's a pretty harrowing account of stories of medical errors and malpractice in the field of ophthalmology. And I just want to quickly talk about one of the cases, Mr. Houston, a 73-year-old gentleman who was getting along just fine. He had pretty good vision out of one eye, pretty poor vision in the other, but wasn't really experiencing any compromised activities of daily living. He could drive, he could read. One day, his local ophthalmologist suggested, hey, maybe you want to think about getting that treated and send him to a major referral center. The ophthalmologist there suggested surgery. He agreed to it. Of course, what ended up happening was that he had surgery on the wrong eye, experienced a number of complications. Two years later, he was completely blind. You can see this issue is, it's pretty important. And taking a step back, just looking at the broader issue of medical errors in general, really came to a forefront. In the late 1990s, there were a number of media exposés about how serious this problem really was. In response, the Institute of Medicine put together a task force that basically came up with, to errors humans, huge guidebook that looked at where are all these errors coming from, what can we do to stop them. It's from there that we got this definition of medical error. We also got things like the timeout protocol, which is of course that mandatory pause where we take a second and just make sure, do we have the right patient? Do we have the right materials? Are we operating on the right side of the body? Do we have everything we need? And these interventions were helpful, but you can see from this slide, they really didn't fix the problem. This scope of the problem is still pretty significant. That number one statistic right there is pretty interesting. It would be the sixth leading cause of death if the CDC was able to actually list medical error as a cause of death. So what we wanted to do is look at this within the scope of refractive surgery. And we felt like it's particularly important in refractive surgery because there is no margin for error. These patients are coming to us, and yes, they might have glasses or contacts, but for the most part, they can see just fine. They have no underlying ocular pathology. There's no true medical indication for surgery. So we're taking someone with the best corrected visual acuity of 2020 almost every time. We're operating on them, and then if there's an error, we're risking permanent visual compromise. So we reached out to ophthalmologists across the country and asked them for cases of preventable errors during late-six surgery or refractive surgery. And by preventable, we were looking at only those errors that happened in the preoperative period. We collected 18 patients, 15 patients experienced unilateral errors. Three were bilateral, and there was one unfortunate gentleman who had two errors in the same eye during two separate procedures, which I'll tell you about in just a second. We analyzed the data based on where these errors were coming from, how the patients did afterwards, and then we also looked at how can we prevent this in the future. So we identified three major sources of error. Cylinder conversion error, which I'll talk about in just a second. Data entry error, which is pretty self-explanatory, although there's some interesting stories associated with that. And if we have time, I'll be happy to share them with you. And then timeout protocol failure, which again is pretty self-explanatory. The timeout protocol was not performed as it should have been. So I just want to take a second and talk about cylinder notation because I'm sure the vast majority of you in this room are quite familiar with it. But as a medical student, it was something I was pretty new to. And also, if we work through sample conversion, I think you'll see, even though the math is extremely basic, there's a number of opportunities for errors to be made. So there's no standard on notation that can be positive or negative, and it really varies widely across institutions. So we have a sample MR here of minus 350, minus 150, at 80, and negative cylinder. If we wanted to convert it, the first step would be switching the cylinder sign from negative to positive. So now we have minus 350 plus 150 at 80. Our second step, we're going to add the cylinder to the sphere. So in this case, it's minus 350 plus a minus 150. So you can see already, if you happen to forget that pesky little negative sign, it's already a difference of three diaphthers with the sphere. So our patient is now minus 5 plus 150 at 80. Our last step is to convert the axis. So we look at that last number. If it's less than 90, we're going to add 90. If it's greater than 90, we'll subtract. So for our patient, it's less. We're going to add 90. And we have our final and positive cylinder notation. Again, this is not rocket science, but there's three steps. An error can definitely occur. And our first case is an example of that error. A 57-year-old gentleman presenting for lasik surgery with a pre-op MR of minus 6, minus 2, at 180. And you can see here clearly there was an error when his notation was converted. He was supposed to be treated at minus 8 plus 2 at 90 and was instead treated at minus 8 plus 2 at 180. His post-op, he experienced basically a doubling of his astigmatism. This gentleman was offered retreatment with lasik, and I guess he was feeling trusting of us if he agreed. Unfortunately for him, a second error of cylinder conversion occurred. Some lucky gentleman ended up with a post-op MR of 125 plus 725 at 130. If you remember, his initial astigmatism was plus 2. So he experienced some pretty significant visual compromise. He required multiple follow-up procedures. He had two relaxing incisions and a PRK. He was still only able to achieve a best-corrected visual cutie of 2030 with contact lenses. So a pretty significant outcome for a man who before surgery saw a fine 2020 with contact lenses. Second example is an example of data entry, and I'll just draw your attention to the OD. You can see he was a hyperopic pre-operatively at 380 minus 50 at 75, and somehow that pesky negative sign came into play here, and a hyperopic patient was treated with a myopic procedure resulting in almost doubling of his sphere. He again also experienced pretty significant visual compromise post-operatively. He was not a candidate for surgical follow-up and was treated with contact lenses. And here's actually an example from the chart of that very error. You can see in the chart he's plus 380 who's entered in the computer at minus 380. Our third case is a timeout failure. I'm not going to go over his MR for you. I'll just tell you the story of how this happened. Patient had a name that was quite similar to another patient that was having refractive surgery that same day. When his name was called in the waiting room, he responded to the other patient's name. During the timeout procedure, when the doctor went to verify his name, he said, oh, I see your name's Joe. So if it's okay with you, I'm going to call you Joe, okay? Proceeded to calm Joe throughout the procedure. It wasn't until the very end that the patient said, you know, Doc, you keep calling me Joe. My name's actually John. Clearly, at that point, the mistake was realized and the patient had been treated with the wrong patients' MR. Fortunately for this patient, the only thing was under correction of myopia. He was a good candidate for re-treatment with LASIK and didn't experience any long-term visual compromise as a result. So overall, definitely the most common error we discovered was cylinder conversion error, followed by data entry, and then the timeout failures. Interestingly, all four of those timeout failures, it was the wrong patient that was operated on. So overall, these patients entered surgery with the best corrected visual QD of 2020. They saw just fine. The average postoperative best corrected was 2025. So that's a pretty significant finding, again, in patients who didn't need to have surgery. Six patients lost at least one line of best corrected visual QD. 16 of them were good candidates for surgical follow-up. Three of them said, thanks, but no thanks. I'll take my chances with contact lenses. The patients who were treated surgically, they did okay. But interestingly, 18 total procedures were needed to correct the errors in these patients. And again, we're talking about an elective procedure. So it's really pretty a significant number. Nine patients were treated with contact lenses. The ones that did really poorly were the non-surgical candidates. As you can see, their best corrected visual QD, even with contact lenses, was significantly compromised. Our best outcomes, as I said, were seen in the surgical candidates. If we break that down by error, it's the cylinder conversion and timeout protocol had pretty similar outcomes. We think it's pretty coincidental. We don't really think that the nature of the error was predictive of how the patients were going to do. We think basically, given the small sample size, it was just luck. Poor outcomes were seen in the non-surgical candidates. And it just so happened that most of these fell in the data entry category. So overall, our recommendations are we want to look at maybe standardizing the cylinder notation since it seems like the vast majority of our errors were observed during this portion. And also, if you look in the literature, there's really a paucity of data available on these kind of errors. But there is three case reports, and they all report cylinder notation error as the source of error. Secondly, you're going to want to make sure that proper timeout is performed every single time and that the patient is saying their own name. And then third, we need to be verifying their refractive treatment parameters every time, integrating this into part of the timeout protocol so that you're looking at the chart, looking at what's entered in the computer, and making sure that everything is matching up. And that is it. Be happy to answer any questions you have.