All right, everybody. I think we'll get started again. It is a distinct honor and a pleasure for me to introduce Dan Resnick, who is going to speak on the topic of degenerative spondylolisthesis. Dan is well known to all of you. He's the former chair of the spine section. He's the current president of NASS and is always full of phenomenal things to say. Dan is the winner of the Volker Sontag lecture. Congratulations, Dan. Well, thanks. I would agree that I'm certainly full of something. Will you guys be the judge of that in just a few minutes? I'm going to be talking a little bit about improving the quality of spine care with regard to spondylolisthesis. But a lot of the themes I'm going to touch upon are universal and apply pretty much to all fields of neurosurgery and all fields of spine surgery. I'm just using spondylolisthesis because it's a very good concrete example of something that we sort of think of as a slam dunk. It turns out it's not. These are my disclosures. I am a spine surgeon. I'm in favor of doing spine surgery. It's how I earn a living. I'm involved with Zoe in a project called NITUS, which really doesn't have much to do with this. I am president of NASS. I'm the fifth neurosurgeon who's been president of NASS. Volker was the second, and so I follow in his footsteps there. And I'm a huge fan of Volker, and I'm very honored to be able to give this lecture, which is named for him. He is really responsible for a great deal of what we're allowed to do as neurosurgeons and as spine surgeons, and we are all tremendously in his debt. So why is this a timely topic? If you have not seen this map, you've been living in a cave someplace in the northern Andes for quite some time. The performance of lumbar fusion has been questioned by many. This is the Dartmouth Atlas, of course, which shows that there's a significant amount of variation in the performance of lumbar fusions in different patient populations based upon county of residence. This high variability of rates has been interpreted as a lack of agreement regarding the indications for surgery, and there's been significant questions by critics of spine surgery as to the motivation for the high rates of surgery in some counties without a demonstrable improvement in health quality outcomes. There's been a huge shift in our conversations about health care away from fee-for-service, away from quantity towards paying for things such as value and quality. These are very appealing words. They seem to make sense. It's sort of the Justice Potter definition of pornography. I can't really describe it, but I know it when I see it type of thing. It's been very difficult to operationalize, and currently we're generally measured on process measures and patient satisfaction scores, which I'm going to spend just a few moments discussing why I think they're useless, expensive, misleading, manipulatable, and, quite frankly, dangerous. These are my avatar scores. This is the HCaTS version that we use at the University of Wisconsin. As you can see on the top slide, I've got green boxes. Green boxes means that I'm sitting with my patients. I'm holding their hands and staring soulfully into their eyes. I'm feeling their pain. I'm doing everything in my power to make them feel better. In the bottom panel, I have lots of red boxes. This means that I'm actively insulting my patients, perhaps even assaulting them, as I kick them out of my office, telling them to lose weight, quit smoking, and get a job. If you look at the actual numbers, I'm about 30 points away from myself. If we assume that the standard deviation on this scale is plus or minus five points, that means that I'm six standard deviations away from myself. Those of you who have any background in math will realize that that means that I cannot exist. So not only does this mean this is me, same guy, same basic patient population, two different clinics. What am I supposed to do with this information? I have no idea what I'm supposed to do with this information. Cancel the one clinic? In reporting this type of data, that is very expensive. You have to hire these companies to do these surveys, to mail them out to get a 6% return rate, but it still costs millions and millions and billions and billions of dollars. Four specialties, primary care, pediatrics, OB, and I forgot, and internal medicine spend $15.4 billion. That has nothing to do with that's only four specialties out of all of medicine, costing $15.4 billion to report these measures, which I think I want to demonstrate to you are completely worthless. These patient satisfactions have repeatedly been shown to have no correlation whatsoever with actual medical outcomes, and when there is an association between patient satisfaction measures and actual medical outcomes, it's almost always the inverse, and I'll demonstrate that to you. The top article is the Fenton article, which is really the first article that came out, looking at the relationship between patient satisfaction measures and actual medical outcomes, and what Fenton described was that the patients who were in the highest quintile in terms of patient satisfaction, so the ones that filled those surveys and they loved their doctor, they were the ones that had the highest medical cost, the highest morbidity, highest mortality, and the most frequent admissions to the hospital. So the patients who were treated appropriately, got better, went home, didn't really like the physician they saw. The ones who were sick liked their physicians and rated them higher. General surgery, same thing. I know you can't read the text, but basically when you're looking at hernia surgery, cholecystectomy, routine general surgery cases, patient satisfaction is inversely related to mortality, morbidity, and costs. In addition to being somewhat misleading with regards to the correlation with actual medical outcomes, these systems are very manipulatable, and in fact, there are five ways you can use to improve your patient satisfaction surveys if you wish to do so. If you hire a different consulting company, they'll recommend six ways to improve your patient satisfaction scores. If you really want to get fancy, you can get seven ways to improve your patient satisfaction scores. If you're going to go all the way over the top, there are actually eight ways you can use to improve your patient satisfaction scores. All you need to do is to hire one of these companies for about a couple hundred thousand dollars, and they'll explain five, six, seven, or eight ways that you can use to improve your patient satisfaction scores. Based upon my review of this literature, it appears that the best way to improve your patient satisfaction scores is to be a very attractive, racially ambiguous woman who appears in every single one of these articles. As you go through this, you start to get a little bit fed up, but what really matters, it turns out that you don't have to do any of those things. All you have to do is bring your heart to work, and that will improve your patient satisfaction scores. So we can just bring our hearts to work. I mean, please. It's hard not to get nauseated when you start going through this type of stuff. Then, of course, there are the independent quality sites, the MedMD, all those types. This is the vitals.com. I looked up myself because I was a visiting professor at another place, and I learned that at that institution, their pay was directly related to their patient satisfaction scores on one of these websites. So I'm like, well, geez, I've never even checked mine, so I figured I'd check mine. So I checked mine. My scores suck. Put that out there. But what was interesting is as I scrolled through these things, I realized I only had three people who had reviewed me. There are three people, three physicians with my name in the I'm not sure if these reviews were even of me. And when I got to the bottom, it recommended that if you weren't really satisfied with Dr. Resnick, a similar physician available in our community was Dr. John Sandin. Dr. John Sandin wasn't my partner. He died in 2005. So the reliability and usefulness of this type of information is highly questionable. It's also dangerous. By trying to chase these numbers, we are incentivized to do stupid things. We are incentivized to offer patients surgery who may not need surgery. We are incentivized to offer patients narcotics inappropriately. We are incentivized to order expensive tests and order expensive therapies that are inappropriate. Erica recently reported that increased patient disability and recommended for non-surgical treatment both independently predicted poor patient satisfaction scores. So this is the exact opposite of what we're supposed to. We already have enough perverse incentives in our lives. We don't need another one. I did my internship in Philadelphia at a hospital. I had a very busy vascular surgery practice. And the most popular surgeon there was a surgeon whose name was Ron, or could have been Ron. Ron was not a very good surgeon. The classical vascular patient in his practice would go with an aorta by fem, fem pop, four-foot amputation, below knee amputation, above knee amputation, death. And this occurred generally about a seven- to eight-month period. It is widely known that Ron was not the technical genius of a surgeon, but his patients absolutely loved him. And the reason was Ron would meet with every patient beforehand, and he would spend the night in their room the night before surgery having a cigarette with them, literally, talking about how he was going to do the absolute best he could to help them with their problem. And they absolutely adored him, even though his outcomes clinically were absolutely horrible. And we are being incentivized to do that by these packages. So instead of patient satisfaction measures, Zoe, myself, Tony Asher, many of us for years have been calling for the use of patient-reported outcomes as a better metric for determining quality of care. We feel they're more relevant, they're more controllable, they're more comparable across different treatment sites. Results can be risk-adjusted so that if you're taking care of a sicker patient population, you don't necessarily get hammered. So if the ASC down the street is cherry-picking all the 32-year-old athletes with fresh hernia discs and you are left with the, you know, high practice, the 56-year-old morbidly obese diabetic who smokes with multi-level stenosis, you know, you can correct for that a little bit. So what's not to love? Well, it turns out there's a whole bunch of these patient-reported outcome measures out there. And it turns out that different outcome measures are sensitive for different things, and different outcome measures can tell you different things, and you have to be very careful which ones you pick for which disease states that you're treating. And I'm going to now talk specifically about the issue of spondylolisthesis, because that was the title of my talk. And the reason why this is relevant is that the issue of to fuse or not to fuse in spondylolisthesis has been something that we've been arguing back and forth about for decades at this point. To this day, it is still considered, whenever we get in a group and talk about what should we measure in terms of something that we can be sure that is good for fusion, we always talk about spondylolisthesis. Well, it turns out it's not quite as clear as it would seem. There's my Photoshop-challenged version showing the first title, the first cover from the original guidelines articles, which we published back in 2005. And back in 2005, this is 15 years ago now, there was no direct evidence to say that surgery was better than non-surgical management for the management of stenosis or spondylolisthesis. We had nothing, right? So this is one of the mainstay procedures that we all do, and we have no direct evidence saying that it makes any sense. We did have some evidence that decompression for stenosis alone made sense, and we had some moderate evidence from historical cohort studies that if a patient had a solid fusion after a non-instrumented fusion, their outcomes were better than if they had a non-union after a non-instrumented fusion. And we really had nothing in terms of instrumentation, but pretty much the most common operation being done at this time was an instrumented posterolateral fusion for these patients. So this was somewhat disturbing. In 2014, we updated the guidelines, and at that point we had the sports study. The sports study, you know, I still vividly remember Paul McCormick, before the results were published, becoming apoplectic about the problems with the sports study in terms of its study design and biases and all that type of stuff. But it turns out the sports study was the best thing that has happened to us, or up until recently the best thing that has happened to spine surgery, because it definitively showed that surgery was better than non-surgery for this problem. However, it didn't really tell us much about the use of fusion or the use of instrumentation. In fact, if you look at all the patients who were operated upon in the sports study, a certain percentage of them underwent decompression alone, a certain percentage underwent decompression and non-instrumented fusion, a certain percentage underwent decompression, non-instrument fusion, and instrumentation, and a certain percentage of them underwent all of the above plus inner body fusion. Turns out the overall results in all those patient groups are exactly the same. So if you're a policymaker or you happen to own an insurance company, you might look at that data and say, well, hold on a second. If a simple decompression has the same outcomes as a circumferential fusion, why should I pay the extra $50,000 for the circumferential fusion if the outcomes are the same? Fortunately, the authors at this point had realized that this wasn't a randomized control, they realized that this was a non-controlled cohort study and presented their data in a very rational way, so that scenario did not happen. In 2016, there were two randomized controlled studies published looking specifically at this issue, so now we have level one evidence looking at fusion for spondylolisthesis. One was those studies, which I'll talk about in a minute, the other was from Europe. So finally, we have high-quality data looking at the exact question that we've been trying to answer now for 15 years and been unable to answer. So we're all very excited when this is announced, when this is coming out. So we have these studies and what have we learned? We learned that it's kind of confusing and we're not quite as sharp as we thought we were. The first study, by Forst et al., was a randomized controlled study looking at patients with stenosis with or without spondylolisthesis who underwent fusion or either went fusion or no fusion in that patient population. And this is a randomized controlled study. They had 247 patients. The patient population was fairly poorly described. They were only required to have stenosis. Spondylolisthesis was something that was noted in association with the stenosis. They didn't do any sort of assessment of factors that we usually think about when we think about doing fusions on these patients. They didn't talk about dynamic instability. They didn't talk about disk space height, facet pathology, or anything that we usually consider when we're considering doing a surgery or a fusion or which technique we're going to use. And they also had some methodological problems that the surgical procedures weren't standardized. Forty percent of the patients had two level fusions as opposed to one level fusions and the numbers dropped off precipitously after two years and at five years, after two years they only had 56 percent follow up. But what they found was when they looked at the ODI, which was their primary outcome measure, there was no differences between the fusion and the non-fusion groups. Therefore, it is clear that fusion is not necessary in all patients with spondylolisthesis. The question remains, is fusion potentially beneficial in some patients? And that's where Zoe's study comes in. Zoe also did a randomized controlled study looking at fusion versus no fusion for patients with stenosis associated with spondylolisthesis. This is a very different study. These patients were very tightly controlled. They were all felt to have stable spondylolisthesis, so a panel of experienced surgeons, Volker being one of them, Ed Benzel being another one, went through the studies and determined that you could go either way with these patients. So when I first reviewed the methodology, I was actually very worried that this was going to be a negative study because of the requirement for equinominy on the panel. The primary outcome measure in this study was the SF36 and everybody had the same operation, the standard post-terolateral decompression, post-terolateral fusion, medial fascitectomy with Iliac crest autographed. What they found was that patients treated with fusion enjoyed better patient-reported outcomes using the SF36 score at every time point following surgery. Turns out the ODI showed a trend towards improvement but was not statistically significant. But very interesting, the rate of re-operation, I don't know why that's cut off, was significantly higher in the decompression alone group at four-year follow-up compared to the decompression and fusion group. That actually made, that actually really helped skew the results in favor of the decompression and fusion group. So what does this mean? So this means that there is a population of patients with spondylolisthesis who do benefit from fusion. How do we identify these folks? And this is an issue with classification is what this is. And just a message to take home about those two different studies, if you take two different patient populations, you do different things to those patient populations and you measure them differently, it's not necessarily surprising that you're going to come up with results that differ. And that's why it's so important that people who actually do this work participate in these studies and participate not only in doing these studies but in the interpretation and promulgation of medical information regarding these things. You can't leave this to methodological experts alone. So when we see these patients, we still have to make a choice. I included this slide, this is in Cave Creek right by Volker's house. I thought he might be familiar with the sign. But we still have to make a choice of what we're going to do and if we make a bad choice, there are consequences of that action. This is a patient of mine. This is a 74-year-old woman with neurogenic claudication. She has stenosis. She is not healthy. She is not thin. And she is not active. Does this woman need a fusion? I didn't think so. I just did a decompression alone on her. How about this person? This is a 34-year-old athlete with disabling back pain and inability to run with some radicular symptoms only brought on by exercise. Does this patient need a fusion? I thought so. I did an inner body fusion on this patient. What about this patient? A patient with a dysplastic spondylolisthesis, had back pain for years and now has a bilateral foot drop. Does this patient need a fusion? I think so. And did this patient need a fusion? This patient is labeled with bilateral L5 radiculopathy, had lateral recess stenosis at 4.5, and foraminal stenosis at 5.1. Did this patient need this operation done? Obviously I thought I did. But these are all four very different patients, but they're all classified as spondylolisthesis. So I did three, one, I did four different operations on these four patients. They all have spondylolisthesis. They all did pretty well. But when we try to communicate that in written form or in registries or in randomized studies, it is clear that there's a lot of heterogeneity in spondylolisthesis and not everybody needs a fusion. And one of the problems is that our classification system that we use is horribly out of date. This is the main classification that's currently used by QOD, Spine Tango, the NAS registry, and was used in both of the randomized studies that I described. This is a myridine classification system which everybody knows. It's easy, it's simple, and it has absolutely nothing to do with the pathophysiology of spondylolisthesis. It has nothing to do with the symptomatology of spondylolisthesis, and it has nothing to do with the surgical management of spondylolisthesis. The myridine classification was published in a gynecology journal in 1932 to discuss whether or not women could safely have childbirth if they had spondylolisthesis. So 80, 90 years later, we're using an obstetric paper to describe our patients who are treating for surgical intervention for low back disorders. The classification system was updated 40 years later. The Wiltzi classification system, most of us are pretty familiar with this one, and the nice thing about the Wiltzi is he really began to identify the pathophysiology of the spondylolisthesis. He realized that there's isthmic, there's degenerative, there's dysplastic, pathologic, secondary to bone disease, and iatrogenic, secondary to your kerosene going through the PARs at an inopportune time. And that helped clarify the issue to some extent. These are some illustrations from his original article, a type 1 on your left and a type 2 on the right over there. Marchetti and Bartolosi in 1997, so another 20 years pass. They took the Wiltzi system and they focused more on etiology and divided their cases between acquired and developmental causes and came up with these causes. And the only real modification was now high dysplastic versus low dysplastic, which is important in the pediatric world, not necessarily so important in the adult world. The advantages, it addressed the underlying pathophysiology again. It was relatively simple and reliable. It did not address severity, did not incorporate modern concepts of sagittal balance, and again, not helping to guide therapy. This is the latest classification system that I was able to find. This classification actually has been published in multiple journals at multiple times by the same group of authors, but the order of the authors changes from publication to publication. So this is the reference that I found. It may be referred to by other authors in other publications, but this was the earliest version of the one that I could find. This is Macklin on the book LaBelle. And they did a really good job of taking that Marchetti and Bartolosi classification system and breaking it down to subcategories, incorporating concepts of sagittal balance, and they came up with eight categories in total based on slip grade, degree of dysplasia, and sacral pelvic balance. And they've been able to demonstrate very good inter- and intra-observer agreement in several prospective studies that have been published. And this is their classification system, and this is actually a very elegant system. It addresses pathophysiology, it addresses severity, it addresses symptoms, and it also helps with planning your surgical intervention. You're going to do something different in someone with a low pelvic incidence, a low sagittal slope, and someone with a high pelvic incidence and a high sagittal slope. These are two different patients, both still classified as spondylolisthesis, but clearly this can help guide your surgical management. So what's the problem? The problem is that I never heard of this classification system until I put this talk together. In this room, how many of you use this classification system on a regular basis? That'd be nobody. And what's the reason for that? The reason is outside of maybe three centers in the United States, the vast majority of, I don't know how to go backwards on this. Do I go backwards that way? Nope. Shoot. Oh, yeah. The vast majority, 98% of what we do is one. Maybe another one or two percent, two. Everything else is a fraction of a percent in terms of what's actually done in the United States. So the classification system is excellent from a pathophysiology, sagittal balance standpoint, but it's pretty much useless for guiding therapy for the vast majority of what we do because virtually everything we do is in the first two categories going forward. To this date, and I mentioned this before, all of the modern registries and all of our best randomized controlled studies are still using the 1932 Myrding classification system for describing spondylolisthesis, and I think that is a reason why you get such disparate results when you look at different studies looking at the overall population of spondylolisthesis. Zoe's study was a splitter. He took a very specific population, studied that patient population. The European study was a lumper. They took all comers and treated them all that way. Both approaches have their pluses and minuses, but the more we lump, the worse our outcomes look. So what we decided to do is we decided to come up with a classification system that would incorporate the things that I, Zoe, and the surgeons at our place generally look at when we're looking at a patient with spondylolisthesis and neurogenic claudication or spondylolisthesis and radiculopathy and what we use to make our decisions in terms of fusion versus no fusion. We talked about things like the patients, whether or not the slip is mobile, whether or not the patient is mobile, whether or not there's regional kyphosis, how tall the dissipate height is, the site of compression, how much of the facet are you going to have to resect to get an effective decompression. If you're going to remove the whole facet, you're going to destabilize things. If you just need to do a midline decompression, maybe you don't. Facet pathology, and particularly what we thought the facet competence was going to be after we did the decompression. And we put all these things into a, what we call the Wisconsin Spondylolisthesis Surgical System, or the WSS system, and there's a, this is the chart that shows what we did. We went ahead and did a reliability study. We took 30 cases. We had 10 surgeons go through all the cases, separated in time by at least two weeks. We got a statistician, we got all the website, we got all the IRB stuff done, and it turns out, this is just a, you can't read this, but this is the spreadsheet that we used to do the study. Turns out that we have a lot of work to do. We disagreed with ourselves a fair amount of the time. We disagreed with each other a fair amount of time. So we're in the process of version two of this right now to try to improve these interrelated reliability scores. And the importance of this is that, you know, when we are talking about, we always talk about wanting to compare apples to apples. Well, in this case, we may be comparing apples to apples, but they're not even both fruit. If you take the ischemic spondylolisthesis, you know, with regional kyphosis, you know, and dysplasia, that's a different patient than the stable spondylolisthesis in the 88-year-old woman with disk-based collapse and simple stenosis. So we need to be very careful in terms of defining what we're doing more precisely so that we can actually develop clinically useful information so that we can offer patients the operations they deserve. Thank you very much for your attention. This is what it looked like in Madison just a few weeks ago. It's a lot nicer right now. And I'll be around later for any questions. Thank you. APPLAUSE

degenerative spondylolisthesis

spine care

lumbar fusions

patient-reported outcomes

classification systems

Wisconsin Spondylolisthesis Surgical System