← Foot & Ankle International
Foot & Ankle International
FAI December 2025 Podcast: Impact of Prior Ipsilateral Arthrodesis on Subsequent Ankle and Subtalar Fusion Outcomes: A Propensity-Matched Cohort Study
00:26:43
Loading summary
Transcript25 lines
- [00:03]
A
This is Mark Easley, and I have the distinct honor of talking with Dr. Avani Chopra, the lead author, and Dr. Michael Anardi, the senior author of December's lead article in Foot and Ankle International. Manuscript is titled Impact of Prior Ipsilateral Arthrodesis on Subsequent Ankle and Subtalar Fusion Outcomes. A Propensity match cohort study. Dr. Avani Chopra is a recent graduate of Penn State College of Medicine and is completing her general surgery intern year at Westchester Medical Center. She also serves as an International Program Coordinator for steps to walk. Dr. Michael Anardi is an Associate professor of Orthopedics at Penn State Health Milton S. Hershey Medical center in Hershey, Pennsylvania. He serves as the director for both the Foot and Ankle Fellowship and for Clinical Research. Dr. Enardi is the current chair of the American Orthopedic Foot and Ankle society's research committee. Dr. Saltzman and I selected this article to highlight a use of a large network database to compare subtalar or ankle fusion rates in patients with or without prior ipsilateral ankle or subtalar arthrodesis. Welcome to the program, Avani. And Mike Ivani, as lead author, would you please provide the listeners a brief summary of your paper and its major findings?
- [01:18]
B
Yeah. So our study was set out to determine whether a prior ipsilateral ankle or subtalar arthrodesis influences the risk of non union when a patient later undergoes a fusion at the adjacent joint. So some of the prior literature combines all the patients with previous fusions into a single group without distinguishing between those who had an initial fusion that had healed successfully and those who had fusion that failed. We think this is an important distinction that might actually influence the outcomes. So, using the Trinex Research Network, we found that a successful prior fusion does not increase the risk of non union in a subsequent adjacent fusion. However, a failed prior arthrodesis does dramatically raise the risk of non union of a subsequent adjacent fusion.
- [02:07]
A
Perfect. Thanks, Mike. Let me turn it over to you. Could you please just in simple terms, just describe your current surgical technique for an isolated ankle arthrodesis and for an isolated subtalar arthrodesis.
- [02:19]
C
Yeah, Mark. So I, over the last probably five years, have moved to much smaller incisions. And honestly, a lot of that has come from one of our basic science researchers who, especially in his diabetic mice models, has been showing the importance of the periosteum and stem cells and healing. So with all of the buzz about minimally invasive burs, I sort of jumped into joint prep with the bur over, let's say arthroscopic prep, I think that's relevant in protecting the soft tissue envelope, but for an ankle it will come. So if you're doing with a bur, make an anterior medial arthrotomy. I start with curettes, make sure I have my bearings, especially if there's trainees with us, and then I begin to denude with a curette just like Myerson would through his too small incision approach. And then I'll make a small incision laterally and do an osteotomy of the fibula. We had one of our residents a couple years ago, Chris arena, write a really nice paper showing you get much more compression at the joint when you do that. And then through small incisions. I will then place the home run screw from posterior to anterior and then usually an anterior medial screw across the tibia talar joint. I do preserve that fibula though. So then I also use 2, usually 4, 0 or 3, 5 solid screws for transfibular graft. I tend to use more lateral ankles, especially in a younger patient. Or you're trying to convince, hey, maybe down the road we can always convert this. I like to save the fibula and then for isolated subtalar fusion, same thing. Small sinus tarsi incision, begin scraping with the curette and then use the bur for preparation and cannulated screws.
- [03:52]
A
Perfect. Thanks. Yeah. We'll talk some more about technique during this podcast. So Avani, this is a collaborative effort between Hershey and Rutgers. Please tell us about the Trinetics Research Network that you used in your study. Based on two recent database studies that use the National Readmissions Database, you and your co authors state that for the current study it is okay to use the terms non union and subsequent revision to draw your conclusions when using this network OR database. These two 2025 studies that you cited as references 3 and 4 in the current study are from Rutgers and you are co author for both. Please give us a little background on the National Readmissions database for the two reference studies and the TriNetX research network used in this study. What are the advantages of using these databases in conducting orthopedic and especially foot and ankle research?
- [04:49]
B
So the National Readmissions Database or the NRD is an all payer inpatient database that captures very large national samples, but it's only over a single calendar year. The strength of the NRD is its size, which allows us to study rare complications like non union by tracking the incidence of the specific ICD10 code. For this project we use the Trinex Research Network, which however is smaller than the nrd. It does still allow us to study those rare complications. It is unique because it allows us to look at longitudinal EHR record or EHR data which is essential for this study. So the longitudinal structure allowed us to track the patients across multiple years and identify the individuals who underwent the ankle and the subtalar arthrodesis within a five year period. Something that the single year database like the NRD we wouldn't be able to do. But in general using these databases, the NRD and Trinetics, they have a large amount of de identified data which is incredibly valuable in orthopedic research, but especially in foot and ankle surgery. Many of these clinical scenarios that we study are relatively uncommon and even high volume centers may not see enough cases to have statistically significant comparisons. So by using these databases we have these large data sets that combine numbers across institutions. They overcome the rarity of certain procedures, complications and reliable study trends and risk factors that would otherwise be impossible to evaluate.
- [06:27]
A
Great, very helpful, thank you. Mike, what prompted you and your co authors to embark on this particular study? My understanding and experience suggest a lower fusion rate for adding an arthrodesis to a segment with an existing arthrodesis and or fusion. Examples include extending spinal arthrodesis to a previously fused segment and adding a hallux IP joint arthrodesis to a previous ipsilateral MTP joint fusion. Were your personal results with subtalar and ankle arthrodesis to prior ipsilateral ankle or subtalar fusion respectively better than what we had previously reported?
- [07:03]
C
That's a great question mark. And when Avani and Sheldon, Lynn and I began sort of talking about this paper, we had been collecting data from the sources of ANI and you just went through and we quickly realized we had a massive volume. But in my specific practice I always shied away from these ipsilateral fusions because I was taught and trained. And there's a lot of good papers out there, you're on a couple of them that showed these are dangerous fusions and they have a high non union rate. And Sheldon and I sort of spoke about, well, what's changed in the last 20 years in foot and ankle and you know, I think minimally invasive preparation of a joint and things like pdgf. I don't think hardware has been a substantial change. I don't think a solid stainless screw is better than a titanium one at increasing fusion rates. But some of these Other biological things sort of just led to the question. And I certainly have my fair share of non unions. I see a lot of Charcot. It was more of like, you know, is this true? When we look at a large population that has a lot of heterogeneity with many different surgeons, like what's the actual trend? And that was sort of the genesis for this idea.
- [08:14]
A
That's very helpful. Yeah, I'll come back to that too because I was curious about the timing of it, things we've learned since those papers have been published. So that's great. And I don't know, I was part of them. I don't know if they're great papers. You have a lot more volume, ours are smaller, but we'll get to that. Vani, I believe that you and your co investigators address this, but do you think it is the failure of prior arthrodesis that led to the increased non union rates for arthrodesis performed or is it the patient's comorbidities that are the root of the problem? Based on the TriNetX database or network, you and your co authors identified that patients requiring both ankle and and subtalar arthrodesis are often medically more complex and carry a higher comorbidity burden. Were you able to tease this out in the Trinetics data analysis? What is your institution's routine for surgical optimization? And do you think that surgical optimization is even more important in these patients?
- [09:12]
B
Yeah, so we did find that patients who require both an ankle and subtalar arthrodesis do carry a higher comorbidity burden, which is why we did the propensity score matching to balance these comorbidities between the groups and minimize their influence on the results. But the interesting thing is, even after matching, we still found a significantly higher nonunion rate in patients whose initial arthrodesis had failed. Which strongly suggests that it's the failure rather than the medical comorbidities alone that is driving this increased risk. So I think the question becomes, well, what do we do then for these higher risk patients? These individuals may be excellent candidates for fusion enhancing strategies. So we talked a little bit about the biologics like augment bone grafting techniques or even other measures to support the biological environment for healing. So I mean, in other words, knowing that patients are predisposed to non union because they had a prior failed nonunion allows us to be more proactive and intentional in selecting techniques and adjuncts that might give them the best possible chance of achieving a fusion. The Next time.
- [10:21]
A
Perfect. All right, well, thanks. And Mike, back to you. As you noted in your introduction, prior to your study, and we talked about this just briefly already, but prior to your study, several investigations with limited numbers of patients suggested lower fusion rates for subtalar arthrodesis performed in patients with prior ipsilateral ankle arthrodesis. Yep. And including a study for which I was co author, which we mentioned, I acknowledge that my investigation has a lower number of subtalar arthrodesis in patients with prior ipsilateral ankle fusions. And likewise, other similar previous studies also have relatively small patient cohorts. So you suggested that the previous studies did not distinguish between healed and failed arthrodesis. And while I can't speak for the other studies, I can confirm that in our study with limited patients, for the cohort in question, all ipsilateral ankle arthrodesis were solid fusions. So why do you think that the TriNetX data or network data suggests the opposite? Are the surgeons at the institutions previously studying the procedure simply not as surgically skilled as those that were in the trinetics database? I know there's more to it, but I just. I had to ask that question.
- [11:31]
C
Well, Mark, that's a very loaded question. I'll be walking on eggshells here with my answer. I think. Now, in all seriousness, I think there are some fantastic surgeons who have addressed this topic and those are memorialized in our references cited. What we meant in our discussion specifically was if you look at each of those papers, there are a few where it's not so well stated as your paper with Mark and Jans Trnka and Lou. And sometimes they don't have CAT scans. Other papers did. And so technology has changed in our understanding of biology and some of our biologics have changed substantially. And again, the way we prep joints. And I think, you know, this is 35 years from your data set. Right. I think you guys were late 80s and early 90s in the cohort and, you know, has some of the things that are seen, you know, across orthopedics now and especially utilization of PDGF and minimally invasive burs. Have they made a difference? And looking at a large population I think is one great way to study that. Now, I would also admit, though, that some of the major shortcomings and concerns that Avani and Sheldon I have with our data is that we don't have granularity in the data. And that's the big caution with this. And everyone sort of knows it is. Okay, well, which of these cases were prepared with an mis bur, you know, I might prepare my subtalar joint slightly differently than you do, you know, or certainly much different than Lou. And I think that stuff does matter. But if you have a large volume of patients and a great source of the data like these two databases and especially trinetics, it allows you at least observe these large population trends. And I think if nothing else, this paper at least will question that sort of old dogma of, well, don't be very careful because, you know, it used to be we thought it was the mechanics, right? It was too much stress on the joint and that's why we'd get a non union and maybe you'd put a couple extra screws in, you know, subtellar fusion or maybe use a hindfoot nail even to compress or something, you know, to address the mechanical stability. But I think this paper sort of, at least what I inferred from it is that there's, I think more to this is biology, especially to Avani's point where if you had a non union, even when we isolated out all of these other risk factors for non union that are well described in all the previous papers, the risk of a non union at one joint previously really increases your risk at another. And that kind of goes back to what we've been talking about, which is the biology and not disrupting a lot of the periosteum and things like that.
- [14:06]
A
Got it. Very helpful. And you're so good at your answers that some of my questions may seem they're coming from the Department of Redundancy department. So I may be hitting on some of the things, and you even use the term old dog and I'm going to bring up that I'm an old dog. But first let's give Ivani another chance at this. And along the same lines, and again, old dog question, but just want to confirm using the tri netics database or network, is this method of study reliable? So in other words, can your results be trusted? I recognize that I could never generate the cohort numbers you have, and like I said, I'm an old dog. To confirm conclusions, I would like to follow these patients myself, perform my own clinical exams and interpret my own X rays and ideally have CT scans to Mike's point, but I just want to make sure that yours is a safe message delivered to the readership with conclusions based on the TriNetX database being reproducible. Can I relax? And I know it's okay for surgeons to employ the same exact techniques for subtalar arthrodesis with the adjacent fusion segment that they would for isolated subtalar arthrodesis, much to Mike's point.
- [15:18]
B
Yeah, so we did kind of talk about the limitations of trinetics is that it does not provide that the imaging or the patient reported outcome. So we can't directly confirm the radiographic union. But in this study we do rely on the fact that when a non union diagnosis code is entered into the medical record, we're hoping that it reflects a careful clinical and radiographic assessment. And so if you think about it, a diagnosis of non union typically triggers further management, whether it's additional imaging, referral, revision, surgery. So clinicians have a strong incentive to be certain before signing that code. So in other words, the threshold for documenting non union is high, which makes it a reasonable reliable proxy at a population level for studies of this scale.
- [16:07]
A
Perfect. Thank you. Yeah, Mike, I'm going to bring up my old dog one more time and I think you've answered this pretty well, but I have one more specific question in there. Their authors, including myself, reported on a concerning fusion rate for the subtalar arthrodesis in patients with prior ipsilateral ankle arthrodesis. So surgeons are included in the TriNetX database. That may have been before some of this information was published. And one of the studies that I was involved in was 2015, so about 10 years ago. But your data goes back 10 years, even before that. So is there something that the TriNetX database can tease out to suggest that there was a group before we were made aware of this concern and then afterwards. So can the TriNetX database identify that level of detail? Do you think that there could be a difference in the TriNetX database for these procedures prior to and then after the previous studies raised the awareness of a provincial increase in non union rates in patients with prior ipsilater arthrodesis? You gave us your technique earlier on how you do these procedures. Could you come back to and just give a little more detail of if you have a complicated situation, do you add anything else like orthobiologics or. Even though you're doing minimally invasive techniques, are there times that you would maybe add a little bit more robust fixation?
- [17:27]
C
Yeah, those are all excellent questions, Mark. And I think, you know, the observer results from this large population may ultimately just be a reflection of all of us as surgeons approaching these patients with a heightened sense of carefulness, knowing that the potential for non union is increased. I mean, even if you look at the data in Our paper, there still is a non union level that is present. It's just not as high as those previously reported and that may reflect some of the later years in our study. And as you pointed out, your 2015 paper does have overlap. And you know, PDGF was not around at the beginning of our study. So we can't just sit here and say, oh, it must be pdgf. I bet you it's a combination of all those things. And unfortunately one of the shortcomings of this paper is we can't point to exactly what with the level of data we have and say what is making the non union rate with a successful ipsilateral fusion lower than previously expected. Because I do believe all of the mechanics that have been stated in previous series are a very rational explanation for why this doesn't heal. But I think if you were to then go back and say, well, why is the rate of fusion a little bit higher when you do have a successful fusion above? Or the converse why is the non union rate so much higher? Why is that such a powerful predictor of failure of an ipsilateral joint fusion? I think going back to biology really in my mind is the explanation for this. It's just we don't have enough observational study essentially retrospective in nature, so we don't have control over well, how did I prepare my joints versus you and which biologics were added and when, and we're lacking that data. So I think that's a major shortcoming that all the readers should be aware of. But as studies move forward and if we ever had a national database, that information from OP notes would get uploaded into it and then we could maybe even revisit this study again with more detail to try to tease those things out. We did briefly look at the use of PDGF because that is tracked, you can find that data in trinetics. We didn't find much of a difference and we had trouble in the course of the whole study. And then there was half of the study that didn't have it until the product was released. And so the use of PDGF does not explain in isolation the results. That was the first question. The second was going back to if I have a patient who let's say had a nonunion previously and then eventually got that healed and now they come back with adjacent segment arthritis, the subtalar joint. I am thinking very carefully about, number one, my construct. And so, you know, there are hindfoot nails on the market that have a compressive mechanism within them and sometimes I'm using multiple screws across the joint. I do like the use of PDGF in some instances. I do carefully protect these patients afterwards. We have a fantastic total contact cast maker. And I'll consider a bivalve total contact cast to improve mechanics and then even make them be in a brace and not just abruptly wean them from a cast or immobilization into regular shoe wear. I'm conscientious about rockers in the bottom of their shoes to mechanically offload the fusion. And I think that if we're correct in our guests that it is more biology not stripping the periosteum. And I think a lot of people will say, well, how do you know how much you've prepped? Have you prepped correctly with the Bureau? When you use minimally invasive bur and you evacuate, it's low, low number of seconds with the bur on to avoid any thermal necrosis. So it's less than eight seconds. And then copious irrigation throughout. And then I will evacuate the first round of the bur. I use fluoroscopy and I look at the subchondral plate, and I just want to go through the subchondral plate. And then after I've burred, I'll insert a small curette and actually feel ensure that I've gotten through the subchondral plate in several areas. And I use several different types of burs. And then after you've gotten to the subcontral plate, the slurry you create will have a different color to it, and it'll be almost like a red toothpaste. And I actually harvest that and then use either tumi syringes or our spine partners have this delivery device where you can actually deliver the bone graft mixed with PDGF back into the joint. So I'm not wasting any graft, but that's just me. I mean, this database is thousands of patients which represent hundreds of surgeons. So you know that we can't extrapolate some of my techniques to the general mass, which is, I think I would love additional data. I think it would really make this an even better paper. But I think take home point, everything Avani has said is absolutely true. I think we can trust these results. We just can't pinpoint the exact reason why we demonstrated a difference from some of the prior work. These are our best guesses.
- [22:09]
A
Perfect. That's a great answer and a lot of good personal experience I'm sure would be valuable to the listeners. Avani, one more for you. Just for the listeners who might not be as familiar with propensity match scoring. Would you just briefly describe this comparative analysis method?
- [22:24]
B
In this study, we found that patients that had both the subtalar and ankle arthrodesis had more comorbidities than those with just an isolated arthrodesis. And so the way around that disparity was to do propetzi score matching. So this is a statistical technique that we use in retrospective studies that reduce the impact of confounding variables to ensure balanced comparisons between the cohorts. So it calculates the propensity score for each patient, which is the probability that they belong to one group or the other based on factors like age, sex, bmi, smoking, sass, diabetes. And then we match patients from each group that have very similar scores so that essentially we have two completely balanced cohorts. And then this allows us to compare the outcomes, like non union rates in groups that are similar.
- [23:18]
A
Perfect. Thank you very much for doing that. And Mike, a little tangent question for you. So I trust, even being a Hopkins grad, right, I trust that you have developed a Penn State loyalty over the nearly 10 years that you've been there. So this is a multi center study between Rutgers and Penn State. And Avani, you've rotated at Rutgers and both of you have co authors here, including Sheldon Lynn, who have loyalty to Rutgers. So no pressure, Mike, but the listeners will know the answer by the time this podcast is released. With the Nittany Lions football team having shown more promise after a relatively or quite dramatic slow start to the season. Will the Nittany Lions beat the Rutgers Scarlet Knights this weekend and become bowl eligible?
- [24:03]
C
Mark, my prediction is that they will be victorious. I think as Terry Smith said in his post game press conference after the Nebraska game. I have the locker room was the direct quote. And he's a fantastic individual and they've overcome a lot of adversity this season, as you mentioned, it's definitely been a trying season, but I will make a bold prediction that Penn State will be victorious. And the final score, 35 Minnie Lions 14 Scarlet Knights.
- [24:36]
A
Wow. Okay listeners, you hold them to that. Okay, so we'll know the answer by the time the podcast is released. But you heard it here, right from the horse's mouth. Okay, good. Avani, I've got one for you too. You have now spent time in Harrisburg, so I need to know, is Hershey's chocolate really the best in the world?
- [24:56]
B
Absolutely. There's no question about it. The highlight of Hershey is the chocolate.
- [25:01]
A
All right, well, good. Well, I can't argue with it it's pretty darn good. Mike, as our podcast comes to its conclusion, is there anything else you would like to share with the listeners?
- [25:11]
C
I think we covered a lot of ground here and really beat this paper up. I would say that this study and those using these large databases are a great place to start and I think large population data is fantastic, especially in any surgical field. And when you have an interesting result like this paper did, go back and really scrutinize the data like we had to make sure that you're presenting the but it really is the truth and I think Avani and the rest of the team did a fabulous job with that and Sheldon and I fine tooth combed it as well before we submitted this abstract to ensure because it was this is a little bit of a paradigm shift from our previous understanding of this condition. So I think these databases can be very helpful as a place to start and I would argue we need additional studies that are controlled isolating some of the variables we talked about and I think take home point here is that a patient with a non union and you're doing ipsilateral fusion still requires a lot of care and a lot of attention and meticulous surgery and aftercare to get a good result.
- [26:15]
A
So great concluding statement. So I would like to thank Dr. Chopra and Arti for sharing insights on their foot and ankle international publication impact of prior ipsilateral arthrodesis on subsequent ankle and subtalar fusion outcomes A propensity match cohort study and I would like to thank everyone for participating in this month's FAI podcast. This is Mark Easley. I look forward to next month's podcast.