Telephonic and Electronic Messages in Total Joint Arthroplasty

[00:00:02] Speaker A: Welcome to your Cases on Hold, the JVGS podcast hosted by Andrew Schoenfeld and Aisha Adkeen. [00:00:08] Speaker B: Here we discuss the best of what each issue of JBJS has to offer with the usual dose of entertainment and pop culture. [00:00:16] Speaker A: Take us with you in the gym, on the commute, and as ever, whenever your case is on hold. [00:00:26] Speaker B: Welcome back to another episode of youf Case Is on hold, episode number 114. My name is Aisha Abdeen. I'm Chief of the Division of Hip and Knee Arthroplasty at Boston Medical center and Associate professor of Orthopedics at Boston University, and I have with me my co host. [00:00:42] Speaker A: I'm Andrew Schoenfeld, professor of Orthopedic Surgery and Vice Chair of Education at Harvard Medical School. [00:00:49] Speaker B: As a reminder, the opinions expressed here are exclusively our own and do not represent those of JVGs, the editorial board, the Board of Directors, nor the affiliate journals of jbjs. This episode of youf Cases on Hold is brought to you by JBGS Clinical Classroom. If you are listening to this on the day we drop it's September 15th for the September 16th episode. Kids are back to school so we should all likewise take advantage of the fall academic reset and revamp your orthopedic learning with the JVGS State of the Art Orthopedic Learning System. No matter what stage you are in, your orthopedic training or career, log on and amp up your orthopedic knowledge at the JBGS Clinical Classroom. So we are going to get into this issue. At the top of the pile we have what's New in Hip Surgery by Ektiari this is permanently free. The Functional Vital Sign Interpreting Physical Function to Guide Treatment by Baumhauer this is a highlight. The Unexamined five Philosophical Concepts in Orthopedic Encounter by Parisian Thoughts at Hand by Ron Translateur this is permanently free. Compliance with the CONSORT AI Extension in Orthopedic Randomized controlled trials by Stressor Non Traumatic Osteonecrosis of the Femoral head an update by Mont Orthopedic Management during the first 24 hours of the October 7th MASH Casualty Event in Israel. The Experience of a Frontline Trauma center by Ryzinski Diluted Poviodine iodine irrigation for prevention of implant related infection A comparative analysis of concentration and frequency in a rat model by ima this is a highlight. There is a commentary and it is 30 days free, so we will get right into our headlines. Andrew will be presenting minimally invasive sacroiliac Joint funct fusion compared with sham operation the SIFSO trial. A concise follow up at two years of a previous report by Kibbsgaard. This is the lead article. There is an infographic and it is 30 days free. [00:02:52] Speaker C: Yes. [00:02:53] Speaker A: So I don't know how it's supposed to be pronounced, but I was going with CF so trial they say it's a concise follow up at two years. It's not that concise. It's almost the length of a regular manuscript really at the end of the day. And recently I submitted a paper to a different journal where I got a comment back from editorial board member who said that there needed to be less throat clearing in the manuscript. Wow. And that came to mind with this work as well as I was reading it. So the premise here that I was not familiar with the CFSO trial and that is because it was published in the E Clinical medicine journal in 2024 which seems unusual to me that it would publish there and then this concise follow up publishes in JBJs. [00:03:52] Speaker B: But interesting. [00:03:54] Speaker A: Yeah. So the SI joint fusions are becoming increasingly common. I wouldn't say they're especially common in the neck of the woods where we practice, but in other parts of the country. I do know they're used quite frequently and with an increasing frequency. They start with there's only three randomized trials. Two of these compared SI joint fusion with non surgical management that demonstrated superior results for SI joint fusion. Both trials transitioned into what they call observational studies beyond the six month time point due to considerable crossover which we see in a lot of versus non op studies that plagued the strict intent to treat interpretation of the sports studies. For example, this study, which I'm not sure that you could do this in the us I don't think an IRB would approve it, but they were able to do it in their location compared SI joint fusion to sham surgery. And in their initial study which only looked at six month outcomes, which also I thought was a little bit unusual, typically you include six months on the way to at least one year. Two years is kind of the particularly enjoy joint replacement. Two years is kind of the expected. Like you're not going to get something published who doesn't have two year data. Am I right? [00:05:18] Speaker B: Yeah, absolutely. [00:05:19] Speaker A: Yeah. So this is a follow on to the six month results and at six months there was no real difference between the SI joint fusion and the sham surgery. Now what may be important about the six month time point in this unique context is that is where they revealed to the patients what surgery they had sham or and they had promised the patients if you had the sham, then you could go on to have SI joint fusion if you wanted to. But we'll get into why that kind of complicates things even further down the road. So the fusion device that they used was the sibone eye fuse triangular titanium implant. The patients and examiners were blinded to group allocation until the six month reveal. That also is especially important. So what was the sham surgery? It was really a skin incision and a blunt pin was inserted through the muscle to the ileum and then removed. And that was it. So they initially were sampling, determined a sample size looking for a difference in set at two points in the NRS score with standard deviation of 2.5. Standard type one error precautions two sided test aiming for 80% power with a dropout rate of 20%. 30 patients were planned totaling 60 patients. And then they analyze the data according to intent to treat, per protocol and as treated principles. That doesn't really apply to the first six months of course, because there was no way that the people could violate their assignment. They didn't know what they were assigned to. It does apply after the fact, but is sort of really obviated by the fact that just about everyone, as you might expect, I mean people are coming here, they want to be considered for the surgery. They didn't get the surgery. They were told they could have it later. So of course they want the surgery, right? Like they're like, all right, I waited the six additional months, let me have the surgery now. So they had 31 patients in the sham group and after those 31 patients were told you didn't actually have the surgery, 84% or 26 of them went on to have the surgery. So all told, there were only five patients that continued in the sham group. They never got the surgery. Six of the 32 patients in the initial SI fusion group and six of the 26 in the crossover group underwent SI fusion on the contralateral side in the time window of observation at two year follow up. The mean reduction in the NRS score for pain in the operated SI joint compared with baseline was 3 points with a confidence interval of 2 to 4.1 in the initial SI fusion group. The five patients who remained in the sham only group had a mean score of 3.8 with a range of 0 to 8 at 6 months and 5.6 at 2 years. So there was certainly deterioration on that front. 26 patients in the crossover group had a mean score of 6.4 right before unblinding, which then increased to 7.4 right after they learned that they didn't have the surgery and then went on to wanting the surgery. At two year follow up, the crossover group had a mean reduction of 3.5 points. For secondary outcomes. The SI fusion group interestingly did worse than the crossover group. The analyses using intent to treat per protocol and as treated definitions are just an exercise in futility may not exactly be the right word, but that's how I would say it. They do report them, as they say, for transparency. But when you're talking about all but five patients went on to have like, there's no point in even going through that process or considering it as having some kind of informative outlook. So they say crossover makes the long term results challenging to interpret concerning the actual effects of the SI joint fusion. But when you put this all together, this is not really a ringing endorsement of SI joint fusion. That's where it is. At baseline at 2 years, 45% of the participants reported feeling much better. So less than half and 25% indicated they were somewhat better. 71% of the participants noted a decrease of just 2 points in pain. So after a fusion based surgery we're getting two points and 41% again, less than half achieved improvement equal to or better than than the predefined mcid. These are not astounding. Knocked it out of the park metrics for sure. Also, there was no decrease in the consumption of pain medication even among those who said that they experienced clinical improvement. [00:10:19] Speaker C: So I'm just left kind of scratching my head as to like, what is [00:10:21] Speaker A: the, like what is the population that we're really dealing with here. [00:10:26] Speaker C: What is really interesting, they call this [00:10:28] Speaker A: moderate rates of surgical success. [00:10:31] Speaker C: I'd say that's a very optimistic set of glasses that they're looking through to call this moderate success. It is apparent that selecting candidates for SI joint fusion remains challenging and the placebo effect is considerable in this context. When surgery is performed, they say that the high frequency of crossover is a challenge. The placebo component was removed after the [00:10:56] Speaker A: primary endpoint of six months. [00:10:58] Speaker C: Patients became aware, obviously they wanted and expected to have surgery, so they're moving into that group. And the placebo effect really only persists, as they showed in their data. While the patients are under the impression that they had the actual surgery. When unblinding and crossover have altered group composition, the use of standard statistical models such as intention to treat really has no informative outlook. And that is true. Then their conclusions Surgically treated patients experience significant pain reduction and and improved Physical function compared with baseline. They completely gloss over the fact that less than half were meeting the mcid. Nobody changed their prescription opioid use data. And this is from like Norway and Sweden or. Well, I don't know, I inferred opioids, but that, that may not be proper. Anyway, they said that there was no decrease in the consumption of pain medication, whatever they were, so it may not be opioids. Again, this being Norway and Sweden, this is not a ringing endorsement that this is the idealized surgery that you're looking for. I think in the right patient, in the right hands, with the right indications, there are really good outcomes. I don't think it's the kind of procedure that's going to lend itself to, given the complexity of both SI joint and chronic low back pain, that this is a catch all diagnosis for patients with a lot of phenotypic differences. And that's why you see things don't work out quite so well, as you know, in these randomized controlled trials, so to speak, because the efficacy of the surgery is going to be distilled through the prism of those baseline patient characteristics. And even when there is SI joint pain and they meet all the criteria, have failed all the non operative measures, it's not necessarily the only pain generator. [00:12:54] Speaker A: And then there's also central sensitization and the other comorbidities and all these other factors that often are comorbid in patients with chronic SI joint pain and that would diminish the efficacy of the surgery. It doesn't mean the surgery isn't useful in the right hands and it doesn't mean that it doesn't have applications. But I think a lot of people who are at the forefront of using these have a lot of experience, they're very good at selecting patients, so they have excellent outcomes. And again, the randomized controlled trial was the sine qua non of excellence in research. It doesn't always work in every surgical context. And the fact that you don't see significant differences doesn't mean that there aren't cases where it is appropriate and it will be successful. I think what it is showing is that it cannot be expected to have the same level of efficacy in the general population. [00:13:54] Speaker B: Yes, very good points. I have not much to add to that, but it just still is astounding that this past muster for their ethical board to have these patients have a sham operation and then have the majority of those go on for a second operation. You know what I mean? Not to mention you mentioned central sensitization and the pain that can happen from having two operations, and it's astounding that the patients consented to that. But interesting. [00:14:20] Speaker A: Yeah, a very interesting study. One you're not likely to see again, I would wager. Or with any kind of frequency. [00:14:30] Speaker B: We'll move on to the next paper I'll be discussing Enhancing Osteosarcoma Survival Predictions A comparative study of a multicomponent model machine learning approach integrating SE R and NCDB data sets versus Conventional single data set modeling by GALOA and there is a commentary. This study was performed at the Massachusetts General Hospital by my friend and colleague Dr. Santiago Lozano Calderon as a senior author. The study seeks to address the shortcomings of machine learning for prognostication and osteogenic sarcoma. Osteosarcoma, as many know, is the most common primary bone tumor in adults, adolescents and children. There's a bimodal age distribution. It has a five year survival rate of 60 to 70% and only 15 to 30% in metastatic cases. There is a need for accurate prognostic assessments to guide treatment. The authors note that current single data set models are population specific patterns rather than generalizable disease characteristics, which limits the clinical applications of these machine learning models. To address this shortcoming, the authors developed a multi component model ML framework using domain adversarial training across two national registries, integrating structured variables with text based patient data to learn generalizable disease patterns and achieve reliable survival prediction across diverse populations. So I had to look up domain adversarial training of neural networks. This was developed by Gannon and Associates and it was published first in the Journal of Machine Learning Research in 2016. So typically in machine learning models are trained on one data set which is considered the source domain, with the intention to use them on slightly different data sets which is the target domain. However, the model's performance can degrade if there are differences between these domains and this problem is called a quote unquote domain shift. And so domain adaptation in machine learning is the process of adapting a model trained on a source domain to perform well on a target domain. The study is a retrospective study using the SEER database which is the surveillance, epidemiology and end results. In that database they had an N of 4,278 patients from 2004 to 2015 and the NCDB, which is the national cancer database with an N of 4,049 patients from 2004 to 2018. The SEER and the NCDB are the two main databases used for cancer research in the United States, differing primarily by whether they are population based or hospital based. SIRA is run by the National Cancer Institute and covers entire geographic regions while the NCDB is run by the American College of Surgeons and the American Cancer Society, capturing cases mostly from Commission on Cancer accredited hospitals. So in this study, cross data set performance of a single model versus multi models were compared to primary outcome measures were performance metrics for two year and five year overall survival predictions. They used the SEER database to identify patients with osteosarcoma diagnosed using primary site codes. They applied the following inclusion criteria to the initial cohort of 15,241 patients histologically confirmed osteosarcoma diagnosis, tumor location in a long bone, short bone or trunk and osteosarcoma as the primary malignancy. These criteria reduced the cohort to 7,252 patients. They subsequently excluded patients lacking complete prognostic and treatment data, resulting in a final cohort of 4278 patients. Extracted variables included demographic characteristics, tumor characteristics, the TMN classification, metastatic patterns, the presence and anatomic sites of these metastatic patterns and treatment modality, surgical procedures, radiation therapy sequences and chemotherapy administration along with survival duration. From the NCDB data they included patients with bone and joint malignancies diagnosed between 2004 and 2018. A search for the International Classification of Diseases for Oncology histology codes were included and they identified 11,643 patients diagnosed with histologically confirmed primary osteosarcoma after further excluding patients with tumors outside the appendicular skeleton and pelvis in patients with insufficient prognostic and treatment information. The final Data set included 4,049 patient variables extracted from the NCDB or identical to those extracted from the seer. The median patient age was 17 years. The sex distribution was similar between data sets, 56% male in the seer versus 55.5% male in the NCDB and significant differences existed in the racial distribution with the SEER having higher Hispanic representation and NCDB having more white patients 76% versus 55%. The median tumor size was slightly larger in the SEER database 9.5 cm versus 9 long bones of the lower limb represented the most common tumor location in both datasets, 71% in the SEER versus 69% in the NCDB. The multimodal domain adversarial approach generally improved performance across each validation scenario compared with the single model approach for two year survival prediction for the NCDB trained models validated on the SEER data. The multimodal approach rather achieved an AUC of 0.843 compared with 0.665 for the single model approach representing an improvement of 0.178. For the SEER trained models validated on the NCDB data, the multimodal approach achieved an AUC of 0.708 compared with 0.606 for the single model approach representing an improvement of 0.102. Receiver operating characteristic curves illustrating the performance differences between single model and multimodal approaches across all dataset validation scenarios were depicted and the curves demonstrated consistent separation between multimodal and single model performance both for two year and five year survival predictions, with the multimodal approach showing superior discrimination across all threshold values for the five year survival prediction. For the NCDB trained models validated on SEER data, the multimodal approach achieved an AUC of 0.798 compared with 0.599 for the single model approach representing an improvement of 0.199 and for the SEER trained models validated on the NCDB data, the multimodal model approach achieved an AUC of 0.648 compared with 0.563 for the single model compared which represented an improvement of 0.085. So the main limitations of the study are the limitations of the national registries themselves, so the lack thereof of certain data, specifically histological data, surgical margins, adjunct of chemotherapy and radiation regimens, as well as geographic and temporal variability in treatments. The study was retrospective and these findings need to be validated with prospective analyses. It's a very high level study. I would have liked a little more description on the machine learning technique they use. I had to search for the original Gannon paper on the domain adversarial training of neural networks and I got a little lost in the weeds. So a paragraph or two that I would say sort of dumb it down for those of us mere mortals that are not experts in machine learning and neural networks, it would have been appreciated. But I think overall, you know, obviously it's a very meaningful contribution to the literature for me. Obviously this case is not on hold. But Andrew, I'm looking forward to your thoughts, you know, as our sort of methodological guru. And I know you know a lot about machine learning and its applications. [00:22:24] Speaker A: Yeah. So there's nothing really to put on hold with respect to what they did and how they went about it and ultimately what they found. I mean those are, those are the realities that the questions that I have for this one. I think there's a little bit of a kind of a shell game going on here. Because the reality is, is that with any kind of discriminatory model, you're always going to get more discrimination if you have more data points or more data. So the fact that the multimodal one does better than the single one is that's almost always going to be the case. The only situation in which that would not be the case is if the additional data points that you were putting in had really nothing to do with the outcomes. You know, like, like you have a robust data set on prediction of survival and then you're putting in like what the person's horoscope is or what sign are you, or, you know, do you like, who's your favorite designer? What shoes are you wearing currently? Like, you know, just things that have absolutely nothing to do with the outcomes. That's the only way you're not going to get an increase. So, you know, it's a little bit of a, hey, look what we did. And it's sort of like, yeah, but that goes without saying. Like, that's the expectation. The second thing is that these data are really old. Like the data that ended in 2015 and started in 2004, and the NCDB data started 2004 and ended in 2018. I just think about other aspects of medicine and would you treat somebody with coronary artery disease based on data from 2004? Would you treat somebody with diabetes? It's like, what about the GLP1s? GLP1s? We're gonna use this 2004 data to take care of you. You know, the, this field is advancing very rapidly. In the last 10 years, there have been major advancements and much of that is not going to be represented in the data that they're considering now. Granted, it is probably the best they can do with the data sets that they have available. I think that some of these have stopped collecting data in such a way that you can kind of compile it in aggregate. And it's a rare condition, obviously, so you're not going to get a lot of patients in any one year. It's not joint replacement, it's not elective spine surgery. The other thing is that because this is de identified data and osteosarcoma is a very rare condition, obviously it's not completely overlapping populations, but we cannot completely eliminate the fact that there are going to be some patients in SER and some in NCDB who are the same individual, and that is just going to artificially like enhance the performance because that person is essentially represented twice with the same outcomes. And then when you're merging them together, it's one person has been duplicated, so they can't tell you how often that was or how rare it was, or they can't quantify that number. It's probable that there are some Is it a lot? [00:25:45] Speaker C: Is it a little bit? [00:25:46] Speaker A: Nobody knows. So, I mean, those are just the kind of thoughts that came to mind when I was looking through this. I don't think that there's a reason to put anything on hold. Again, in terms of what they did, I think they're being a little bit coy and not sort of saying yes, it's not organically increasing the AUC performance. Like that's naturally going to happen because of your enhanced data and larger data elements. [00:26:16] Speaker C: Right. [00:26:17] Speaker B: So they proved the expected trend. [00:26:20] Speaker A: Yeah. I'm really excited about your thoughts on this next one. [00:26:26] Speaker B: I know. Likewise, actually. So let's without any further ado, we'll get to our your cases on hold featurette. This is entitled Surgeons May Be quote touched out the ever increasing workload of telephone calls and electronic messages in total joint arthroplasty. This is by Lam and colleagues. There is a commentary, a visual summary, and it is 30 days free. This study comes to us from the group at Jefferson, the Rothman Institute. They address the issue of increased communication demands of arthroplasty care teams. The increased volume of arthroplasty of the total hip and knee orthopedic practices are assuming a greater responsibility in managing patient coordination and care due to cost containment and have shifted away from inpatient, hospitalization and rehabilitation fees. The authors define these communications as touch points or encounters made through telephone calls and electronic messages. We've all increased a surge in these messaging in all practices. Though contemporary literature highlighted the additional workload of telephone encounters during the perioperative period, electronic touch points also contribute to a much larger workload for orthopedic practices. So I think this is a study that's very timely and their purpose was to analyze and quantify the total number of preoperative and postoperative touch points that patients utilize over time after primary total hip and knee arthroplasty in a very high volume arthroplasty center. This was a retrospective study of patients undergoing primary total joint arthroplasty of the hip and knee from January 1, 2016 and January 31, 2022. Touchpoints related to the procedure were captured for all patients during 30 days preoperatively and 90 days postoperatively. Communication touch points included telephone encounters and online electronic messages sent to patients by the patient staff, received from patients or received from another party on behalf of the patients regarding their care. Electronic messages included encounters tracked through the online portal which patients could use to directly interact with the surgical team. Administrative staff, physician extenders and nurse navigators had processed and responded to most of the touch points and and patient demographic and comorbidity variables, insurance type and procedure setting, inpatient versus outpatient were recorded. Inpatient facilities included multiple regional hospitals affiliated with the Rothman Institute and the outpatient facilities consistent of ambulatory surgical centers or specialty hospitals. So patients scheduled for primary total arthroplasty or total nery arthroplasty received surgical information through methods that slightly varied according to surgeon preferences and their respective teams workflows. All patients met with a surgery scheduler and received information packets containing relevant appointment details, preoperative instructions, and discharge protocols related to the surgery and this included telephone numbers for patients to contact regarding questions about billing, scheduling, perioperative concerns and some surgeons called the patient the day before surgery or within 24 hours post op to address concerns. So we don't have a breakdown of which of how many surgeons did that and those outcomes. Patients were also informed about the institutional online portal which could be accessed 24,7 to communicate with the respective orthopedic team by electronic messaging and the primary outcome measure was the mean touchpoint utilization. This was calculated per patient on the basis of total telephone calls made and electronic messages exchanged and the number of patients who underwent total aparthoplasty or total nearthroplasty each year. There were 45,216 patients included in the study, 42% of whom or 42.5% of whom underwent total hip arthroplasty and 57.5 underwent total knee arthroplasty. 84.8% of cases occurred in the inpatient setting and 63.1% of patients had commercial insurance. Mean age of cohorts were 66.2 years ranging from 20 to 98 years and the mean body mass index was 30.4. Age, BMI, ASA classification, Charleston comorbidity index were significantly higher in patients who underwent total knee arthroplasty and than the patients who underwent total hip arthroplasty, so the initial query yielded 302,949 touch points within the perioperative period. They excluded 8,254 touch points from self pay or non commercial and non Medicare patients with 89858 touch points and 205 63. Seven post operative touchpoints were included in the final analysis and this consisted of 277,729 telephone calls and 16,966 electronic messages. That's huge. So from 2016 to 2022, electronic messages increased by 31% in patients who underwent total knee arthroplasty and by 66% of patients who underwent total hip arthroplasty. For post operative touchpoints, patients who underwent total hip arthroplasty, the mean number of telephone calls was 3.2 in 2016 and 10.2 in 2022, with a positive increase of 220%. For patients who wondered about total knee arthroplasty, the mean number of telephone calls was 3.8 in 2016 and 11.1 in 2022, which is 192% increase. Electronic messages doubled with an increase from 1.2 to 2.5, which is a 108% increase for total knee arthroplasty and from 1.0 to 2.1, which is 110% increase for those who underwent total hyperthroplasty. So when trending the preoperative and postoperative touchpoints per patient over time, the mean total increased from 8.7 in 2016 to 17.8 in 2022, which is 105% increase for those who underwent total knee arthroplasty and a 121% increase from those who underwent total hip arthroplasty. So you get the point that over time there has been a considerable uptick in messaging from the preoperative to the post operative period. It's a retrospective review and it documents a very high volume of arthroplasty practice and I think it reflects what we're experiencing across the board in academic practices. I think also in community practices it speaks to the increase in volume of total joint arthroplasty. But also even though they had a majority of patients having inpatient surgery, as we increase the outpatient surgery, there's more and more questions that arise in the early post operative period and we just need additional staff to be able to manage these questions. They mentioned that the majority of these interactions were performed by physician extenders such as apps and nurse navigators. But it just speaks to the huge volume of resources that we need to now attribute in arthroplasty to address these calls. I think it would be very interesting to see if practice variations such as the surgeon calling the patient the night before if there's any variation in how they manage the educational components. It sounded as though there was pretty consistent education of patients in terms of the materials, but was all that material consumed? Right? So what percentage of the patients are actually really absorbing the educational material? It would be interesting to determine whether patients sort of disparities in social determinants of health play any role, such as language barriers or health literacy, and they didn't delve into that. But in my world that's pretty important in a health safety net hospital. And that plays a big role also just in terms of surgeon variation and variation in practice techniques, whether that plays a role. So sort of a multivariate analysis would have been helpful in that respect and or at least knowing what the variations were per practice. But I think it's very, very eye opening. It reflects what a lot of us are experiencing and put numbers sort of to what our experience has been. I think it speaks to how we need to sort of think about resource allocation in arthroplasty. I think it makes many of us lament the fact that reimbursements are dec, whereas their resource demand is increasing, and how we're going to deal with that conundrum moving forward. Andrew, what are your thoughts? [00:34:43] Speaker A: Yeah, I thought it was a very interesting study. I thought, you know, quantifying this is certainly very, you know, important for folks to understand. I think that there's a little bit of tone deaf here because what we're suffering in orthopedics, and I don't. Suffering isn't the right word. What we're seeing in orthopedics is dwarfed by what they're seeing in primary care. The primary care doctors are just inundated, as I'm sure you can imagine, with primarily the difference here is the ease and the ubiquity of electronic access through the medical record. And that's why, you know, we're seeing this sea change in 2015, 2016. That's where everybody was transitioning to primarily epic, but some type of electronic health record and that, you know, if we go back to ancient times when, when I was in residency, you know, part of our job, when you were on call, you were holding the pager for like the, the practice, you know, essentially overnight stuff, patient request calls that was coming to that pager. It happened. This was, you know, you were covering a practice of like 30 surgeons. It didn't happen all that often when there was the barrier of you knew that your phone call was going to disrupt somebody, whether it was you were calling your doctor in the middle of the night, or you were calling someone in the middle of the night to say, I'm having this problem. There was some forethought into does that problem merit me calling? Or even if they were going to call in the daytime. If you're contacting the practice, it tended to be, the wound is draining. I'm having more pain. I can't feel my toes. Like something that was really like, yes, we need to address this. By and large, you didn't get the messages that sometimes are even to like, I have a friend who wants to see you. How can I get them in? Which, which is a delightful and a very easy thing to, to address. But also you're supposed to like, remind them like, you know, this is only supposed to be for like patient health care communications. It's not, it's not email, essentially. And that's how people relate to it. So the, the two real guardrails that were removed, as I see it from a historical perspective in this context, was the first is that the institutions created this expectation amongst patients that this is, you can communicate with your provider anytime. You can just send them whatever you need and like it. You know, there was the expectation that there was no question or no comment that was not worth sending to your provider. And, and then also that there was the expectation that you will hear back and you will hear back pretty soon. And at least at our institution, there is an emphasis on, you got to answer these in 48 hours. [00:37:55] Speaker B: Yes. [00:37:56] Speaker A: You can't just leave it building up. You got to answer them. And then that goes into patient satisfaction scores. And just the business of medicine is such that it is a. The power is in the purse of the consumer, so to speak. But as we've discussed on many, many an occasion, it's not that individual who's really paying the lion's share of the. [00:38:24] Speaker B: Are we going into the tuition? [00:38:26] Speaker A: No, no, we're not touching. But it is, that is, you know, if it's a cash only practice, right, then you know for sure you're going to be answering these right. When you're like, pay me $60,000 to do the total hip. And they're like, here you go. You know, if they want something, it's like, absolutely. Oh, yeah, you know, what do you need? The part of the moral injury here is that people feel, I'm not getting paid enough, quote, unquote, to do these procedures, or I'm getting paid less than I used to, or the payment isn't covering enough in terms of everything that goes on with the practice. And there are all these additional demands which I fully get and I'm empathetic to. But that's really where these things come from in the last several paragraphs of the discussion is what they're talking about, which is there's a limit to the number of touch points that a surgical team can reasonably manage without systemic changes and appropriate valuation of this workload. So let's just go to the primary care doctors have the same problem who's talking about valuating their workload in terms like. Right. Like it's like we need this to build. This has to be factored into our lump sum payment for the RVU for doing the total hip. Well, what about the primary care doctor? Like they're not getting a lump sum payment, they're only getting the payment when they do the CPT for the office visit. Right. [00:39:53] Speaker B: That's true. And I think we're very good at advocacy in orthopedics. Right. But it's not a zero sum game. I think the PCP should do that. Of all people, they should be doing that. [00:40:02] Speaker A: Yeah. I mean I think what it really comes down to from a pragmatic standpoint, they're very helpful here and they have some CPT codes that they say they're non reimbursable. But you could use these to, to quantify the check 9800-898-016 to show check in services to highlight touchpoint volume. I don't know what that's ultimately going to achieve. What I, the reality I think is, you know, they talked about education and things like that. I'm not sure it's education on when is it appropriate to use the gateway, when is it appropriate to use the messaging service. You know, these are the things that we will prioritize in terms of answering. If you're sending a message about you read something in Reader's Digest that you want our opinion on. You know, I googled this and you put in, you know, a such and such millimeter size construct and I read that it should be at this construct or you get the ones where it's like the radiologist said you said that I had a disc herniation, but the radiologist said it was a disc protrusion. So what's going on? We're trying to pull here. Yeah, I mean it is just a reality of. And I, I don't, I don't think it's going to change, of course. And I don't think that this article is going to lead to policy relevant changes in terms of reimbursement around this because ultimately the payers have no vested interest in. You know, essentially their answer would be adjust how you manage your like it. It's really the hospitals and the institutions and maybe the practices that are putting the urgency around these have to be answered. These. You know, you don't answer every email that you get necessarily. Maybe you don't need to answer every gateway message, but they're saying no, you do need to answer every gateway message. So then you have to modulate that with the person who's sending the messages. [00:42:05] Speaker B: Yeah, agree. Very interesting. Okay, so let's get into our honorable mentions. We have medial collateral ligament injury and posterior cruciate ligament tibial avulsion fractures, an unrecognized finding by Wang There is a commentary in this article. The study was performed in a single center in China to identify fractures characteristics and soft tissue injuries in patients with posterior cruciate ligament tibial avulsion fracture. This is a relatively rare injury. The fracture is often accompanied by soft tissue injuries, most commonly involving the medial collateral ligament, and they aimed to determine the rate of MCL injury and its association with fracture characteristics in patients with this injury. They noted that patients with PCLIF were identified with CT and associate ligamentous and meniscal injuries. They were evaluated with magnetic resonant imaging. Fracture morphology was assessed with heat maps and quantitative measurements. Receiver operating characteristics and logistic regression analyses were utilized to identify predictors of MCL injury. They had a total of 148 patients with this injury, 28.4 had a concomitant MCL injury and MCL injuries were significantly associated with posterior horn tears of the medial meniscus and patients with MCL injuries exhibited a larger fracture distribution area on heat maps and the quantitative analysis showed that these patients had a significantly smaller fracture medial border and a significantly larger fracture AP diameter. Multivariable analysis identified a fracture AP diameter percentage of more than 50% as an independent predictor of MCL injury. They concluded that MCL injury is relatively common in patients with a posterior cruciate tibial avulsion fracture and it tends to occur concomitantly with a larger evulsed fragment. Next, we have outcomes and complications of vertebral body tethering in skeletally immature patients with idiopathic scoliosis. By imbrault There is a commentary on this article. Vertebral body tethering or bt aims to gradually correct scoliosis using patients growth while preserving spinal motion. The authors reported a five to eight year outcomes and complications in skeletally immature patients. It was a prospective single center COHORT study with 74 patients with idiopathic scoliosis with more than or equal to 5 year follow up. The preoperative first post op visit and 1 year, 2 year and more than 5 year radiographs were analyzed and a more than 5 or equal to 5 degree increase in the inner screw angle suggested tether breakage. All 74 patients there were 5 male and 69 female were skeletally immature at surgery. The mean age was 11.8 and the mean follow up time was 63 months. The findings were that 66% of patients had a radiographically suspected tether breakage after five years and 13.5% of patients required posterior spinal fusion. The VBT yielded significant correction at the coronal plane and transverse plane with a reoperation rate of 21.6%. So there you have it. That is this issue of JVJS and this concludes your cases on hold episode 114. If you like what you heard, please like and subscribe wherever you receive your podcasts. And join us next time when Andrew [00:45:45] Speaker A: will be the host and we'll be in October. Halloween will be right around the corner. [00:45:51] Speaker B: Thanks everyone. [00:45:53] Speaker A: Sam.