All right, so we'll go on to our first paper. Robbie Westerman from Iowa is going to be talking about the true learning curve in hip arthroscopy according to patient-reported outcomes. Dr. Westerman. All right. I'd like to thank the AOSSM for the opportunity to present, and I'd also like to thank them for not putting us at the end of the day on specials day like usual. Here are my disclosures. None of them are relevant to this talk. So we know that well-performed hip arthroscopy has tremendous potential to help patients in terms of improving patient satisfaction, patient outcomes, and providing high rates of return to sport. Surgery, however, needs to be properly indicated and performed in order to achieve these outcomes. The problem is surgeons experience a dramatic learning curve when gaining competence and learning how to treat FAI arthroscopically. The problem is we don't really know what that learning curve is. Is it 20 cases? There was a discussion around the arthroscopic Latter-J is only 20 cases, and you have it just like a hip scope. Is it 519 cases, which is what the New York State databases might suggest? And what do you measure? Is it operative time? Is it traction time? Is it complication rates? Is it revision rates? Is it conversion to total hips? We don't really know what we're measuring in terms of the learning curve. PROs have been very sparsely reported as it relates to this. So what about re-operation? This is a New York State database study that demonstrated you need to have over 518 cases of hip arthroscopy before your re-operation rate goes below 10%, which is an incredible amount and discouraging to a lot of young surgeons. What about yearly volume? This is a Deegan paper, another New York State database study, showing that two-year, five-year, and 10-year follow-up, you basically need to be doing over 300 cases a year to have good outcomes and low re-operation rates around the hip. So this is the Dunning-Kruger effect. So this is confidence on the vertical axis and knowledge on the field on the horizontal axis. You see, you start as a medical student with very low knowledge and confidence. By the time you're a chief resident or fellow, you pretty much know everything. By your first or second year in practice, you realize things are getting a little more complicated. So surgeons that are doing hip arthroscopy need some guidance about what the learning curve is and how long it's going to be. This is courtesy of Josh Harris, this Dunning-Kruger effect. So the purpose of this study was to evaluate the true learning curve in hip arthroscopy for FAI according to patient-reported outcomes. This was a retrospective analysis of my first 1,000 cases in practice over the first four years. Cases with concurrent PAO or femoral osteotomy were excluded. We took convenient samples from each year in practice and evaluated them according to WHOS scores. We had WHOS pain, physical function, pre- and post-op at minimum one year. And then we had MCID for WHOS, which is 10 points, and we used simple statistical methods. The mean age in the patients for these first four years was 28, 56 percent female, labral repair 95 percent, osteochondroplasty 100 percent. The mean one-year post-operative WHOS pain scores in my first year of practice was 82, and then years two through four it got better, around 90 to 93. My annual volume also increased from 150 to about 325 in year four. What about minimally clinically important differences? During my first 250-case period, 22 percent of patients did not meet MCID criteria by WHOS, and between cases 251 and 1,000, we had 11 percent that did not reach MCID. So a big shift after about 250 cases. If you graphically look at the WHOS outcome scores at one-year follow-up between cases at one and 1,000, you look at patient acceptable symptom state here, you see that after about 150 to 250 cases, more patients are reaching PASS. The trend line continues to go up even after 500, and there's still some whiffs later on as well. Again, annual volume is low at the first part of this graph and higher at the second part of the graph that may be an important confounder. This is an example of a 17-year-old American male football player with severe bilateral FAI for people that don't treat hips. This is a very severe case, alpha angles over 90 degrees, very limited internal rotation. I treated this kid. His right hip was my 39th hip scope, and his left hip was my 200th hip scope, and you can see a residual FAI on his right hip scope that required revision at 1.5-year post-op. So what about confounders? So during my first year in practice, I was only closing about 75% of capsules, but Shane Ngo and his excellent work around the capsule demonstrated why I should routinely close this every time. So that's an important confounder that could influence the increase in PROs throughout the learning curve. Chris Larson demonstrated why you should do an anterior inferior iliac spine decompression in cases of large deformity or low femoral anteversion. That's something that I incorporated in my last couple of years as well. And then Mark Philippon and Chris Larson also showed me some pearls about dynamic assessment of the hip before and after FAI correction in order to make sure you're comprehensively assessed for adequate resection of FAI and restoration of motion. Finally, the indications are different. During your first year in practice, you get a lot of low back pain with a labral tear on their MRI or SI joint pain. And the further you get along, you have better indications, narrower indications, and better referrals. So this is another important confounder in trying to do a learning curve analysis. There's limitations. There's a big increase in volume, which may confound the results based on some early statistical analysis. The patient report outcomes are not obtained in all patients. And changes in practice like capsule, sub-spine, dynamic assessment are also potential confounders. So in conclusion, there's great improvements in the one-year PROs after 150 cases, more likely to reach patient acceptable symptom state. After 250 cases, MCID was cut in half from missing at 22% of the time to missing at 11% of the time. And this corroborates well with some of Shane Ngo's work published a couple years ago. So the take-home points for those educating fellows, the PRO learning curve is probably around 250. Continued improvements still happen between 750 and 1,000. Higher annual volume is something that should be considered. In my personal opinion, I think people that are proficient in hip arthroscopy, it should be about 50% or more of your practice. But this can be variable depending on your annual case capacity. So thank you for your time. Thank you.

Dr. Robbie Westerman

hip arthroscopy

learning curve

patient-reported outcomes

re-operation rates