Listen to Andrew Duckworth & Xavier Griffin discuss the paper 'Effect on health-related quality of life of the X-Bolt dynamic plating system versus the sliding hip screw for the fixation of trochanteric fractures of the hip in adults: the WHiTE Four randomized clinical trial' published in the February 2021 issue of The Bone & Joint Journal.
Click here to read the paper
[00:00:00] Welcome everyone to our BJJ podcast for the month of February. I am Andrew Duckworth and a warm welcome from your team here at The Bone & Joint Journal. As always, we would like to start by thanking all of you for your continued comments and support as well as a big gratitude to our many authors and colleagues who take part.
As we move forward this year, which we hope will see us moving onwards and upwards from a very difficult 2020 for all of us, we aim to continue to deliver a range of topics through our series with our primary aim, always to improve the accessibility and visibility of the studies. We publish at the journal, for both you as our readers and listeners, as well as for our many authors.
We kicked off the year in January with a great discussion from Mr. Mike Whitehouse on their meta-analysis on the effect of antibiotic-loaded bone cement on the risk of revision, following hip and knee arthroplasty. And this month we will continue in the same vein with another excellent high quality research study being covered.
As you know, over the next 20 to 30 minutes or so we'll cover a range of aspects of the chosen paper, emphasizing the important points of how the study's been put together, as well as some take-home messages from the paper and how these potentially fit into each of your day-to-day clinical practices.
So today I have the pleasure of being [00:01:00] by Professor Xavier Griffin from Queen Mary University, London, to discuss his study, entitled The effect of the X bolt dynamic plating system versus the sliding hip screw for the fixation of trochanteric fractures of the hip in adults: the WHiTE Four randomized clinical trial, which will be published in the February edition of the BJJ.
Welcome Xavier and a big thank you for taking the time to join us today.
Thanks, Andrew. My pleasure.
If we go straight to the paper, the aim, obviously, this was a multicenter randomized controlled trial. It was to compare the clinical effectiveness of the novel x-bolt hip system, with the sliding hip screw for the treatment of fragility hip fractures.
So if you could just give us a brief introduction and background to the paper and maybe some literature regarding the well-established workhorse, that is the sliding hip screw.
Yeah, course. So, everyone listening to the podcast today will know how big a problem hip fractures are for us, but you know, the headline numbers here, we're talking approximately, 70,000 hip fractures across the UK per year. And the cost of [00:02:00] that problem is estimated to be in the order of 1% to 2% of the total healthcare expenditure in the health and social care budget in the UK.
Hip fracture inpatient admissions take up more bed days than stroke and schemic heart disease combined. So we're talking about a colossal public health problem here.
So looking into the detail of how we as surgeons treat the hip fractures, well we're interested here in trochanteric fractures and they make up about 50% of the entire hip fracture burden. And typically we would fix those fractures, and the workhouse historically and in many countries around the world today is the sliding hip screw.
So this is a device with a plate and a barrel that is screwed to the side of the femur once the fracture is reduced and then a lag screw is placed across the fracture. And the great innovation here is [00:03:00] that the lag screw is able to slide in that barrel and in doing so allows the fracture to collapse.
Well, when the fracture collapses, you then end up with a load sharing construct where you've got reasonable bone on bone aposition across the fracture. And you have the implant clearly crossing the fracture as well.
That's important because most of these fractures that we're fixing are fragility fractures. So they are a consequence of reduced bone mineral density or osteoporosis. And if we were to use historical, non-sliding implants, where we're relying on purely the fixation in the fracture with the implant itself, they tend to cut out because the bone can't withstand the interface forces.
So the sliding hip screw is the workhorse and you know, listeners will be familiar with some fracture patterns in some countries are using an intramedullary nail, but the features are similar. Again, it's a lag screw, with a [00:04:00] fixed ankle device.
Now the big innovation here is the x-bolt system. That's what we were testing. So what does that really look like? Well, everyone could imagine a sliding hip screw screw I'm sure. It's a barrel fixed to this side of the femur, as I was explaining earlier. The x-bolt is the identical in that sense, but within the barrel, we have, instead of a screw, we have a bolt. And that bolt crosses the fracture and where it terminates in the femoral head instead of a screw thread, we have essentially an expanding wall bolt. If you imagine that when you screw it into a lightweight brick or a breeze block, you have those expanding bolts and the flanges that expand up. It's actually developed from the technology that's used to create a car Jack. So people who have jacked their car up to change their wheel on the side of the road will remember that you screw. [00:05:00] And as you screw down the flanges deploy and raise the car. It's the same technique, but here, the flange is expanded into the bone.
And the concept is that those flanges generate improved rotational control and they also improve the risk of cutout those. They reduce cutout because they get a better fix in this parotic bone.
And that was the implant that was brought to us several years ago. And, in fact, in The Bone & Joint Journal previously, we've published some exploratory work, in a small pilot study. And, there's some biomechanical work which the manufacturers have done to demonstrate that the pullout strength is improved in comparison to the sliding hip screw.
So that's sort of background to the study and fundamentally our research question was, in a multi-centre pragmatic study, could we demonstrate that when comparing standard of [00:06:00] care, sliding hip screw with the x-bolt system, could we demonstrate that this effect that we found in the preliminary work was maintained?
Yeah. Yeah, absolutely. That's a really nice background to it, Xavier. I think, like you say this procedure of the x-bolt implant. And if that actually has a clinical effect and a positive clinical effect for our patients.
So if we go on to the methodology, which obviously with a trial of this nature really important and very robust as the listeners will see, but it's a multicentre, multi-surgeon two group parallel group randomized controlled trial. It was carried out in 10 acute hospitals throughout the UK. So Xavier key to any study of this nature, what were the inclusion, exclusion criteria you used for the trial and why?
So, those of you who are familiar with the WHITE studies and many of the other trauma studies, that have been done over the last 10 or so years in the UK, it was a very expansive eligibility criteria. We want to really mirror surgical [00:07:00] practice rather than end up with a very kind of unique bespoke data set, which is not generalizable.
So, essentially the inclusion criteria, those people that we're interested in are any patients presenting with a trochanteric fracture of the hip here in the opinion of the surgeon confronted with that case in a trauma meeting would plan to do a sliding hip screw fixation.
We excluded patients younger than 60 years of age because we're trying to focus on fragility fractures, and that's our sort of working proxy definition of that. We excluded patients with a subtrochanteric fracture because that's probably now generally agreed to be best treated with an intramedullary device. And we excluded patients clearly who were managing without operations. So the non-operative group.
Yeah, so very broad criteria there really. Yeah. To include as many and we say fit the generalizability in the normal clinic surgical practice that we see.
Yeah. Okay. And in terms of the sort of pathways that they went down, you know, it's a [00:08:00] pragmatic trial, the treatment pathways in the two arms, were they standardized in any way, or were they very much a pragmatic view as well?
So we did describe for the reader what we did, but essentially we mirrored NICE guidance, blue book guidance, and then we just wrote that down. So we've really followed local practices. That's pre-op assessment and aesthetic was achieved in accordance with the local policies, perioperative analgesia with nerve blocks or intravenous paracetamal or opiates, typical anesthetic and perioperative care.
The surgery was not specified beyond a closed reduction of the fracture on a traction table. And, once you are happy with your closed reduction, you proceeded to one of the two randomized arms, whether that be the sliding hip screw or the x-bolt hip-plating system. We did allow surgeons to augment fixation in the way that they felt was appropriate. [00:09:00] And we've reported that in the paper, but essentially we recognize that some surgeons are augmenting DHS with D rotation screws, or they may need to cable a wire or a component of the fracture together. Some people tend to try to capture the lesser trochanteric fragments, and we also recognize that there's certainly a use amongst some surgeons of trochanteric stabilization plates, particularly in fractures, which tend to be described as A3 under the OTA classification. But those fractures at the level are less trochanteric or in reverse obliquity.
So we reported on all that fully in the paper in a separate table. But actually, you know, 90% of all the operations were just either sliding hip screws or the dynamic plate. And then rehab was much as you expect, so early returned to weight bearing with physiotherapy.
Yeah. And as you say those sort of caveats you described, they're relatively few aren't they? [00:10:00] And if you look at table one and the listeners can see it in the paper, you know, the baseline characteristics of those two groups are very well matched as well.
In terms of the primary outcome then what did you choose and why, how did that sort of determine your sample size as well?
Yes. So I think this is actually a really crucial question in how you design the study and how you answer the question because ultimately the primary outcome is the only outcome against which we could hang our hat in terms of making the assessment of to what degree there is a difference between the interventions.
And so the choice that exactly which instrument you use to measure outcome is crucial, right? Because that's the definition of successful failure in that sense. So we chose health rates and quality of life as our outcome of interest. And we used the EQ5D instrument, which I'm sure listeners will be familiar with, which is the sort of [00:11:00] typical European option. Often in the States it is measured with the SF, six, 12 or 36, but here we use the EQ5D which is the nice recommended tool. So they behave very similarly. Why did we use that? Well we used it because that's what patients tell us is the most important thing to them following a hip fracture.
So we've done a lot of work over the last few years with patients and indeed experts across the spectrum of the multidisciplinary team that look after people with broken hips and it's really about return to independent living. It's a return to your pre morbid or pre fracture, quality of life. That's what they're interested in. And so that's what we used as our primary measure. Now you can argue about that, but it seems to me that if you're going to judge a new implant or a new technology, then you should measure it against the, uh, the things that are important for our patients fundamentally.
Yeah four months was our primary measure and the reason we chosen four months is [00:12:00] we've done some exploratory work to see how recovery tracks so you got about a, you've got a very severe decremant, decrease in your quality of life after fracture. That's not surprising you're in bed, you're in pain. You don't know what the future holds, and you're very anxious. So it's no surprise that patients after fracture have a very low quality of life. And then it recovers and it recovers to about a 20% reduction at one year, compared to where you were pre fracture and that's a big, that's a big hit in quality of life, it's similar to a big stroke A.nd at four months you pretty much recovered all the recovery that you're going to experience.
The issue with waiting 12 months is that this is a very frail population and we start to get a lot of attrition in our data because of people unfortunately dying. So if you wait too long, you lose a lot of data. If you go too early you probably don't represent the absolute recovery and four [00:13:00] months seems to be about the right time. And that is similar to what the national hip fracture database are doing. So, that seems to be the sweet spot four months.
Absolutely. And in terms of your sample size and that required, what did you calculate for that? And what was the difference you were sort of looking for between the groups?
Yeah, so this is really important when we discuss the results. This is the sort of magic number we have to keep in our head. The clinically important difference that is thought to be around the kind of number that patients can perceive to be different and that they value as a difference is 0.075 on the scale. So 0.075 is the number that we should judge our confidence interval later on. It's really important that we specify that upfront so that we can't kind of fudge our results later, but that's the number we used.
And we went for 90% power. So he wrapped it right up and a standard 5%, P-value that the readers will be familiar with. And when you run that through, you come up with just shy of 340 [00:14:00] participants in each group. And those that have read the paper think hang on 340 times two, that comes to fewer than you recruited. That's because we anticipated that we lose about 40% of patients through death or withdrawals. And so when you run all those numbers up together, you come to 1,128 participants. And that's what we collected.
It's quite a thing with the hip fracture trials. Isn't it? You do have such a big drop-off with expected mortality, but anyway, with these injuries, it's amazing. Isn't it? So that's very clear. And in terms of just before we move on to the results, then what about secondary outcome measures? What sort of things were you keen to look at? Obviously the EQ5D was the primary outcome, but obviously something talked about was, you know, the perceived benefits of the x-bolt. There were other things you looked at that were maybe potentially been beneficial in terms of maybe failure or revision surgery as well.
Absolutely. So we tracked quality life out to 12 months. We looked at mortality, but then we looked at some of the slightly more interesting things. So functional status using the scale that we collect in the [00:15:00] National Hip Fracture database about how mobile people are. We looked at residential status in terms of whether people back in their own home, whether they'd gone to a care home or a nursing facility. We looked at all-cause revision surgery because, you know, ultimately that's probably one of the top ways that we might judge the technical capability of the surgery and the implants. And we looked at all cause complications. We did look at radiographs as well because, you know, I'm not sure that I really believe it, but lots of people think that radiographic outcomes are important. And, I guess as a technical measure of our capacity to influence fracture healing, it's a reasonable thing to look at.
Unfortunately, because we so rarely take x-rays in the postoperative period from our hip fracture patients across the country. Most of these data are not available because we're only collecting radiographs that were acquired for clinical [00:16:00] uses, we didn't mandate it as part of our protocol. Those data are available. We haven't published them in this paper here because it's quite a complex data set. We'll probably do a supplementary publication later, but that was our set of secondary outcomes.
Yeah, absolutely. So if we move on to the results, sort of keeping those in mind . There were just, like you said, 2,105 patients were assessed for eligibility, and eventually that whittled down to 564 eventually randomizing into each group. It's a very large number. Well over a thousand, as we've already said, the baseline demographic data was similar in both treatment arms, as you'd expect for the randomization particular that size. Participants had a mean age of just over 85, 86 years and just over 70% were female and the pre-injury median EQ5D was 0.57.
So all very in keeping with the normal patient population we'd see with hip fractures. And just before you go into the sort of key findings, can you tell us about any sort of crossovers between the groups? And, what about the, you know, the sort of follow-up [00:17:00] rates at four months and one year, were they sort of what you were expecting and anticipating?
Yeah. So, our consult diagram are laid out in figure three for the people that are reading the paper, but you're absolutely right. There was a crossover rate. So there's a group of patients that we didn't know what they got because they withdrew and we weren't able to collect data from them. It's this sort of feature of the fact that we're including patients without capacity. We do get this post randomization withdrawal phenomenon.
So. 25 in the x-bolt group and 30, we weren't able to collect any data about those people because they withdrew their consent. The headline, I guess though, that people will pick up on it is that within the x-bolt group, 564 randomized 476 received the x-bolt, which means that 60 odd people, did not receive the x-bolt. The vast majority of those actually received a sliding hip screw.
On the other side of the flow diagram, you'll see that the vast bulk [00:18:00] of people allocated to sliding hip screw received a sliding hip screw. And there wasn't very much deviation from that. There were one or two people that on table, the surgeon decided to do an intramedullary nail. That was the most common cause. So there is a crossover. We've got some sensitivity analyses to handle that. We expected that in surgical trials, it's not uncommon to see a 10 to 20% crossover rate here it is just that 10%. So we felt that was reasonable. We've got a stack of sensitivity analyses pre-specified, which we reported in the supplement to the paper.
Essentially the take home is, before we kind of get into the weeds of exactly what the results show, the take home is every single analysis we did, cutting it in each different possible option, all showed the same result, which really draws out the experiment was successful. You know, the randomization was successful.
In terms of follow-up, the four month timepoint was our key time point. In the [00:19:00] x-bolt group, we had 324 completed their form at the time point of the 564. And on the other side and the sliding hip screw group 318 and the 564. Yes. That's not an enormous number butvover a hundred patients died in the two groups. And when you add all that together, we grossly exceeded our attrition. Well we greatly exceeded our follow-up post 60% assumption. So we had more data than we had anticipated at the time. So our team in the office did a fantastic job of collecting data. We're really thrilled about.
Absolutely. I think for that volume of patients it's really very impressive. And like you say, it's a difficult patient population to try and get all that data on.
So if we do move on to the key findings. I suppose we should probably look at the primary outcome measure the EQ5D and then we can look at the secondary outcome measures. What did you find?
Well, before we get into the analysis Andrew, the key finding for the primary outcome was [00:20:00] that there was no statistical or clinically relevant difference between the two treatments. I mean, that's the headline finding in the study . Now I said the number we needed to keep in mind was 0.075, because that's what we specified right up front as our clinically relevant difference. And when you look at the primary outcome, with the stats plan that we had, we had a difference of 0.029. So the point estimate much lower than 0.075. So essentially less than clinically detectable. And when you look at the confidence interval, this is the really important bit, the confidence interval goes from minus 0.013 to plus 0.07. So both of those extremes still less than a 0.075. Yeah. So really tight confidence tools. And essentially what we're saying is, with 95% certainty, no [00:21:00] possible effect exists that is clinically relevant here. So I mean, an absolute slam dunk result. People say it's a negative finding. Yes, it's a negative finding, but it's really a precise negative finding that we can hang our hats on and say actually nothing important to find here.
Yeah. Yeah, absolutely. And then in terms of your secondary outcomes, I suppose the 12 month EQ5D and mortality and revision.
Yep. So, no differences in any of the other secondary outcomes. Now because, the data become a little bit less available. Sometimes those confidence intervals stretch up a little bit more, but essentially we're looking at really everything in our secondary outcomes pointing to the same finding as our primary analysis for our primary outcome, which is the difference between them.
So, you know, very reassuring that all of our secondary measures supported are our principal conclusion. I guess people will be, [00:22:00] because we're all surgeons I understand that, there'll be particularly interested in what happened in terms of the all cause revision surgery. So, because this is an analysis, which is based upon adverse events, typically we use a per protocol analysis here. So instead of saying, what were they allocated to? We are classically and certainly in our pre-specified analysis, we'd say, what did they actually receive? Did they follow protocol? So here we had the two group sizes, which are a bit different from the randomized sizes. We're talking about 401 people in the x-bolt group, 493 people in the sliding hip screw. And this is table four for people that read in the paper. Yeah. But we're looking at all cause revision surgery risk. And this is percentages now 2.7% in the x-bolt group and 2.6% in the sliding hip screw group. That's all-cause revision surgery. And [00:23:00] one of the things that people have said to me. And I think it's a really great point is. Actually the bit that we're most interested in probably is what's going on in the femoral head because that's where the implants differ. And essentially we're seeing no difference, so failure of fixation, 1.7% in the x-bolt group, 1.8% in the sliding hip screw group. So essentially these are behaving in exactly the same fashion in terms of revision or failure.
No, I agree. I mean, you didn't go down that table four like you described, they're almost identical numbers, really all the way down, aren't they, in terms of the two different implants and the reasons for revision.
Great. So if we go onto what it sort of means, I suppose, you know, the strengths of study are clear without question. It's a huge trial and methodology, very robust, it's large, it's pragmatic, it's multicentre, multi surgeon, randomized trial, you know, it's also a randomized study that sort of nested in the WHITE cohort, low risk of confounding. And I agree, [00:24:00] the findings are very likely generalizable and conclusive should we say. Well, what do you feel I suppose, the key take home messages and maybe what can you take home from what you learned from the trial? You know, sort of a newer implants and investigating that and any sort of potential limitations of the data to caveat that?
Yeah. So I think the first thing I say is that I've really enjoyed, this has been a long journey and I've really enjoyed going through the journey with x-bolt. They've been a fantastic partner. They first came to us with a device that was on the shelf, but had no clinical evidence and WHITE one was the pilot study. We helped them a little bit with the biomechanical stuff. And, the team in Bath University also were involved in that. And it was really, really, I think, laid the challenge for the bigger implant companies out there to say, this is how we should be doing business going forwards. If you come to market with a new implant, well, we want decent randomized clinical [00:25:00] effectiveness studies before we start rolling it out in a broad, NHS sort of pan NHS way because that's how we best protect our patients' safety. And also when we do find effective interventions, we accelerate their introduction to practice.
So I was hugely impressed by x-bolts desire to move through this pathway. And, and I haven't seen it happen with many other implants. So, that was, it was enjoyable journey. It's not easy though and I'm hugely thankful to the 10 hospitals that participated because, you know, hip fracture, sliding, hip screw, some people are doing those. They've done several when they were SHOs and now Caltrain. Right, this is cut your teeth type of orthopedics. And so to bring a new implant into that arena is tough and it's challenging for surgeons to pick up something new when actually they've got [00:26:00] something that works well. So hugely indebted to the teams in those 10 hospitals, but it did show that the UK can deliver at speed and at scale when we have a question to answer, so I think that's a big up for the NHS system in terms of how we can do studies.
I think that the answer is pretty definitive. I don't think we need any further studies. And I think the big take home for me was when you look at the patients that were included in the study in the baseline demographics table, you'll see that about half the patients had A2 fractures. So that's a culimuted , three, four part fracture in the trochanter but a standard obliquity. About 35 or 40% of the patients had an, A1 fracture. So that's a classic two-part fracture standard obiquity. And about 5% of patients had A3 fractures, reverse obliquity or fractures at the level of the lesser trochanter. So we really were including the whole [00:27:00] range of patients and we're finding revision rates in NHS practice of between 2 and 3%, which is just fantastic.
And a little bit of extra work that I did looking around these data to see if these are really believable, because you know, they eclipse a lot of the published data. If you look at the Cochrane 2010 report by Martin Parker. You're you're seeing revision risks exceeding 12 and 20% in the studies, comparing them with intermedullary nails. And, here we're finding a revision risk of 3%.
Actually, when you plot those studies out and you look at the funnel plots, it's exactly where you'd expect it to be. It's bang on what the weighted average should be. And so I think it is believable and it's a huge positive for us as a system of care that we're doing something really fantastically well. That actually the sliding hip screw is a great [00:28:00] implant used in the hands of surgeons in the country for patients that are going through standard NHS treatment pathways. Things are going well, really well.
And I think that's one of the big home take home messages from me as well as that just that low risk of revision surgery, you know, from a range of hospitals, huge number of patients. And I think that's the thing about it. Isn't it it'd be hard for any device really to improve on that. They're such good numbers in terms of how the sliding hip screw is doing.
Yeah. And you know, it's sort of, it's another query because you know, the data here are fantastic. We approached all our patients at four and 12 months, either they are a participant themselves orthey are a carer. And we asked them if they'd had any revision surgery. We contacted every hospital for every single patient. And we asked them to do a medical reviews at 12 months to see if they'd had any revision. And we asked everyone to have a look at their PAC systems to see, because, you know, you don't get a revision without having a pre-op [00:29:00] x-ray and typically a post-op X sheet. And so we asked everyone to have a look at their PAC system. So we did a triple whammy to try and pick up revision. And you can't tell me that that registries are gonna outperform us, in terms of the quality of the data, of course, the size of the data, the total number of people they can eclipse us. But randomized trials are great for measuring effectiveness, but that we'll say fantastic systems for collecting reliable data. And so I think the, the revision risk estimate here is very reliable and it's consistent across 10 hospitals and it's probably representative of what's happening in the UK.
And so you've got to ask yourself, well, you know, do we really need to use anything other than a sliding hip screw to treat these fractures? Are we going to get outcomes that are better as you say. And I'm not sure. It's tough. It's a tough implant to beat.
Yeah, absolutely. No, I agree.
[00:30:00] I think that's a good note to finish on actually so uh, that was a really great discussion and thanks so much for taking the time to join us and congratulations on an excellent, excellent trial and the group of WHITE studies as a whole. So it's great to have you with us.
Thanks very much.
And to our listeners, we do hope you've enjoyed joining us. And we encourage you to share your thoughts and comments through social media and a like. Feel free to tweet or post about anything we've discussed here today. And thanks again for joining us. Take care, everyone.