Five-year follow-up of the Scaphoid Waist Internal Fixation for Fractures Trial (SWIFFT)

Listen to Andrew Duckworth and Joseph Dias discuss the five-year follow-up of the Scaphoid Waist Internal Fixation for Fractures Trial, the results of which were published in the January 2026 issue of The Bone & Joint Journal.

Click here to read the paper 'Clinical effectiveness of early surgical fixation versus cast immobilization for adults with a scaphoid waist fracture'.

Click here to read the paper 'Cost-effectiveness of early surgical fixation versus cast immobilization for adults with a scaphoid waist fracture'.

Click here to read the paper 'Radiological outcome of early surgical fixation versus cast immobilization for adults with a scaphoid waist fracture'.

Find out as soon as the next episode is live by following us on X (Twitter), Instagram, LinkedIn, Tik Tok or Facebook!

Transcript

[00:00:00] Welcome everyone to our first BJJ podcast of 2026. I'm Andrew Duckworth and a warm welcome back to you all from your team here at The Bone & Joint Journal. As always, we'd like to thank you all for your continued comments and support, as well as a big gratitude to our many authors and colleagues who take part in the series that highlights just some of the great work published by our authors each month.

So today for our podcast, I have the pleasure of being joined by an author from not one but three papers published in the January edition of the BJJ. Looking at the five year results in terms of clinical and neurological outcomes, as well as the cost-effectiveness of the SWIFFT trial, which looked at early surgical fixation versus cast immobilization for adults with a scaphoid waist fracture.

Today I have the absolute pleasure of being joined by my colleague and friend, Professor Joe Dias from Leicester. Joe, so pleased to have you with us today. Thanks for joining us. Thank you, Andrew, for inviting me to discuss the SWIFFT five year results. It was a great way to start 2026 that on the 1st of January

we had three papers. I'm so grateful that they've all come out together. Absolutely. No, I think it, it's, they really [00:01:00] do fit well together, don't they? And well, obviously that, that's what we'll be talking about a bit today. So, but, so, Joe, to kick off maybe, I thought though, for our listeners, you know, I'm, I'm just gonna take you back and, you know, it'll be

probably over a decade now, you know in when this all, all started and it would be chewing over in your mind, I'm sure before the SWIFFT trial. Can you give us us an idea of what the current state of the literature and practice was prior to the study and, and what the landscape was like at that stage before you started SWIFFT?

I think it all started 42 years ago. In. In 1984, Tim Herbert described the headless Herbert screw. It was an excellent innovation. It simplified the method of stabilizing a fracture of the scaphoid, which is a bone that is covered with articular cartilage, which you want to preserve. Tim went on to recommend its use for all acute scaphoid fractures and colleagues acutely fixing

[00:02:00] scaphoid fractures claimed various benefits for this method, including that union was quicker, there was no need for immobilization, earlier return to function, less time off work, and earlier return to sport. So all these suggested benefits sounded very attractive. The surgical technique itself was quickly well established, and colleagues from around the world, Europe, Australia, Canada, US started fixing scaphoid fractures acutely as a treatment of choice.

This got taken up within the UK and this led to where in the nineties and the early part of this century, with several of us starting to investigate in small, randomized controlled studies, whether there were benefits, whether the benefits that were suggested were actually [00:03:00] seen. We ourselves reported a randomized study in a small cohort of 82 patients in 2005, and demonstrated that at eight weeks, patients who had internal fixation were performing better than those that were put in a cast.

But bear in mind that at eight weeks, you've only just taken patients in a cast out of the cast. So that was not unexpected. After eight weeks, there was no difference in pain function or wrist movement. Between six and 12 weeks of the 44 that were treated in a cast, in ten we were not certain about union, and we investigated it on CTs and also offered surgery, which got taken up by seven.

We operated on these and four of these, there was no movement at the fracture site and all went on to join. We reviewed this particular group seven years [00:04:00] later and established the rate of arthritis after this injury, but there were no other differences that were noted between those that were fixed and those that were not fixed.

But at around that time, there were other studies published which had different conclusions, some showing benefit to having early stabilization. So there was an element of uncertainty. In spite of this uncertainty, there was a rapid adoption of early internal fixation of scaphoid based fractures.

In England, for example, the hospital epi episode statistics showed that of the expected 4,500 acute scaphoid fractures in England, 1,500 were internally fixed. That is a third in 2007. By 2009, 2,500 were [00:05:00] internally fixed, so the rate had increased to over 60%. The rate of internal fixation was increasing, even though there was no clear evidence of benefit for internal fixation.

Both these factors, the uncertainty from previous small studies. And the increased rate of internal fixation of acute scaphoid fractures led us to propose the SWIFFT study. Yeah, Joe, thank you. That's such an amazing overview of, and, and do I remember then in terms of that huge explosion, I didn't really know those numbers about from his data.

That is incredible that increase over that two, three year period to the number of people that were being fixed. And I think you're right. I think, I think what was driving that was the, the, this sort of data from those smaller studies and particularly that maybe quicker return to function that was perceived was a, a big driver of that.

That's really interesting. I think really sets up nicely why you developed the SWIFFT study. So if we go onto that. Joe, and, and we move on. Before we move on to the, the five-year studies, which we're gonna discuss, which [00:06:00] are in this month's Journal, if we then just maybe to discuss the SWIFFT trial, you know, you know, how what, what the design of it was.

You know, what the primary aim was and any sort of key, key factors in the design or the protocol. The Scaphoid Waist Internal Fixation for Fractures Trial. The SWIFFT was a pragmatic randomized controlled superiority trial, which aimed to see if surgery for acute scaphoid fractures improved results.

We randomized patients with a clear bicortical fracture of the waist of the scaphoid to either having internal fix screw fixation within a fortnight, or to be initially mobilized in a below elbow cast for around six weeks and only to fix those where the fracture was not uniting. At six weeks if on scaphoid X-rays, there [00:07:00] was any doubt about the state of union of the fracture,

this doubt was resolved on a CT scan. The main emphasis was to confirm the state of union using the CT scan. If there was a clear gap across the entire surface of the fracture, then this was defined as a nonunion, and if a nonunion was confirmed, that patient was offered immediate internal fixation.

The main outcome of the SWIFFT study was the patient rated risk evaluation or the PRWE score, which is based on 15 questions, and it allows you to have two domains around pain and around function. The score ranges between zero and a hundred and zero is a normal wrist without any symptoms. The PRWE minimum clinically important difference that had been [00:08:00] established prior to the study, but only for distal radius fractures was 11.5 points.

We used six points, which is almost half the previously established minimum clinically important difference, but we were not certain whether the previously established difference would apply to scaphoid fractures. So we were being quite conservative and knowing that we might over recruit but also gave us the confidence that we were more likely

to be certain of the results that we, we got. We did the power calculations and based upon the smaller difference we set out to randomized 438 acute scap patients with acute scaphoid fractures so that we could retain at least 350 of them at one year. The key secondary outcomes were fracture union, which we assessed on CT scans at one year, and [00:09:00] complications establishing the harms of each method of managing the broken scaphoid.

That's really interesting, Joe, and I think he, it's amazing how, like, how from the, the, the literature you had the basis and then the, the trial has like very clearly been set up to try and address those concerns and questions, particularly with the size of the trial like you've done and also looking at that key outcome, like you say, of nonunion.

And in terms of that, Joe, for people who, obviously the, the main results for the one year results were published in the, in the Lancet, I think it was in 2020 it was. What were the key findings from that? And, you know, anything in particular in terms of the secondary outcomes you'd wanted to highlight as well?

The, the key findings that we reported at one year for the SWIFFT study were that there was no difference in pain and function between the scaphoid fractures that were fixed and those that were initially treated in a cast. The scores were around 11 for those treated with fixation and 14 for those treated [00:10:00] in, in a cast.

There was also no difference in nonunion rates, which was surprisingly very high. But moderate or severe complications, things like infection, screw penetration, and CRPS 1 were more common in those that had surgery as one would anticipate it was 12% with those fixed initially and 2% in those that were initially treated in a cast.

There was quite a big difference in the harms. Surprised us. Was the very few days difference in the time to return to work. Yeah. Which was given as one of the benefits of early internal fixation. The time of work was 16 days after fixing the fracture, and 18 days when treated in a cast.

Mean times. The median time of work was nine days after screw fixation [00:11:00] and five days after cast immobilization. So our patients went back to work. Even when they were still in a cast and a third of them took no time off work. Yeah. Yeah. This is an amazing population. Yeah, absolutely. I, I tell you, that was the biggest surprise to me from results.

'cause it, it really was. Contrasting to some of the previous trials that had really advocated that this was the big thing about early fixation, wasn't it? What do, was there any hypothesis about why you think that was at all that you, you, you felt, or, or just, just the findings there as they were.

No, I think it was how we practised medicine and in our previous study in 2005, the mean time off work was five weeks and six weeks. So the difference, the thing that changed between then and now was how we advised patients. So we advised, gave them confidence telling them, yep, get [00:12:00] back to work.

Do as much as you can manage the bone will take six, eight, 12 weeks to unite, but you can still get on and work. Yeah. Yeah. And those in the cast, what we'd be saying to them. Is, look, the important thing is to avoid movement. So if the cast broke and it was allowing wrist movement to come back early, yeah. So they could cast reinforced.

I think it was how we advise patients that changed in, in between the two studies. That's very, that's very interesting. Yeah. And actually, totally, totally makes sense. And, and I think just to highlight for our listeners, Joe, is that that key point you said about the design is that, is that early assessment and potential intervention, if it's going on to nonunion, that's a really key part of it, isn't it?

That is a key part. Not comparing cast only. Yeah. Versus fix only. Yeah, absolutely. Comparing cast plus treat any failures early. Yeah, absolutely. To early [00:13:00] fixation of all. Yeah, no, absolutely that and that, I think that's a really key point that you highlight really well. And I mean, you mentioned about the nonunion, but I think we maybe come back to that when we look at the five year outcomes.

'cause that's quite interesting how that has progressed. And so if we do move on to the three publications in this month's Journal and these were looking at. The clinical outcomes, the radiological outcomes, and, and the cost-effectiveness at five years, which for, you know, for our listeners, I'm sure everybody's aware, but to, to, to get data at five years in a trial like this and the quality of the data as we'll come onto is actually incredible, I think.

And it's a real testament to Joe and the team to, to have been able to achieve this, particularly in what is more of a predominantly young and quite mobile population. You know, it's not, not generally as easy. So, Jo Joe, just for our listeners, you know. What were the key findings and sort of the follow up that you attained?

Really mainly firstly with, with regards the clinical effectiveness. At five years, we, we reviewed all the patients and repeated exactly the same analysis as we had done at after one [00:14:00] year. So at five years, the pain and function assessed using the PRWE score was 9.6 after early fixation, and it was slightly better, 9.1 after initial cast immobilization.

The union this was not statistically different. The union rate was 1% different between the two groups. It was 98% in those treated with fixation versus 97% in those treated initially in a cast, and this difference too was not significant. We were really surprised at the rate of union that we saw, particularly as this was a pragmatic study.

Yeah. And did not meddle with the thresholds for decisions in the different units that were recruiting into the, into the study. Yeah. So that was. In how you would expect these decisions to [00:15:00] behave in real life? What, what we were surprised about initially, just as you mentioned we were looking after predominantly young working men.

And given that population we had planned when we were planning the study to have an 80 to have a 20% attrition at one year. And further 20% of ones that were left. So overall, we were anticipating that we would hopefully retain at least 66%

at one year. What we what we were amazed about was that at five years, 345 of 423, that's 82% responded. Patients responded. That was amazing. Yeah. Course from the hospitals, we had 412 out of 419, so that is 98% response rate. And this has provided us with great [00:16:00] confidence in the results.

Yeah, absolutely. I think having five year follow-up data at that level, Joey, I think it's just absolutely incredible and I say it is such a testament to the team and in and, and in terms of, maybe we'll come to the radiological union paper last actually, in terms of the cost-effectiveness. What did it show?

The patients randomized to early surgical fixation had higher costs to the NHS of 1,295 pounds more than those treated in a cast early. They also had marginally better quality of life or a one year period of 1% better QUALYs. And this difference was not significant. But that led to a cost per QUALY of 82,000 if you fixed all

fractures acutely. Mm-hmm. This is well above the [00:17:00] accepted threshold for cost-effectiveness used in the UK, for example, and most other countries as well of between 20 and 30,000 per QUALY. Yeah. To commission treatments. The probability of early surgery being cost-effective was very tiny in this model of about 5%.

At one year, our Markov model, which simulates the lifetime impact of the injury on the patient was based on many assumptions because there was no data on the impact or of arthritis, the impact of nonunion, et cetera. Five years after randomization, most patients in both groups returned to the quality of life level, which was consistent

with their age adjusted norms. Yeah. And did not require any further healthcare. Yeah. It also confirmed in the Markov model that was now updated on [00:18:00] actual data rather than assumptions that the initial cast immobilization was the most cost-effective strategy with the cost to our system of 1066 pounds less per patient compared to those who initially had surgery.

Yeah. And annual saving even now of between seven and nine million pounds a year. Yeah, absolutely. These findings confirmed clearly that initial immobilization in the cast with fixation only for nonunion that is seen in the tiny number of patients is the optimal cost-effective treatment for these patients.

Yeah, no, absolutely. And I think, like you say, you like, you do a lovely overview of that cost-effectiveness and it's so clear on every marker that you've looked at in terms of whether it be the cost per QUALY and the overall cost savings as well. And Joe, just finally, before we sort of wrap up, the final thing was the radiologic outcomes.

And I think [00:19:00] there's a really interesting data you've, you've gathered here at the five years regarding that progression to union and, and looking at arthritis as well. We were surprised about this as well, which is the partial unions, what, what we had done was we looked at the proportion of the area of the surface of the fracture that was bridged with bone.

So a hundred percent above 95% if it was bridged, it was full union. Yeah. And if it was 0%, it was a nonunion. Yeah. And partial union was between 20% and 70%. And what we found that partial union, all the patients progressed if there was greater than 20% bridging. Yeah. And this occurred without any intervention and any modification or, or protection.

And this is consistent with the finite element analysis work, both from [00:20:00] Switzerland and from the UK. Mm-hmm. So from a union point of view, we found that partial unions progressed. Mm-hmm. The other surprising thing that we found was the ones that were fixed, they still had a screw or an implant inside.

Mm-hmm. And therefore, of that group, the ones that were fully joined at one year, which is greater than 95%, more of them were less than 95%. And what was happening was trashielding. Yeah. Yeah. And how so? And more of them. But of course it was a functional scaphoid. Yeah. Yeah. And had no impact. But it was interesting to see that even the act of putting in a screw changes how the, how the physiology works and how the bone remodels.

Yeah. Over five years. Yeah. The, so that was union and both those were interesting, [00:21:00] interesting observations. One of which is so reassuring for patients. The second thing that we found that was quite surprising was at baseline. That is at the time they came into the study, 10% had arthritis at 9.8%.

Yeah. Around 10%. So one in 10 10 had preexisting arthritis in one of the joints of the wrist. At one year this increased to 27%. And at five years it was 52%. So that is half of them was showing cha changes of arthritis. Mostly mild, yeah. And mostly around, but mostly around the scaphoid and particularly around the SCA for trapezium trapezoid joint.

Mm-hmm. Which was seen more frequently in the ones that were fixed. And I think it is to do with the approach and possibly leaving a thread or two or in the joint. Yeah, it's true. Yeah. [00:22:00] So there was an impact mostly, although the rate of arthritis was higher and we think this probably reflects. The injury to the cartilage because the rate is the same in both groups.

It, it tends to be mired at five years. Mm-hmm. But it did have an impact on the PRWE. The PROM Yeah. Yeah. Very tiny impact. Yeah, no, absolutely Joe. I think that, like, that to me, what the fascinating, I mean all amazing radiological data that, but that, that, that, that, that line of fractures with a minimum of 20% union at one year consolidated with the passage of time without intervention is really interesting data, isn't it?

And actually. I think, you know, in my clinical experience, anecdotally that does occur. You do get people who just are very slow to heal, but they do heal and I think that gives great reassurance. And going back to the original protocol in terms of intervening at six weeks, if you're concerned, these are really for ones where there is nothing at all, there isn't [00:23:00] it?

It's for that there is a nonunion, but if there is some consolidation there and certainly 20%, you're probably okay to run with it. Is, is that your sort of take with it as well, Joe? Pretty much. Yeah. And we know that the amount of bridging that you need to do activities of daily living Yeah. Is between 20 and 30% percent.

Yeah, yeah, yeah. You know, so you don't need to protect. No, absolutely. Absolutely. Well, Joe, and that's great overview of those studies and, and I suppose just to wrap up just a few final quick questions. So. Do you feel, you know, you know, this has been like sort of your baby for five, ten years now. Are there any limitations or caveats to the key take home messages?

I mean, I think we've covered most of 'em, but anything you want to highlight in particular? I think the important thing with all these studies are the exclusions. Yeah. Yeah. In the study. So the results to be absolutely pure about this. The results do not apply to this the ones that were excluded.

Mm-hmm. [00:24:00] So that those were the ones that were displaced more than 2 mm. Yeah. Or to proximal pole fractures. Yeah. But both these are uncommon and account for less than one in 20. Yeah. Of acute scaphoid fractures. So the results are relevant to the vast majority of scaphoid fractures we see.

Yeah, absolutely. And looking back now, is there anything about the studies that were, I always ask this of a lot of our authors, anything you would change now knowing what you know, in terms of the design or anything? From the study itself? Not much. Yeah. Yeah. And I think the, the team has been fantastic.

Yeah. From. All the elements of the team. And also the colleagues, the clinicians that contributed to this have been amazing. Mm-hmm. Mm-hmm. Because we have discussed all our uncertainties as we went along with the whole team. Yeah. And resolved them as we [00:25:00] went along. And I think, and that was not just the clinicians, but also patient advisors and that was really useful. How we practise

is different. We spend more time explaining to the patients why the scaphoid fracture an injury that looks so trivial when you look at it has the potential to affect this young man's future wrist and hand function. And we ex explain to them that they are part of our team to achieve union. Yeah. And that by achieving union, we protect the function of the wrist.

Yeah, absolutely. It's the first thing. The second thing that is clear to us is that we should be advising all our patients about activity, even when they're recovering. Yeah. Really. And this is what I mentioned earlier, that we need to explain to them. That there is no harm and we now have the proof there [00:26:00] is no harm.

Yeah. With the returning to work. So we are very much more confident about saying, get on with life, use your hand for day-to-day stuff. If the cast is soft, then come back and get it reinforced. Yeah, we reassure them that partial union, the thing that looks not quite right, even at one year goes on and consolidates and we are not worried about it.

And actually it just proves that nature is an absolutely beautiful thing. Totally. And although there is the consequence of the injury with arthritis rates going up to one in two by five years. The ra, the, the severity of arthritis is very, very mild. Mm-hmm. So you're not, we, we are not overly concerned about that.

Yeah, absolutely. I mean, Joe, that's a, a lovely way to actually finish off in terms of that. I think those are all those lovely key take home messages from SWIFFT. And actually just an example of how, you know, really [00:27:00] well-designed big you know generalizable, pragmatic trial has really, you know, changed the landscape of what was

normal practise, you know, or was becoming normal practise back then and is now we have a practice based on real robust data. And I, you know, I've really enjoyed talking to you today, Joe and. I think our listeners have learned so much from, not only from the work, but hopefully from what we've chatted today, and just a massive congratulations to you

and all the team on such an incredible body of work, which has added so much high quality literature in this area. Not just in terms of the the clinical outcomes, but the radiological outcomes as well, and the cost-effectiveness, and it's been amazing to have you with us. Thank you, and thanks a lot. And to our listeners, we do hope you've enjoyed us joining us, and we encourage you to share your thoughts and comments through our social media platforms and alike.

Feel free to post about anything we've discussed here today. And thanks again for joining us. Everyone. Take care.