Guidelines for the follow-up of total hip arthroplasty: do they need to be revised?

Listen to Mr Andrew Duckworth interviewing Prof. David Beverland about his paper "Guidelines for the follow-up of total hip arthroplasty: do they need to be revised?", published in the May 2019 issue of The Bone and Joint Journal.

Click here to read the article

Transcript

[00:00:00] I am Andrew Duckworth and a warm welcome to now our sixth podcast from your team here at The Bone & Joint Journal. I would like to thank all of our readers and listeners for their comments we have received regarding the series to date, as well as to our authors and guests interviewers who have taken part in our series over the past few months. We really do appreciate all your efforts.

So far this year, we've covered a range of topics, including the role of robotic unicompartmental knee replacement with our editor-in-chief at the journal, Professor Fares Haddad. We've had an excellent discussion on the role of DDH screening with Dan Perry and Alex Arbolt, and last month, a really fascinating dialogue between Ian Murray and Dr. Scott Rodeo on cell therapies and orthopaedic surgery. We do hope these podcasts are improving the accessibility and visibility of the studies we publish both you as our readers, as well as for our many authors. 

As you know, we hope that during the next 15 to 20 minutes or so we'll cover a range of aspects of the chosen study, emphasizing the important points of how the work has been designed, as well as the key findings from the study and how these potentially fit into each of your day-to-day clinical practices, with this month's [00:01:00] discussion, I suspect being relevant to many of our listeners. We also hope the discussion will give you a behind the scenes insight into how the authors have developed the study and give them an opportunity to put forward the key findings of their work. 

So today I have the pleasure of being joined by David Beverland from Belfast to discuss their study entitled Guidelines for the follow-up of total hip arthroplasty. Do they need to be revised? Which will be published in the May edition of the BJJ. Welcome David and a big thank you for taking the time to join us. 

Thank you, Andrew. 

So David, moving on to the study. So it was obviously a retrospective study that was used to determine the roots of referral, all presentation of patients requiring revision from primary total hip arthroplasty in your centre over an almost 10 year period from 2005 to 2015. You detailed very nicely in the introduction to your paper, the increasing numbers of hip replacements being performed worldwide and how this is going to increase with time. And how important it is that healthcare providers adopted a cost-effective model for the long-term follow-up of these patients as well as the current national guidelines. So David, to start us off, [00:02:00] could you give us a brief background on the history of the paper including what the current guidelines are for total hip replacement follow-up currently?

Okay, Andrew. Yeah. So, first of all, 10 years ago, I would not consider the concept of discharging any primary joint after only one post-operative year. I did a fellowship with *inaudible* in 1988 and  learned the critical importance of long-term follow-up. 

As soon as I was appointed in Belfast in 1990, I started to do high volume joint replacement. I found myself with  very large review clinics, which were very enjoyable apart from the small number of very irritating patients who hadn't done well or who work unhappy. However, I began to realize that rather than being very irritating, these problem patients were the most important and needed more time than a new patient and not the lesser time that I was giving them in a busy review clinic. 

So in 1993, I set up what was called The Outcomes Unit and was joined by a young, enthusiastic nurse called Seamus OBrian, [00:03:00] who was one of the first arthroplasty care practitioners in the UK and Seamus and his team are still with me. So it happened more that Seamus saw all the post-op reviews on his own, and then sent me the ones that had problems.

So sadly since 1993, I've only seen unhappy patients. At the same time Seamus collected all the outcome scores and put them onto a database, whereas in reality most surgeons who review patients don't do so in a structured fashion. So it's often a missed opportunity to capture information. 

At that time, we saw all hips, irrespective of their age, six weeks, one year, five years and 10 years, and always with an AP and lateral x-ray. However once we got to the ten year review stage with a capacity issue. And because the five-year patients were doing well, we stopped doing five year checks. And then over the years, we had to be more selective about our ten year reviews and started just to see patients who were under 70 at the time of their operation. 

Then in recent [00:04:00] years, we've published our ten-year results for both our hips and knees and quite simply our revision rates for both hips and knees and specifically how those revisions presented didn't justify routine review of asymptomatic patients. I also have to say, this was a very gradual change in my mindset as it went totally against everything Ive been taught. 

Moving on to your next point then Andrew about the current BOA guidelines. As far as I'm aware the most recent guidelines are from 2012. And they stayed that particularly in patients of 65 and under 10, implants should be reviewed with an x-ray at one year, seven years, and every three years thereafter, if they're asymptomatic and if their x-rays are normal. Obviously the implication being that if they aren't, then they should be seen even more often than that.

And of course in this paper, that we're discussing, we suggest that such asymptomatic patients can be discharged at their first post-op review, [00:05:00] without x-ray. And I have to say the emphasis here is on  asymptomatic. 

It's also worth pointing out that in 2012, when the guidelines were written and ODEP 10A implant had a better than 90% chance of survival at 10 years was what we're now discussing is an ODEP, 10A star implant, which is a better than 95% chance of survival of 10 years. So our expectations have changed and perhaps our guidelines also need to change. 

That's interesting. That's a really great, great summary of how your practice has changed over the past few years. And just what the current recommendations are, but just a summary points, so what do you feel the key points are in terms of the arguments for and against routine follow-up then?

Okay, so I have no doubt that the big game changer has been cross-linked poly. I believe it's the biggest advance in PHA since Harry Craven reduced John Charnley to high density [00:06:00] polyethylene in 1961. 

The major issue was that was all-poly was likely to lead to massive bone loss in the asymptomatic patients. And that's just doesn't happen with crosslinked poly. It is quite simply in my view, the argument against routine follow-up is that crosslinked poly has removed the need in the  asymptomatic patients, if, and only if they have a 10A star implant. And therefore, again, in my view, the care argument for routine review is any patient with symptoms and any patient who doesn't have a 10A Star implant.

That's really interesting. That's great. So moving on to sort of how you, how you've done this study then. So obviously, you know, it is based on previously published data, isn't it? That's what it's built upon. So can you just give a brief background on that, on that study that the work is based upon.

Yes. So we published this work in BJJ with a paper entitled Impact of a [00:07:00] learning curve on the survivorship of 4,802 cementless total hip arthroplasties. This was a retrospective review of micro pinnacle implants over a 10-year period from 2005 to 15. The major learning point in that series was that certainly with the CRI you should always use a colour, essentially the colour prevented revision for early postoperative periprosthetic fracture. As you said with a total of 80 revisions in that cohort and for this paper that we're discussing today, we simply looked at primarily how and then when each revision presented. 

Okay. And so obviously you've alluded to it already, but you know, it's over a 10-year period and you sort of mentioned it in the paper that you're following over that period and how you've done that follow-up has changed a little bit over time. Is that fair to say? 

Yeah, so between August 05 and March 15, our standard follow-up for all THA patients with my practice included a five-day post-discharge phone call, clinical review at six weeks and clinical review with x-rays [00:08:00] at one year and 10 years. This was critically outside of the routine review procedure, patients also have had access to a helpline. This is staffed by very experienced arthroplasty care practitioners and it's worth pointing out again, the 45% of patients who were revised in this paper, self-referred using this facility of the helpline.

Yeah. That's very interesting. We actually have a similar setup here in Edinburgh. 

So in terms of, obviously we'll come on to the primary outcome, which is obviously revision in a minute, but can you sort of comment for the listeners on the obviously *inaudible* revision on the robustness of the data for, you know, there's a huge number of patients over a long time period and how sure you are that all the revisions were collected and that none were missed?

Yeah, obviously that's a very important and critical point, Andrew. As mentioned in the manuscript one of the limitations of the study is that the original cohort of patients were not reviewed at the time of publication. And this means that we didn't have a loss to follow-up group. And no information on [00:09:00] patients who may have been revised at other centres, however of the 975 patients who've subsequently been called for review at 10 years. Of those 975, 134 were deceased and that left 841, and we're able to make face-to-face or telephone contact with all but six of those patients. So in fact our true loss to follow-up was only 0.6%. 

Also as you know, Northern Ireland is a pretty small place. I work at Musgrave park, which is the biggest centre, but my other colleagues around the province know that I personally like to hear about problems and particular revisions. So I think it's very unlikely that we've missed any revisions. And even if we have missed a small number, it wouldn't really change the numbers or the message of this paper. 

No, I would agree with that, David. I think that small validation cohort you've done was very nice because it, you know, the chances are you've missed one if that, you know, looking at the paper as a whole [00:10:00] so that's, that's really interesting. 

So in terms of, obviously we said the primary outcome is revision, but I mean, I know you do, but there are other data that you collect as well. Do you collect PROM data normally for these patients as well? 

Yeah. So, as a routine we take an Oxford score on every  patient, pre-op and at six weeks. And another policy is that we send a postal review with the operative scores for patients at one year. I appreciate that that in England, that score is done at six months. We didn't put this data in either this paper or the previous one, but clearly PROMs data is important. 

Yeah, absolutely. Moving on to the results of paper then as you've, as you've alluded to the overall incidence of revision in the series was 1.7%. So, you know, 80 patients of the 4,802 that you treated. And the median time to revision was just shy of two years at 23 months. So David, can you just for the listeners describe in more detail the 80 cases that underwent revision and in particular why and [00:11:00] when they were undertaken? 

Okay. Andrew so yeah the reason for revisions in patients include instability, 27.5 so just over a quarter, infection at exactly a quarter so 25%, symptomatic loosening 18.8% and fracture at 7.5%. And those four things account for over 78% of all of our revisions. The 7.5% for early previously fracture revisions, it was higher in the early part of the series, but essentially it felt the zero when we started to use the CRI of the colour. The colour CRI, we were one fracture out of 4,000 cases. 

Right. 

And this isn't just me, published data from the NJR shows the colored CRI has the same revision profile as a cemented stem in the first year following surgery. So as you can see, not all cementless  or for that matter, cemented stems are the same.

Yeah. 

Again, its worthy to note that the [00:12:00] other causes of revision 6.3%, which is metallosis that occurred with ceramic on metal bearings in combination with cri pinnacle. As you perhaps know that we published our five years results of this ceramic and metal bearing in the BJJ in 2015. And at that time, the revision rate was definitely less than 10%, but we know now that when we publish our ten-year results, it'll be between 5 and 10%. It won't be as high as 10, but certainly therefore given the present definition of a 10A star implant, a cri pinnacle with varying combinations would not be a 10A star. So it's well worth emphasizing that it's not just the stem cup we have to think about. The bearing is also very critical. 

With regard to timing then you also asked about the timing, 38.8% of all revisions took place within the first year of which the majority were infection and instability, which is typical of other series as well. And interestingly, over half of our revisions for infections [00:13:00] were just a latter exchange and to date none of those have become infected again. And then overall 86.3% of revision took place within the first six years. 

Right. Yeah. So that's a majority, isn't it? So that sort of moves on nicely to figure one in the paper which looks at the modal referral for revision. And then you do an analysis of those results regarding the modal of referral and how this related to the timing of the revision post-index procedure. So can you just go in to a bit more detail regarding that David for us?

Okay, Andrew. Yeah. So again, the interesting thing here is that is that 45% of the revisions were as a result of self-referral. And almost half of those were for infection and it's worth pointing out that we really emphasize to our patients the importance of reporting and discharging to us and specifically not to the district nurse or GP. So we feel that we get our infections early and perhaps that's [00:14:00] why all our liner exchanges for infection are successful and we have no recurrence of infections in those cases to date. 

Next up then GP, which have a good number of patients with aseptic loosening. The majority of patients from other hospital referrals, which is instability. In patients were mainly for femoral fracture and emergency department referral patients were also many for instability as with the other hospital referrals. 

Two patients were re-admitted straight from home either infection or dislocation. And as you know, then there were two patients that were picked up at routine review, one for infection and one for liner association. But I really think that both of these would have self-referred if they haven't been seen at those times.

Okay.  Just briefly, what were the actual numbers again? So in terms of the, for the hospital referrals and the GP referrals, what were the most common causes and what were the numbers for those David? 

Yeah. So the majority of other hospital referrals were for instability or just occasional. It was 10 of the 13. [00:15:00] Whereas half of the GP referrals were ultimately for aseptic loosening.

So that's sort of what, you're almost what you'd expect intuitively isn't it? Yeah. Okay, great. So sort of, sort of moving on to summing up the paper and then the context of the current literature, you know, the strengths of the study, I think are without question. There is a large number of patients from a single centre with previously published data in other areas and strengths that are out for that question. And it's clearly, I think, provided important information regarding the ongoing debate and planning to the future with regards to how we carry out the long-term followup of our hip arthroplasty patients. So, David how do you feel the data fits into the current literature as well as the current guidelines? 

Okay, Andrew, I think our data fits well with other published data on reasons for revision from a number of authors and in particular from NJR. Clearly though, as you can see our policy has holes with current guidelines. But I think it has to [00:16:00] be said that very few surgeons in the UK actually adhere to these current guidelines. So, what I'm hoping is that this paper will provoke debate on our surgical community on this subject. 

That's right. And I think that's what we often hope of with a lot of these papers is that I think it's creating that debate and discussion isn't it? So that we can actually move things forward. I suppose one of the things that some people might say is, you know, PROM data or patient satisfaction is that as important or more important than using revision and, you know, does that need to be considered what's your view on that David? 

Yeah, again, a very important question. But I do really feel that revision is important because it does give us a very good objective measure. I think in general, as surgeons, our expectations have been too low. On April 14 that 10A star weighting was produced with a less than 5% revision by the 10 years, as opposed to the less than 10% that [00:17:00] had been replaced by 10A implant before 2014. 

I think today we would all agree that 10% revision rate would be considered totally unacceptable. And although the 10A star implant is less than 5% I think that many of us would also feel that 5% is too high. And as you know, at the moment in the NJR the best performing implants have a revision rate of less than 2.5% in 10 years. 

Yeah. 

I predict that by 2024, which is 10 years on from the last  revision, the 10A star implant will become 2.5% of the 10 years and perhaps they'll give it two stars. By then I, again predict that the best performing implants will have a revision with a 1% of 10 years. That has to be good for our patients. 

And also if our best performing implants are getting towards 1%, that will really focus attention on patient outcome, other than revision. But I think it needs to be more objective. [00:18:00] I think we start to need more than just PROMs. I think things like Fitbit technology and a more simple accessible TKA analysis, I believe will become more important. 

And in reality, our surgery is still very crude. Our implants are manufactured to very high specification using robots. And then we use a hammer to put them in. I think we need to define ideal targets for cup placement or stem placement. And in fact, we're about to submit data showing that hip centre and where you put them actually impacts on patients outcomes in terms of pain and satisfaction. 

And certainly our patients aren't always as good as we think they are. For example, we're all well aware that knees tend to be less happy than hips. This has been well published. And for me, of my knees, 18% or one in five have moderate to  severe pain in  one [00:19:00] year. So that's a really big number, whereas it's half that figure for my hips. So it's only 9% of my hips have moderate to severe pain at one year.

However, interestingly by 10 years for my knees, the moderate to severe pain has dropped to 10%. But my hips at 10 years, 9% of them still have moderate to severe pain. And again, I believe our expectations here are too low and I think we need to do better than 1 in 10 of our patients having moderate to severe pain at 10 years. And I'm hoping that's not just me, but I think that's probably a fairly general thing for all our patients. 

That's really interesting. That's really interesting predictions for the future as well. And in terms of where we go moving forward in terms of how we assess the patients and outcome really. And those are really interesting figures about, you know, the rate of pain afterwards and how it sort of, comes down to similar levels of the total hips, but they haven't changed. That's really interesting. 

So in terms of just to finish off [00:20:00] David, you know, what do you think are the main take home messages of the paper, but also, you know, what the are caveats related to the limitations of the study do you feel? 

Yeah, so as we've said, I think our follow-up is good. I don't think we've missed many revisions. So for me, the key limitations are that this relates to one surgeon's practice and just the one implant. However, as I've said, my intention, the hope is that this provokes discussion. And so the take home message, I think, is we need to consider whether or not we need to review a 10A star implants.

 Yeah.

But the key thing is the symptomatic patients made an easy pathway back into our systems through self-referral. And the sad thing is this often isn't the case. And this is perhaps where the major challenge lies. And this is irrespective of whether you have a long-term review programme in your hospital. 

The problem is that these review programmes usually relate to an individual consultant. [00:21:00] And I believe there's a cultural problem among surgeons. I think maybe across the world where if you like there is no long-term corporate or shared responsibility for patients with symptoms. Often their original surgeons have retired or moved on. So I believe the model of an experienced arthroplasty care practitioner team is a very good one. There's no reason why not that team cannot  see the symptomatic patients quickly and then triage as to firstly who the patients to see. And secondly, and perhaps above all, how urgently they should be seen.

Yeah. And that sums up very nicely. So,  yeah thank you so much. That was a really interesting discussion. And thank you for joining us for our podcast and congratulations on an excellent study. 

Okay. Thak you Andrew very much. 

Thank you. So to our listeners, we do hope you enjoy joining us, and we encourage you to share your thoughts and comments through Twitter, Facebook, and a like, and feel free to post or tweet about anything we have discussed here with you today. And thanks again for joining us.