BJJ Podcasts

The changing incidence of arthroscopic subacromial decompression in Scotland

April 14, 2020 The Bone & Joint Journal Episode 21
BJJ Podcasts
The changing incidence of arthroscopic subacromial decompression in Scotland
Show Notes Transcript

Listen to Mr Andrew Duckworth interview Dr Paul Jenkins about his paper 'The changing incidence of arthroscopic subacromial decompression in Scotland', published in the March 2020 issue of The Bone and Joint Journal.

Click here to read the article

[00:00:00] Welcome everyone to our BJJ Podcast for the month of March. I am Andrew Duckworth and a warm welcome from your team here at The Bone & Joint Journal. As always, we'd like to thank our readers and listeners for the comments and support we receive as well as to our many authors and guest interviews who have taken part so far.

Over this year, we hope to continue to build on the range of topics we've covered through the series so far with our continuing aim to improve the accessibility and visibility of the subjects we publish for both you as our readers and listeners, as well as for our many authors.

For this months study, as you know of the next 15 to 20 minutes or so we will cover a range of aspects for the chosen work, emphasizing the important points of how the study has been designed, as well as the key findings from the data and how these potentially fit into each of your day-to-day clinical practices. We also hope to 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 Mr. Paul Jenkins from Glasgow Royal infirmary to discuss their study entitled The changing instance of arthroscopic subacromial decompression in Scotland, which has been published in the [00:01:00] March edition of The BJJ. Welcome Paul and a big thank you for taking the time to join us today.

Thank you very much, Andrew, for the invite to take part in this podcast. 

So Paul, the aim of your study was to examine the recent trends in delivery of arthroscopic subacromial decompression in Scotland and to determine if this varies by geographical locations throughout the country. So can you give us a brief introduction to the paper and some background to the recent studies highlighting the chronic tendinopathic model of dysfunction for subacromial pain.

So I think as you say this paper was rooted very much in the emerging concepts of the pathogenesis of tendinopathy, affecting the whole broad spectrum of tendinopathy throughout the body. In the initial sort of evolution of tendinopathy, it was sort of first described really around the shoulder by Codman in the 1930s, as a result of intrinsic degeneration inside the tendon. And it was really only later on that Neer described the extrinsic [00:02:00] model, coining the term impingement with impingement sort of structures, such as the hooped aspect of the acromion and the CA ligament leading to rubbing and damage. And it was these concepts that then led to the surgical management by taking them away first with open procedures. And then with the advent of arthroscopy and the advent of arthroscopy was very seductive in that you could take a procedure which was associated with reasonable morbidity in terms of its open setting and undertake it with a very small incision, under much lower morbidity associated with that.

On the back of that, we definitely saw a really rapid expansion in the adoption of subacromial decompression as a technique. And that was described by Professor Carr and his group in a paper in The Bone & Joint Journal in 2014, which [00:03:00] described the temporal trends and geographical variations in England over a period of 2001 to 2010. And they reported a seven fold increase from 5.2 per a hundred thousand to just over 40 per hundred thousand. 

Following on from that, there were two really quite influential, large, prospective pragmatic multicenter, randomized controlled trials. The first was UK based, which was the Seesaw study - can shoulder arthroscopy work? Undertaken again by Professor Carr's group as the chief investigator and published in the Lancet in 2018, but had been published online the previous year. And similarly there was a study from Finland in the BMJ which looked at quite similar study design. 

The [00:04:00] Seesaw study randomized patients into three groups. The first was no intervention at all. And then there were two surgical groups. The first surgical group was children arthroscopy alone which was designed as a placebo intervention group although I think we have to remember that within that group patients did undergo a diagnostic arthroscopy and both the joint and the bursa were instrumented. And it had sealing insufflation and I think has many of the listeners would agree, it can often be very difficult to see at all in the bursa without removal of at least a little bit of the bursa and therefore, you know, possibly a partial vasectomy. And these are points which can be discussed in terms of how we interpret the results.

The Seesaw study showed that both diagnostic, arthroscopy alone and the actual full subacromial decompression and removal of these hypothetical [00:05:00] extrinsic compression structures were both superior to no intervention in terms of the study end points with the Oxford shoulder score, but  there was really equivalents between them in terms of their efficacy. And the conclusion was that the placebo surgery group was as effective as the actual surgery group.

The IMPACT study was slightly different in the patients were initially randomized to either physiotherapy or surgery. And then within the surgery group, they were randomized a second time to the sort of placebo diagnostic arthroscopy or the arthroscopic subacromial decompression group. 

Both these studies really started to be widely discussed amongst the shoulder community. And along with the evolving [00:06:00] concepts of the pathogenesis of tendinopathy as being not a purely extrinsic issue, led the shoulder community to start to really think about the indications again for subacromial decompression and where it was used. And we wanted to look at a national database to see, was there an improvement in, or was there a change in the provisional of this procedure around the time these studies started to be published. 

That's a great overview, Paul, I think just sort of bringing together sort of the background to an SAD and also the recent studies. And I think, like you say your work is very timely in determining certainly in Scotland, how the literature has influenced that. 

So if we move on to how you performed the study. You analyzed national temporal trends in procedures, utilizing the National Scottish Mobility Records or SMR for the period between March, 2014 to April, 2018. So for [00:07:00] our listeners, Paul, who won't be familiar with the SMR, what does that database normally collect and sort of how robust or how accurate is it? 

Well, this database has been run for a long time, stretching back to the early 1980s. And it is run centrally by the NHS Scotland information and statistics division, and every episode within Scottish hospitals, either in the day-case or an inpatient generates an SMR or one code and SMR stands for the Scottish Morbidity Record. And this is very similar to the English HES type data. The data is generated by a hospital quarters who at the end of an episode, will look through the final discharge letters along with your operation notes and any other information in the patient's notes to make a code about the underlying diagnosis with the ICD 10 and the procedure undertaken with [00:08:00] OPCS code.

It's a very useful dataset. And in other settings, because of the use of the community health index number, which is unique to every patient, we can also look at sort of linkage studies and look at linkage to other national databases, but this was a relatively simple and straight forward query  asked of the database just looking at one sort of set of procedures over a period of time.

In terms of accuracy, there have been studies about the accuracy and it has been reported as being an average of 90 to 94% when it has been validated. With any of these sort of databases, however, we have to accept that there will be a degree of miscoding. That was one of the reasons in terms of the time periods we chose, because earlier versions of that will be OPCS coding system, there was [00:09:00] less clarity about the coding of subacromial decompression. Whereas from 2014 onwards, there were specific codes relating to that as a procedure. And that was also again, one of the reasons we didn't look at sort of longer term trends prior to that. We felt that it would be introducing some more potential for *inaudible* that we couldn't take account of.

Also in terms of if there were going to be some errors in coding, which we would accept, we have assumed that they would be consistent during the period of time. And so we're looking at this in terms of internal consistency and we would have some more cautions when extrapolating it to other work. Although it is interesting to note that the incidence of subacromial decompression at the start of hours was in the same region as 40 per a [00:10:00] hundred thousand as had been described at the end of Professor Carr's study in England in 2014. So I think that actually is quite a nice way of  linking those two studies and looking towards the validity of both these datasets. 

No, I totally agree, Paul that's right. And I think the *inaudible* of it is quite clear there and like you say, with using big data like this, there may be some inconsistency there, but they're likely going to cancel themselves themselves out over time.

And before we move on to the results, though, just for our listeners, what were the inclusion and exclusion criteria you had for the study in particular? 

So we looked at in general, we looked at patients who had the specific codes of arthroscopic subacromial decompression. And then we were looking at those who had had an arthroscopic AC joint excision alongside that. We excluded any open forms of AC joint [00:11:00] excision because one suggested hypothesis was that as the ability or the subacromial decompression became a less common procedure surgeons may look to treat other pathology of questionable symptomatology. You know, in some cases the AC joint is fairly much symptomatic in some other cases there may be radiological changes, but it's not pervading sort of symptomatic cost. So we wanted to look at this as a balancing measure to see what was the rate of other arthroscopic shoulder procedures changing over that same time period. 

No. Absolutely. Yeah. I think that's very clear in terms of, I think the exclusion criteria, like you say, just add to the validity of the way you've performed the study. I think that's very clear. So in terms of, if we move onto the results, shall we say, so what were the key findings? You sort of alluded to what it was like at the start. What would you feel the key findings are of your study in relation to the [00:12:00] number of ASTs performed over the study period and how that changed?

I think the key finding was that the rate fell by approximately 35% and that the majority of this fall was seen in the year 2018. These were financial years so that would actually be from april, 2017 to March, 2018. That fell from 41.6 per hundred thousand to 28.9 per a hundred thousand. That again, validates our hypothesis that some of this fall was due to the ongoing work around tendinopathy in these multicenter randomized controlled trials. Now one issue that was brought up during the review process was in terms of the timing of the publication of the RCTs versus what we have observed as this fall.

[00:13:00] Although we have the dates of publication of these RCTs as 2018, they were being discussed and in some cases were discussed in the wider media and then a paper such as The Guardian well before that. And quite a lot of surgeons who had been involved in the process and actually randomized patients are starting to get a feel for the fact that there may well be sort of limitations of the subacromial decompression. So the feeling was that even prior to the publication of the results, there was a, an increasing awareness in the shoulder community of the possible outcomes of the study. And I think actually there was an anticipation and practice did start to change in the year before that. 

No, I think that's right. I mean, certainly that was similar for, as we all know, for the draft study that looked at wrist fractures, the change that started to occur before the study was published. But for [00:14:00] those very reasons you just described, you often get that change happening a little bit before the actual official publication date for those reasons.

So, we've talked about the reduction, but in terms of the geographical variations across Scotland, did you see anything when you looked at that Paul? 

This again was very interesting. And it did actually, again, back up some of the work that's been done by Prof Carr in the 2014 paper where we saw quite significant variation between boards throughout the country. In the year 2017 to 2018 of the 14 boards in Scotland, there were four where the rate of subacromial decompression was over three standard deviations from national average. And there were two where the rate was below that national average.

 When we look at these numbers involved. Some of this could be done to the practice of perhaps one or two different surgeons or units. And in terms of the actual [00:15:00] criteria that people use to decide whether to recommend to a patient, whether to undertake some subacromial decompression or not there can be variation in terms of the amount of time, the number of steroid injections, the amount of physiotherapy that should be undertaken prior to that. 

We would plan to use this sort of data separately to feed back to the individual boards areas and,  as an area of interest for future research to see whether these outlying areas do come within a more sort of national average within the *inaudible* once that is fed back.

 Obviously there are also areas of good practice on show here, but you also have to balance that in terms of when a rate is very low, is there actually restriction in terms of an effect of procedure. And I think in [00:16:00] all of our discussion, we have to realize that there still is a role for this procedure and that,  the danger of some of these studies and some of this work can be for a very binary approach to the problem where we say this operation doesn't work. It shouldn't be offered. And even when we look back at the Seesaw study, there was a approximately 25% of the non-operative group who crossed over into an operative intervention or over this study team period. So I think we do need a more nuanced approach and recommend that it is still there  possibly, bu as a procedure that may be beneficial, but when conservative measures are exhausted after a reasonable amount of time.

No, I completely agree, Paul. I think that's the key for many of these trials is that we sort of want a binary outcome in a one size fits all sort of answer to it. But often I don't think you can do that. [00:17:00] I think it's for more generalizability and also I think in this case, as you say, it's sort of saying that this is procedure we were probably overdoing but there are a small number of patients who will still benefit from this once they fit certain criteria.

So that sort of moves on nicely, Paula. If you had to summarize, what do you feel the key findings of the work are? And I suppose you've already discussed it briefly, but sort of with any caveats, with any limitations of the data. 

I think if we start with the caveats, this looks at, sort of large national level data. There is a potential for the data being slightly erroneous or incomplete though we are reassured by the published validity or validation studies of the datasets being in the region of the 90%. We're relatively comfortable that the trends over time will be valid in terms of any [00:18:00] errors will be consistent over time because there hasn't been any significant changes to the coding during this time period or  the overall methodology. 

What we don't have access to is the number of these procedures that are undertaken in the private sector. We would actually see that one of the potential benefits of undertaking this study in Scotland. In Scotland the private sector really makes up quite a low proportion of the overall orthopedic procedures done within the country on an annual basis. And, I think it makes it probably more applicable or comparable to the national population, although it would be of interest what is happening in other, in other sectors.

Absolutely. In terms of Paul, we've already [00:19:00] discussed that a little bit, but you mentioned the greatest decline was seen 2017 to 2018. We think that's potentially related to the literature. And so the lag time with publication, we've talked about that. What do you feel the explanations are in terms of the variations you saw across Scotland? Do you think are those are variations seen elsewhere in other parts of the literature and do you have any sort of explanations for that in particular? 

I think we see this sort of variation really throughout the literature. And when we look at things like the English Atlas of variation or the NHS  Scotland Atlas of Variation, we see such unexplained variation in the provision of a variety of procedures and other forms of healthcare, you know, from arthroplasty to cataract surgery and understanding this variation is probably key to trying to improve the provision of healthcare relief throughout [00:20:00] systems, as large as the national health service and trying to filter down the recommendations and getting evidence-based medicine into practice is the challenge and that there are a lot of studies coming out like this and trying to get them down to the actual people who are undertaking the procedures in to effect change is probably one of the most challenging aspects of our sort of practice and evidence-based medicine now going forward. 

Well, I totally agree, Paul, I think it's sort of, when you're talking about going forward, you know, relating to your study, how do you feel that the data of these types just studies sort of reflect on what our practices are on a national level? How would you best move them forward? How can you utilize them to sort of influence our day to day individual practices? 

I think as static pieces of work, these are interesting, but I think there needs to be ongoing [00:21:00] audit and review and feedback to surgical teams. We have access to national tools such as the National Joint Registry and the Scottish arthroplasty project. When it comes to arthroplasty I think we need to have extension of these projects too cover non-arthoplasty surgery because actually there is probably the majority of procedures undertaken in orthopedics are not arthoplasty related. And I think focus needs to come to them so that we can look at where variation exists to try and understand that and reduce it.

No. I totally agree. That's exactly right. So,  I think that's all we have time for actually. So, thank you so much for taking the time to join us today and congratulations on an excellent study that I think is without doubt an invaluable addition to literature in this area. And again, thanks for taking the time to join us.

Thanks again for the invite. It was a pleasure to join you.

 And to our [00:22:00] listeners, we do hope you've enjoyed joining us today and we encourage you to share your thoughts and comments through Twitter, Facebook, and a like. Feel free to post a tweet about anything we have discussed here today.  And thanks again for joining us.