Latest / MSKMag OutLoud / Shoulder instability. It's all getting a bit silly

Transcript

Shoulder instability. It's all getting a bit silly · MSKMag OutLoud

0:00
SPEAKER_00: Shoulder instability.
0:02
SPEAKER_00: It's all getting a bit silly.
0:05
SPEAKER_00: Written by Fraser Philp.
0:07
SPEAKER_00: Innovation, ignorance, or insanity?
0:11
SPEAKER_00: How many shoulder classification systems are enough?
0:14
SPEAKER_00: Or rather, how many shoulder classification systems are too many?
0:17
SPEAKER_00: 5 sounds fair.
0:19
SPEAKER_00: 10 seems a little bit too much.
0:21
SPEAKER_00: And 18 would be ridiculous, right?
0:24
SPEAKER_00: Right?
0:26
SPEAKER_00: Well, folks, hold on to your shoulders.
0:28
SPEAKER_00: In 2011, that's 15 years ago, a systematic review identified 18.
0:33
SPEAKER_00: Yes,
0:34
18 published classification systems in the existing literature
0:37
before they went on
0:38
SPEAKER_00: to develop their own.
0:39
SPEAKER_00: So 19 then.
0:42
SPEAKER_00: Over the years, that number has continued to rise.
0:44
Has this abundance of classification systems helped to improve
0:47
the lives of people
0:48
SPEAKER_00: with shoulder instability?
0:49
SPEAKER_00: No.
0:50
Have the proposed rehabilitation regimes and treatment plans
0:53
provided definitive
0:54
SPEAKER_00: answers on the best way to help all our patients?
0:57
SPEAKER_00: Not quite yet.
0:58
SPEAKER_00: More data needed.
0:59
SPEAKER_00: Do I have all the answers?
1:01
SPEAKER_00: No.
1:02
SPEAKER_00: Am I having a go?
1:03
SPEAKER_00: Definitely not.
1:05
SPEAKER_00: Shoulder instability is a small world.
1:07
SPEAKER_00: My experience so far in this space has been positive.
1:10
SPEAKER_00: This is because those who I've engaged with have been welcoming,
1:13
SPEAKER_00: receptive to new ideas,
1:14
SPEAKER_00: giving of their time and willing to engage in the conversations.
1:18
SPEAKER_00: Thanks to all if I've not said it so already.
1:21
SPEAKER_00: They are often very experienced and knowledgeable clinicians,
1:24
all of whom are trying to do what's best for people with
1:26
shoulder instability.
1:28
It's important to remember that a lot of the concepts critiqued
1:30
here came about as
1:32
SPEAKER_00: a result of trying to make sense of the chaos.
1:35
SPEAKER_00: They were the original ideas,
1:36
SPEAKER_00: innovations,
1:37
fledgling hypotheses and the foundations from which other
1:40
concepts could develop
1:41
SPEAKER_00: and compare themselves.
1:43
SPEAKER_00: Standing on the shoulders of giants and all that.
1:46
History tells us that our initial ideas of how the world works
1:49
aren't always
1:50
SPEAKER_00: correct the first time around.
1:52
Think flat earth theory,
1:53
milk blood transfusions,
1:55
and cocaine for hay fever.
1:57
SPEAKER_00: We need to continually revisit our assumptions and understanding.
2:00
SPEAKER_00: It's difficult to do this when there is no new or reliable information.
2:04
SPEAKER_00: It seems like every time I get online,
2:06
someone somewhere has published a new shoulder instability
2:08
classification system,
2:10
SPEAKER_00: rehabilitation plan or course.
2:13
SPEAKER_00: I'm not against that.
2:14
SPEAKER_00: This is a complicated area.
2:15
SPEAKER_00: Anything that brings more clarity and certainty can only be a good thing.
2:18
SPEAKER_00: But we have to be honest about the state of our clinical practice,
2:22
the evidence that drives it and whether these growing
2:24
contributions are actually
2:26
SPEAKER_00: helping or even making things worse.
2:28
SPEAKER_00: Is the information really new or is it fit for purpose?
2:32
To help answer this,
2:33
in this article I want to put forward three main points.
2:36
One,
2:37
existing classification systems are conceptually useful but
2:39
practically useless.
2:41
SPEAKER_00: Two,
2:42
SPEAKER_00: research in this area is thin,
2:44
SPEAKER_00: not well joined up and does not follow the life of the patient.
2:48
Three,
2:48
we seem to be repeating the same mistakes and it's time to try
2:52
something different.
2:53
SPEAKER_00: Point one is the longest.
2:54
SPEAKER_00: If you make it through that, you're on the home straight.
2:58
SPEAKER_00: Why is this important?
3:00
SPEAKER_00: There is a phenomenon in artificial intelligence,
3:03
SPEAKER_00: AI,
3:03
SPEAKER_00: known as iterative degradation or quality decay.
3:06
You may have seen funny examples of it on the internet where
3:09
chat GPT is asked to
3:10
SPEAKER_00: create exact replicas of a picture over and over again.
3:13
SPEAKER_00: The final result ends up worse than the original.
3:16
The cause of the problem comes from the fact that every time a
3:19
new updated version
3:20
SPEAKER_00: is fed back into the model,
3:21
SPEAKER_00: small errors and noisy data get larger.
3:27
SPEAKER_00: Without new or accurate information to correct itself against,
3:30
SPEAKER_00: the system just gradually becomes a parody of its former self.
3:33
SPEAKER_00: Can you see where I'm going with this?
3:35
At the patient level,
3:36
this potentially results in them not getting the best care.
3:39
Misdiagnosis and wrongfully attributing things as the cause of
3:44
their problems may
3:45
SPEAKER_00: result in delays or withholding of needed investigations or treatments.
3:49
SPEAKER_00: It may also result in unnecessary or ineffective treatments being used,
3:53
all of which make the overall health care burden on people
3:55
larger and longer.
3:57
SPEAKER_00: At the therapist level,
3:58
SPEAKER_00: it makes entering the world of shoulder instability seem overly complex.
4:02
SPEAKER_00: As an exercise,
4:03
SPEAKER_00: I always ask myself,
4:05
if I was a student trying to learn about this afresh and all I
4:08
had available to me
4:09
SPEAKER_00: was the published evidence,
4:10
SPEAKER_00: would I arrive at the same conclusions as someone else?
4:13
SPEAKER_00: Is the data or evidence clear enough?
4:15
If I had to pick which classification system or treatment plan
4:18
would I use and why?
4:20
SPEAKER_00: How much of current practice is product of tacit knowledge,
4:23
SPEAKER_00: departmental or institutional norms?
4:26
SPEAKER_00: More recently,
4:26
SPEAKER_00: social media,
4:27
course and masterclass conference presentations are playing a
4:30
larger role,
4:30
SPEAKER_00: and so we need to make sure we can help people discern the truth.
4:34
SPEAKER_00: So here's my attempt at doing that.
4:36
One,
4:37
existing classification systems are conceptually useful,
4:40
but practically useless.
4:43
The purpose of classification is to identify distinct groups to
4:47
which people or
4:48
SPEAKER_00: things can be assigned on the basis of some predetermined measures or
4:51
SPEAKER_00: characteristics.
4:52
SPEAKER_00: For example,
4:54
SPEAKER_00: in figure two,
4:54
SPEAKER_00: which you can see online,
4:55
SPEAKER_00: you could classify Lego blocks by their color or their shape.
5:00
These features,
5:01
shape and color,
5:01
are practical,
5:02
cheap and easy to observe or measure.
5:05
You could argue that shape would be more robust method for
5:08
classifying Lego blocks.
5:09
SPEAKER_00: This is because misclassification may occur by people who are colorblind.
5:14
SPEAKER_00: And there's another figure, three.
5:17
SPEAKER_00: The main point here,
5:18
SPEAKER_00: however,
5:18
SPEAKER_00: is that any object could be classified in multiple ways,
5:20
SPEAKER_00: depending on the measure,
5:21
SPEAKER_00: feature or combination of measures and features used.
5:25
So figure 3 is a picture of a variety of fruit and vegetables
5:31
in colours and stuff.
5:34
Classification systems are therefore only as good as the
5:37
measures on which they are based.
5:39
SPEAKER_00: If the measures or features aren't reliable,
5:42
SPEAKER_00: are subject to interpretation or can't be quantified,
5:44
SPEAKER_00: it may result in misclassification.
5:47
SPEAKER_00: We also need to consider,
5:48
did we make the classification systems and then assign measures
5:51
we think should go
5:53
SPEAKER_00: best with them,
5:55
SPEAKER_00: top down?
5:56
Or did we let the data and its features objectively and without
5:59
bias tell us if
6:00
SPEAKER_00: there are actually distinct or overlapping groups,
6:03
SPEAKER_00: bottom up?
6:04
SPEAKER_00: My feeling is that in healthcare, we too often do the former, top down.
6:08
Another thing to consider is what purpose does the
6:10
classification system we are using serve?
6:13
Is it a conceptual and theoretical model to help us make sense
6:17
of an incomplete
6:18
SPEAKER_00: understanding of physiological or biomechanical processes?
6:22
Or is it used as a definitive treatment algorithm based on
6:26
objectively quantified
6:27
SPEAKER_00: measures or tests?
6:29
SPEAKER_00: Are we guilty of confusing the two?
6:31
SPEAKER_00: Do we erroneously give equal weight to both when we shouldn't?
6:35
SPEAKER_00: So what is actually needed for a classification system to be any good?
6:39
SPEAKER_00: 1.
6:40
SPEAKER_00: The end treatment, investigation or management plan has to be different.
6:44
SPEAKER_00: If all people,
6:44
SPEAKER_00: irrespective of category,
6:46
SPEAKER_00: get the same investigations or treatments,
6:47
SPEAKER_00: is the classification serving a purpose?
6:50
SPEAKER_00: 2.
6:50
SPEAKER_00: It has to put people in distinct groups, i.e.
6:53
SPEAKER_00: it has to be able to discriminate. 3.
6:56
SPEAKER_00: What if a person can simultaneously exist across all possible categories,
6:59
SPEAKER_00: each of which has a different investigation or treatment endpoint?
7:03
Does the classification system have enough discriminatory
7:06
ability to be considered
7:07
SPEAKER_00: suitable for practice?
7:09
SPEAKER_00: 3.
7:09
SPEAKER_00: It needs to be reliable and accurate measurements
7:13
People need to agree on the way these measurements are used to
7:15
combine or group people.
7:17
SPEAKER_00: So do existing shoulder instability clinical classification systems work?
7:21
SPEAKER_00: Not really,
7:22
SPEAKER_00: especially when compared against imaging or surgery,
7:24
SPEAKER_00: which is usually considered the gold standard.
7:27
SPEAKER_00: Here are some quick stats.
7:29
SPEAKER_00: Maroda et al.
7:31
SPEAKER_00: 2020 reported that multidirectional instability was overdiagnosed.
7:34
SPEAKER_00: 10 to 20% of patients had bony changes in their shoulder,
7:37
SPEAKER_00: despite the cause of their shoulder problem being a
7:41
SPEAKER_00: muscle coordination issue.
7:44
SPEAKER_00: Jaggi et al.
7:46
found that 10% of participants were later found to be
7:50
unsuitable for the study
7:51
SPEAKER_00: after an arthroscopy showed no capsulolabral damage or bony injury.
7:58
SPEAKER_00: Clark et al.
8:00
summarised nicely by Adrian Davies also shows an approximate
8:04
error of 20% as well
8:05
when it comes to confirming the direction of instability in
8:07
rugby players.
8:09
SPEAKER_00: Note,
8:09
SPEAKER_00: I have also purposely mixed traumatic and atraumatic instability and go
8:13
SPEAKER_00: interchangeably at times between them.
8:15
SPEAKER_00: The misclassification problem is a reason for this.
8:17
SPEAKER_00: When I refer to shoulder instability,
8:19
SPEAKER_00: I am talking about the less obvious cases,
8:21
SPEAKER_00: although the concept could be extended to any subgroup.
8:25
SPEAKER_00: Now,
8:25
SPEAKER_00: granted,
8:25
there are some barn door obvious cases which arguably don't
8:28
need a classification
8:29
SPEAKER_00: system,
8:30
SPEAKER_00: e.g.
8:31
SPEAKER_00: some traumatic dislocations accompanied by imaging,
8:33
SPEAKER_00: but as a clinician,
8:34
SPEAKER_00: I'm usually more interested in the ones we don't get right or unclear,
8:37
SPEAKER_00: rather than the ones that are obvious.
8:39
SPEAKER_00: It appears we are getting it wrong for about one in every 10,
8:43
SPEAKER_00: if you want to be optimistic,
8:44
SPEAKER_00: or one in five patients.
8:46
SPEAKER_00: Is that good enough?
8:48
SPEAKER_00: Why is this happening?
8:49
SPEAKER_00: What is it that existing classification systems aren't telling us?
8:53
To help illustrate some of the points,
8:55
I'm going to use the Stanmore Triangle.
8:57
The Stanmore Triangle frames shoulder instability as a dynamic
9:01
or shifting
9:01
SPEAKER_00: condition existing between three poles and five states.
9:04
SPEAKER_00: Structural, non-structural, traumatic, atraumatic and muscle patterning.
9:08
SPEAKER_00: Can you see why I said this might be overly complex?
9:11
The idea is that people can move along these poles between
9:14
states with varying
9:16
SPEAKER_00: levels of each state,
9:17
basically along the lines of the triangle and across to other
9:19
corners or poles.
9:23
SPEAKER_00: Seems like a handy framework, right?
9:25
SPEAKER_00: But what happens when we get into the nuts and bolts of it?
9:27
SPEAKER_00: Can it work practically?
9:29
Does it do all of those things that are needed for a
9:32
classification system to be any good?
9:34
SPEAKER_00: Does it put people in mutually exclusive categories?
9:37
If the management options are different but you can't clearly
9:39
identify who needs
9:40
SPEAKER_00: what accurately,
9:41
SPEAKER_00: then what purpose does it serve?
9:43
SPEAKER_00: It took me a while to recognise where else I had seen this model before.
9:47
SPEAKER_00: It's the same one used in theology to depict the Holy Trinity,
9:50
and it's been keeping theologians busy for thousands of years
10:04
as they depict,
10:05
SPEAKER_00: without committing heresy,
10:06
an entity that is simultaneously all things at once and yet
10:10
three distinctly
10:11
SPEAKER_00: different things at the same time.
10:13
SPEAKER_00: Somehow, we've decided it's a good model for shoulder instability.
10:16
I've also seen people use triangle models to describe states of
10:20
selective emotions
10:23
SPEAKER_00: for an upcoming event.
10:25
SPEAKER_00: A triangle model in this case also makes sense.
10:27
Having a combination of different emotions and transitioning
10:30
between them regularly
10:31
SPEAKER_00: seems a conceptually helpful way to illustrate how someone feels.
10:37
SPEAKER_00: So it seems triangle models are helpful for concepts,
10:40
SPEAKER_00: but not necessarily clinical decision making.
10:42
SPEAKER_00: At the end of the day, we are therapists and not theologians.
10:46
SPEAKER_00: Oh, Fraser.
10:47
SPEAKER_00: Theologians.
10:49
SPEAKER_00: Practitioners, not philosophers.
10:52
SPEAKER_00: There's a few more things we need to consider.
10:55
SPEAKER_00: Irrespective of the shape,
10:57
although I have seen someone propose the cube of shoulder
11:00
instability to try and
11:01
SPEAKER_00: resolve this,
11:02
the measures on which the classification is based are arguably
11:06
the most important.
11:07
SPEAKER_00: For example, what does muscle patterning actually mean?
11:10
SPEAKER_00: Are we talking about a coordination issue?
11:13
SPEAKER_00: How would we detect this?
11:15
What if the muscle patterning is a consequence of changes to
11:18
the underlying
11:18
SPEAKER_00: structure rather than the cause?
11:20
SPEAKER_00: As things stand,
11:21
there's not enough high quality evidence based on the state of
11:24
the art 3D and
11:25
surface electromyography measures to differentiate between
11:29
movement patterns that
11:30
SPEAKER_00: could be considered.
11:32
SPEAKER_00: Normal variation within a population,
11:36
different but an adaptation for maintaining stability based on
11:39
changes to the joint
11:40
SPEAKER_00: from instability,
11:41
SPEAKER_00: and different and is resulting slash causing instability.
11:46
we can't yet make this distinction with equipment that allows
11:48
us to go back and
11:49
look at the different joints and muscles at a pace we can
11:52
comprehend and revisit
11:53
how then do we can we do this in real time with our eyeballs
11:57
during clinic there
11:59
are of course other classification systems available the
12:02
frequency etiology
12:03
direction severity feds system is a helpful way of
12:07
standardizing the description of
12:08
SPEAKER_00: someone's instability
12:10
SPEAKER_00: Again, it seems pretty straightforward on the surface.
12:12
SPEAKER_00: However,
12:12
SPEAKER_00: when you look a bit more closely,
12:13
SPEAKER_00: there are a lot of ways,
12:14
SPEAKER_00: 30 different combinations,
12:17
SPEAKER_00: that people can be classified.
12:18
SPEAKER_00: It also doesn't really tell us about the primary mechanism.
12:21
SPEAKER_00: Etiology is discussed in relation to being traumatic or atraumatic,
12:25
SPEAKER_00: but with no reference to what role bony,
12:27
SPEAKER_00: muscular or soft tissue structures may play.
12:31
A lot of the models also do not explicitly mention what role
12:35
psychosocial factors play.
12:37
SPEAKER_00: As a result,
12:38
it seems people will borrow categories or attributes from other
12:40
models creating a
12:41
sort of Frankenstein's monster classification system that may
12:45
have negative
12:45
SPEAKER_00: downstream effects.
12:47
SPEAKER_00: For example,
12:48
SPEAKER_00: as a precursor to the Stanmore Triangle,
12:51
in 1979 Rockwood identified that the presence of psychiatric
12:54
problems was relevant
12:55
for people who could voluntarily sublux their shoulder in type
12:59
3 instabilities.
13:00
SPEAKER_00: Whilst not explicitly stated in the Stanmore Triangle,
13:03
SPEAKER_00: this historical association has continued to permeate into practice.
13:07
SPEAKER_00: The end result?
13:09
Some of our research that used hypothetical patients to see how
13:12
physiotherapists
13:12
SPEAKER_00: make decisions for shoulder instability patients identified that 1.
13:16
Female patients were more likely to have negative psychosocial
13:18
factors attributed
13:19
SPEAKER_00: to them,
13:20
SPEAKER_00: despite having none stated in the cases,
13:22
SPEAKER_00: and 2.
13:23
SPEAKER_00: Male participants were more likely to be offered investigations sooner.
13:27
Stop here and consider the practicalities and real world
13:30
implications of this.
13:33
How can we practically and objectively identify people with
13:37
genuine psychosocial
13:38
and psychiatric factors that are contributing to their shoulder
13:41
instability?
13:41
SPEAKER_00: I mean, what does contributing in this sense even mean?
13:44
What is the mechanism by which these will impact the stability
13:47
of someone's joint
13:48
SPEAKER_00: beyond the pre-existing constraints of their underlying bony morphology?
13:52
SPEAKER_00: soft tissue structures and muscle architecture.
13:55
SPEAKER_00: The risk here is guilt by association.
13:58
The mere presence of non-traumatic instability makes you look
14:00
for psychosocial
14:01
factors more intentionally than you would in other forms of
14:05
shoulder instability.
14:08
SPEAKER_00: If you were constantly worried about your shoulder coming out,
14:11
SPEAKER_00: or worse,
14:11
SPEAKER_00: it did keep coming out,
14:13
SPEAKER_00: your mood or mental health might be affected too.
14:15
It's also likely to be sore and limit what you can do,
14:18
which doesn't help either.
14:19
SPEAKER_00: Next thing you know,
14:20
some therapist has decided the thing that's most likely
14:23
contributing to your
14:23
SPEAKER_00: shoulder instability is some form of life stress based on your gender,
14:27
SPEAKER_00: although they don't actually tell you this.
14:29
SPEAKER_00: We need to be careful about how and who we attach labels to,
14:32
SPEAKER_00: especially in the absence of any data we can point to and say,
14:35
SPEAKER_00: look,
14:36
SPEAKER_00: here it is.
14:37
Classification systems that inform decision making are meant to
14:40
provide objectivity
14:41
SPEAKER_00: in decision making by at least providing a check on our personal biases.
14:46
SPEAKER_00: They definitely aren't meant to reinforce them.
14:49
SPEAKER_00: So what's the summary of all this?
14:51
SPEAKER_00: It seems existing models are incomplete.
14:53
They can't seem to identify and describe in unambiguous terms
14:56
all of the important
14:58
information required to accurately describe and classify all
15:01
forms of shoulder
15:02
SPEAKER_00: instability.
15:03
This may not be a big problem if we use shoulder instability
15:06
classification systems
15:08
SPEAKER_00: conceptually rather than practically.
15:10
But why are we in this cycle of inventing more and more
15:13
classification systems?
15:14
SPEAKER_00: Well, I'm glad you asked.
15:16
SPEAKER_00: The reason for this, I believe, is...
15:18
SPEAKER_00: The research in this area is thin,
15:21
SPEAKER_00: not well joined up and does not follow the life of the patient.
15:25
SPEAKER_00: So what does the research tell us from a rehab perspective?
15:28
SPEAKER_00: Based on some of the better studies, we know that.
15:30
SPEAKER_00: One,
15:31
SPEAKER_00: adults with atraumatic surgery,
15:33
SPEAKER_00: there is not much difference in outcomes between diagnostic arthroscopy,
15:37
SPEAKER_00: placebo or actual arthroscopic capsular shift surgery.
15:42
SPEAKER_00: 2.
15:43
SPEAKER_00: Existing uncertainties and a lack of robust evidence means that clinical
15:46
recommendations were centred around expert opinion in
15:50
arthroscopic shoulder
15:51
SPEAKER_00: stabilisation surgery for traumatic shoulder instability.
15:54
SPEAKER_00: 3.
15:55
SPEAKER_00: In adults,
15:56
SPEAKER_00: general public,
15:57
SPEAKER_00: typically non-athletic sample,
15:59
SPEAKER_00: with a first-time traumatic dislocation,
16:01
additional or multiple sessions of physiotherapy were not
16:04
superior to a single
16:06
session of advice supporting materials and the option to
16:08
self-refer to physio.
16:10
SPEAKER_00: That's the ARTISAN trial.
16:12
SPEAKER_00: Four,
16:13
high load strengthening exercises are more effective than low
16:15
load strengthening in
16:16
SPEAKER_00: mainly adult females with hypermobile shoulders,
16:20
SPEAKER_00: although results were variable.
16:22
SPEAKER_00: Five,
16:23
SPEAKER_00: in adults with multidirectional instability,
16:24
SPEAKER_00: the Watson multidirectional instability program seems to perform better.
16:29
than the rockwood instability program six there is very little
16:32
if any information
16:33
available for young people with shoulder instability even in
16:36
the basic science and
16:38
mechanism space i was surprised when writing up our study that
16:42
all of the previous
16:43
research on muscle activity and movement differences were
16:45
predominantly adults with
16:47
the odd young young person our data set is probably one of the
16:52
youngest recorded
16:54
There are other studies that provide helpful information such
16:57
as the Derby
16:58
Instability Program and other similar single work group
17:01
longitudinal studies.
17:03
SPEAKER_00: But remember,
17:03
when we want to know something actually works or is better than
17:06
something we're
17:07
SPEAKER_00: already doing,
17:07
it's usually well-designed and appropriately powered RCTs that
17:11
are needed.
17:12
Although there are other ways of evaluating effectiveness that
17:14
I'll discuss later.
17:16
Consider then how does the existing evidence inform our
17:20
decision on what to do with
17:21
a young person who has multi-directional slash atraumatic
17:24
shoulder instability or
17:26
SPEAKER_00: an overlay of traumatic and atraumatic?
17:28
SPEAKER_00: Where are the RCTs for them to inform our decision making?
17:32
SPEAKER_00: The truth is getting funding in this area is difficult.
17:35
SPEAKER_00: Compared to other health conditions, e.g.
17:36
SPEAKER_00: cancers,
17:37
SPEAKER_00: hypertension or heart attacks,
17:38
shoulder instability as a proportion affects a small percentage
17:41
of the population.
17:42
It therefore doesn't always tick the value for money box
17:45
funders are after.
17:46
It also means getting enough people through the doors to answer
17:48
the study question
17:49
SPEAKER_00: is more challenging.
17:52
SPEAKER_00: It's also partly a hidden problem.
17:54
Estimating the true instances for some of the more complex
17:57
shoulder instability
17:58
SPEAKER_00: subgroups is challenging due to a lack of robust evidence,
18:00
SPEAKER_00: a recurring theme in this field.
18:02
No one really knows how big the problem is and the
18:05
complications are often delayed.
18:07
Funders also usually want to fund studies that answer a
18:10
question in a reasonable amount of time.
18:14
What's not quantifiable in the data but very important to
18:17
consider is that people
18:18
with more mixed pictures slash complex slash multi-directional
18:22
instability usually
18:25
SPEAKER_00: present as children or adolescents.
18:28
They may have several episodes of instability and do not
18:30
present until it starts
18:31
impacting on them and their parents or carers daily lives in a
18:33
significant way.
18:35
SPEAKER_00: likely that the true size of the problem, e.g.
18:37
number of instability episodes and duration of the problem,
18:40
is underreported.
18:41
SPEAKER_00: 2.
18:42
May be reliant on their parents or carers for communicating
18:44
this information given their age.
18:46
SPEAKER_00: Their voice in the research agenda is therefore not prominent.
18:50
SPEAKER_00: They also likely have other life priorities or issues to sort out.
18:53
Remember how complex being a teenager was and then your arm
18:56
kept popping out?
18:59
SPEAKER_00: Three,
18:59
SPEAKER_00: sorry,
19:01
can spend a long time in the healthcare system going between
19:04
services despite being
19:05
SPEAKER_00: a relatively small proportion of service users.
19:08
SPEAKER_00: They can be complex cases for clinicians.
19:10
They may see multiple clinicians and be referred between
19:13
multiple services,
19:14
SPEAKER_00: high healthcare burden for them and us,
19:16
SPEAKER_00: which may not result in joined-up care,
19:19
SPEAKER_00: although throughout their life,
19:20
how many healthcare providers might someone with recurrent
19:23
instability see?
19:24
It involves telling your story to lots of different people
19:26
again and again.
19:28
SPEAKER_00: Overall,
19:28
a worse experience of healthcare and having to wait for formal
19:32
diagnosis and
19:33
feeling like you're being passed between services consider also
19:36
the potential for
19:36
SPEAKER_00: being labelled as having psychosocial problems.
19:39
SPEAKER_00: Unlike surgical interventions,
19:41
there's not a culture or governance requirement to record and
19:44
report outcomes.
19:46
The end result is we can't actually evaluate what we are doing
19:48
to put a number to the problem.
19:50
SPEAKER_00: There's also a funding gap, which I find a bit mind boggling.
19:54
SPEAKER_00: We know that shoulder arthritis is a big problem.
19:57
SPEAKER_00: We know that in traumatic shoulder instability,
19:59
you are between 10 to 19 times more likely to get shoulder
20:02
arthritis later on.
20:05
What we don't know is how likely arthritis is to develop in
20:08
someone with atraumatic
20:09
shoulder instability who has experienced multiple episodes of
20:13
instability again and
20:14
SPEAKER_00: again.
20:15
SPEAKER_00: There are charities and funders set up to address arthritis.
20:18
There are charities and funders set up to address health issues
20:21
in children.
20:22
SPEAKER_00: Both are opposite ends of the spectrum,
20:24
SPEAKER_00: meaning those with more complex and under-researched shoulder instability
20:27
SPEAKER_00: presentations fall between this gap.
20:31
SPEAKER_00: If arthritis is such a big problem,
20:33
why aren't we focusing on those most at risk to try and prevent
20:36
it if possible?
20:37
If funders are serious about addressing underserved groups or
20:40
ensuring fairness
20:41
SPEAKER_00: through,
20:42
SPEAKER_00: then this surely is one of the groups who should be considered.
20:45
SPEAKER_00: So in essence,
20:46
the reason there's limited new and rigorous evidence is that
20:48
it's pretty hard to
20:49
SPEAKER_00: get the big definitive studies off the ground for the reasons stated.
20:53
SPEAKER_00: Often,
20:54
if there isn't a clear pathway to big clinical trial that
20:58
answers a specific
20:58
SPEAKER_00: clinical question,
20:59
it can make it a bit more challenging to get the initial
21:01
funding that's required to
21:02
SPEAKER_00: develop these projects.
21:04
Research studies therefore tend to focus on small development
21:08
or pilot work,
21:08
SPEAKER_00: which doesn't always progress on.
21:10
The end result is a lot of research,
21:12
not all of which can be used to inform our decision making.
21:15
Doing good research in this area is not impossible,
21:18
just really,
21:18
really difficult.
21:20
SPEAKER_00: There's also a circular logic problem.
21:22
SPEAKER_00: A lack of evidence means getting funding to do more research difficult,
21:26
SPEAKER_00: which is needed to produce more evidence.
21:28
SPEAKER_00: The problem here seems to be... 3.
21:31
We seem to be repeating the same mistakes and it's time to try
21:34
something different.
21:36
SPEAKER_00: Albert Einstein is famous for saying,
21:38
Insanity is continuing to make shoulder instability
21:41
classification systems and
21:42
SPEAKER_00: treatment plans in the absence of new data or RCTs.
21:45
SPEAKER_00: Or something like that.
21:47
Given that we've gone beyond 19 instability classification
21:50
systems and the rehab
21:51
SPEAKER_00: evidence base is likely less than 10 papers,
21:53
SPEAKER_00: it's probably time we did something different.
21:56
SPEAKER_00: The way I see it is there are two possible options moving forwards.
22:00
SPEAKER_00: One,
22:00
this article inspires a drastic change in how shoulder
22:03
instability work is funded,
22:04
SPEAKER_00: leading to more definitive trials.
22:07
SPEAKER_00: I wish.
22:08
Or two,
22:09
we need to try something different,
22:10
innovating and working within the systems we've got.
22:13
I think a fundamental step in this is improving how we measure
22:16
people with shoulder
22:17
SPEAKER_00: instability for both diagnosis and their lifespan.
22:21
SPEAKER_00: Whilst RCTs are an excellent methodology for demonstrating effectiveness,
22:25
they aren't always as well suited practically to less common
22:28
conditions or
22:28
SPEAKER_00: subgroups.
22:29
That doesn't stop us from trying to answer some of the ongoing
22:31
clinical questions
22:33
SPEAKER_00: using alternative methods,
22:35
SPEAKER_00: as long as we understand the implications of these.
22:37
Remember,
22:38
in some cases,
22:39
we won't be starting from a high bar,
22:42
just expert opinion.
22:43
There are some innovative studies going on in this space which
22:45
provide hope.
22:47
SPEAKER_00: Maroda et al,
22:49
SPEAKER_00: 2020,
22:49
used fluoroscopy to classify and confirm different types of
22:52
shoulder instability.
22:53
SPEAKER_00: I agree we can't use that on everyone,
22:55
but it's a helpful starting point from which other technologies
22:57
or solutions may
22:58
SPEAKER_00: develop.
23:00
SPEAKER_00: In other rare diseases or complex movement disorders,
23:03
things like 3D movement analysis and electromyography are used
23:06
to inform decision
23:07
SPEAKER_00: making.
23:08
They also tend to have really good data sets that follow the
23:10
patient through their life.
23:12
SPEAKER_00: Gillette Children's Hospital is an excellent example of this.
23:16
You might say the technology is not there and it's too
23:18
complicated for practice.
23:19
The truth is the technology has been there for quite a long
23:21
time and is already
23:22
SPEAKER_00: used for making decisions about rehab and surgery.
23:25
These arguments were around in the early days of human movement
23:28
analysis itself and likely CT.
23:30
SPEAKER_00: Just run that past me again.
23:31
You want to buy a £100,000 plus machine that emits ionising
23:36
radiation while
23:36
spinning at 120 to 240 revolutions per minute and you have to
23:40
lie in the middle of
23:40
SPEAKER_00: it.
23:41
SPEAKER_00: However,
23:42
as we were able to demonstrate how improved diagnosis improved
23:45
outcomes for people,
23:47
these new ways of diagnosing people were introduced and made
23:49
standard in clinical
23:50
SPEAKER_00: pathways.
23:52
If we can improve our diagnosis alongside improved recording of
23:55
outcomes that
23:56
SPEAKER_00: follow people throughout their life,
23:58
we can then start to make positive steps to seeing not only how
24:01
big the problem is,
24:02
SPEAKER_00: but what seems to change outcomes.
24:05
This is always the starting point for further well-designed
24:08
definitive trials.
24:09
In a world of digital health,
24:11
secure data environments and AI,
24:13
this should be easier than ever.
24:15
SPEAKER_00: However,
24:15
the risk is these just become really good systems that store
24:18
and regurgitate really
24:20
SPEAKER_00: rubbish information.
24:21
SPEAKER_00: We need to try and avoid iterative degradation.
24:26
SPEAKER_00: So what needs to happen next?
24:27
My hope is that if nothing else,
24:29
we stop and take stock of where we're at.
24:31
SPEAKER_00: The answer is too much unusable research is not more unusable research.
24:35
SPEAKER_00: The answer to too much complexity is not more complexity.
24:39
SPEAKER_00: We need to go back to basics.
24:41
If we are to continue to keep rising while standing on the
24:44
shoulders of giants,
24:45
SPEAKER_00: it's important the foundations are right.