Hello and welcome to the BJGP Top 10 podcast.

So this is where we take a closer look at the most read research papers in the BJGP in 2025 and just have a discussion about what they mean for day to day general practice.

I'm Nada Khan, one of the associate editors of the Journal.

And in today's episode we'll be exploring some of the themes that really captured attention with the readership, I suppose.

And we'll be talking about things like consultation compl complexity and workload pressures.

Some work around diagnostic uncertainty and how to look, look after people with multimorbidity.

And I think we're going to have a discussion a bit more, not just about what these papers found, but maybe a bit about why they resonated and maybe give a bit of editorial feedback around that.

And because it's a conversation here between three clinicians as well.

And I'll go around and introduce everyone in a minute, maybe a bit about what they add to the conversations we're already already having in practice and where the gaps still are.

And I guess with that we'll be keeping it grounded in the messy reality of today's general practice as well.

So I've introduced myself and I'm joined here by Tom Round and Sam Merrill, who are both also associate editors of the bjgp.

But I'll go to Tom first.

So, yeah, tell us a bit about who you are and how is your day going?

Great, Nada.

Thanks for having me.

So, Dr. Tom Rand, I'm a GP in northeast London and an academic clinical lecturer at King's College, interested in early disease and cancer detection and also health inequality.

So, yeah, pretty good.

Like everyone, I've got a mild cold at the moment.

I think exactly the same last year when we did this podcast, winter cold season.

So I think we're all sort of feeling that a little bit in primary care with flus and other things and staff, you know, so otherwise good.

Looking forward to having really interesting discussion about these papers which are really fascinating and give a real broad breadth of what we do in general practice.

Great.

And Sam, we'll go to you and you have some really exciting news in the background as well.

So, yeah, tell us about who you are and what you're up to today.

Thanks, Nad.

I think, yeah, you're alluding to the fact I'm on Puppy alert because our new addition to the family in the winter is keeping us busy and making remote working a challenge.

But we're getting through.

But yeah, lovely to be with you guys.

And I catch up and BJGP and wider podcast audience.

So, yes, I'm a GP working in the Northwest of England and a clinical senior lecture at the University of Manchester.

Brilliant.

Okay, so let's get into the top 10 most read research and published papers of 2025 and I'm going to kick off with number 10 and number 8, just because they're on a sort of related topic.

So number 10 is by Michael Anderson and colleagues.

Michael's based in Manchester and at lse.

And this paper looks at prescribing, quality in practices and the role of clinical pharmacists as.

And I'll just point out that I'll put links to all the papers in the show notes as well.

So this paper looks at the adoption of clinical pharmacist roles in English general practice and asks that question of does bringing pharmacists into the primary care workforce actually lead to improvements?

Michael looked at this through a longitudinal approach.

They used national practice level data from 2015 to 2019 and just looked at practices that didn't, didn't have a clinical pharmacist role.

And it's really interesting, the results actually.

So, not surprisingly, the proportion of practices with a clinical pharmacist increased from about 3% to over 20% over the course of the study.

And the, the team found some really significant improvements across several prescribing indicators.

So things like reductions in total medication costs, better opioid prescribing and prescribing for anxiety meds after pharmacists were implemented in pract, I guess, really it would be interesting to hear your thoughts, Sam and Tom, about what do you think really are the outcomes we want most from clinical pharmacists?

And how do you think we should interpret these modest changes at scale?

Because there's a lot else going on in terms of workforce that we need to think about in general practice, like access and continuity and not just meds optimization.

I can talk from practice experience because our clinical pharmacist just left for Canada just in the last month or so.

But yeah, it was really interesting, like having him part of the team.

I think in a lot of ways he took a lot of burden off the gps in terms of meds monitoring, meds management, medication reviews.

He builds a lot of continuity with a lot of patients because he was doing a lot of checking in.

So in a lot of ways he was quite invaluable member of the team and we have sought a replacement since.

At the same time, you know, there was some, some extra challenges in terms of workload and stuff, because obviously pharmacists have different training and the role of a clinical pharmacist in general practice is relatively new.

So, you know, their approaches to prescribing and, you know, how close you stick to guidelines and how much you adapt for individual patient situations is slightly different.

So.

Yeah, but I think that was part of sort of feeling a way out with the role.

But it's really noticeable when he's not around because it does affect how the workload flows and how the practice runs and how the patients, you know, interact with the practice.

So, yeah, it's been really interesting at the local level.

But, yeah, Michael's study also very interesting to look at the wider picture about how it's affecting quality of care.

And I'll just jump now to number eight, unless, Tom, you want to add.

No, just to say, obviously this, this paper is looking at the macro level up to 2019, so it'd be really interesting what happened since, because we only started having a pharmacist after that point with the induction of ARS roles.

So I think, yeah, further, you know, this is giving a signal, we think that some indices are improving, but also I think it's important to be aligned with our own subjective experience, maybe qualitative and other implementation type research.

But overall, I think this trend is a good thing, I think, from my own experience.

Yeah, absolutely.

And then I guess jumping to paper number eight, which was written by William Hollingsworth and his team from Bristol, and this is looking at comparing paramedics in general practice with gps.

And the paper is asking a really practical workforce question, which is, is what happens to patient experience, safety and NHS costs when patients are seen by a paramedic in general practice rather than a gp.

And this team looked at this, they used a prospective cohort study across sites in England and they looked at patients who had an urgent or same day consultation with either a paramedic or a GP and then looked at their outcomes over the next 30 days.

And I guess the headline finding is that really there wasn't a clear difference in patient reported health and well being over 30 days, but there were some differences in that experience right after the consultation.

So patients who saw a paramedic said that they were.

Well, they reported lower confidence in their health provision, they felt there are more communication problems and maybe a lower perception of how the practice promotes safety.

And there were fewer subsequent GP appointments in the paramedic group, but there weren't really any GP savings as such that were offset by higher use of other health care professionals.

So I guess that you could sort of summarize that by seeing.

Seeing a paramedic might lower GP pressure, but it doesn't necessarily reduce overall NHS costs.

So I wonder, yeah, Tom, coming to you, what do you think should really matter when we diversify the workforce?

Do you think it should be workload, cost?

Yeah.

Really interesting discussion, isn't it?

And we talk about testification, isn't it?

Sort of, you know, how do we, you know, how do we help GPs with workload?

Workload, sorry.

Fundamentally we need more GPs, don't we?

We need to have, you know, we've got high 2,300 to 2,500 patients, sometimes even higher deprived areas.

So fundamentally, I think the workforce, we do need more gps.

This also debate, also, you know, obviously there's a slightly toxic now debate about physicians, associates.

You know, from my own viewpoint, you know, undifferentiated initial consultations in primary care are high risk.

We know that from all the evidence and the research.

So you've got to be very careful about patient select selection and triage for this.

And you can see, I think also this links to.

We've got this big increase in the ARS roles, but then we haven't seen that increase in primary care satisfaction.

So I think this comes down to people probably still want to see a GP for certain conditions.

How do we get to that right model of MDT working?

And I think we do need robust safety evidence.

So this obviously is, you know, it's good study, but it's fairly small scale, probably need larger scale and systematic review evidence about this replacement.

You know, what's the safe role?

What are the guidelines?

What sort of cases should these people, should other allied healthcare professionals be seeing, particularly for undifferentiated illness?

And going back to the, obviously, the PA debate, we've obviously got the college position that probably PAs should not be seeing undifferentiated illness in primary care.

So I think it's a nuanced discussion, but we need better, we need further studies like this to help us decide what we're doing.

Absolutely.

And I think that's really important as the workforce in general practice increases to diversify and policy shifts towards an increasing multidisciplinary team as well.

So, yeah, be interesting to see what happens in the future.

Really great.

So I'm going to go over to Sam and Sam, you're talking about paper number nine, but, yeah, talk us through this.

This is a bit a paper that, you know well, so tell us a.

Bit more about it and your involvement.

In it, first author on a BJGP top 10 paper.

I'm very honored.

Congratulations.

Humble to all the readers out there who had looked at it.

So this was a study of asymptomatic prostate cancer detection using PSA in primary care in England.

And we used data from what's called the National Cancer Diagnosis Audit.

This was the 2018 version.

So we had about a quarter of practices in England participate in the ncda and data was gathered using a sort of standardized template on all the new cancer diagnoses in a practice in 2018.

So practices participate were given that list and a template to complete and looking at the record in detail.

So what happened to these patients in the lead up to their diagnosis?

Were they seen in general practice?

What happened?

Were they investigated?

Were they referred to?

And it was not screen detected cases for any of these were specifically cases coming through primary care.

And the strength of this data set is that we have access to both coded and free text data in the record.

So a lot of large primary care research data sets like CPRD don't have free text data.

So it's relying on GP coding, which we know varies between practices.

So the big things that this study found we looked at.

So There were nearly 10,000 prostate cancer cases in the entity.

Overall, when we filtered out all the patients who had symptoms recorded at the time of presentation of primary care and the time of diagnosis, we were only left with about 1900.

So the vast majority of men with prostate cancer symptoms were present at the time of diagnosis, which conflicts somewhat with existing literature out there, the quality of which is pretty variable and often not great.

So that was one interesting finding.

In terms of the differences between practices for asymptomatic prostate cancer detection and PSA testing, there's huge variation, something like 14 fold difference between the practices picking up the most men through asymptomatic PSA testing and the practice picking up the least.

And we didn't see any obvious GP practice level factors, so it didn't matter about geography, list size, number of GPs, cough outcomes, none of that seemed to make any difference.

There were patient level factors, so older men less likely picked up through that route, which kind of makes sense because symptoms are much more common in men as they get older.

And PSA testing, the benefit is less generally depending on their general health, so it might be done less often.

So that makes sense.

Men from deprived areas were less likely to be diagnosed through this route, which we know there are significant inequalities for men deprived regions in terms of prostate cancer outcomes.

Not Just PSA testing, but stage of diagnosis, treatment outcomes, we need to do better with that group.

And interestingly, white men were less likely to be diagnosed through this route.

Even though the sort of stereotypical person being, coming in, asking for a PSA test when there are no symptoms and maybe a low risk is a.

Is an older, wealthy white male, they were less likely to diagnose through this route, which.

That was an interesting finding.

Yeah.

So really interesting study.

Obviously grabbed some interest and is a very, very, very topical issue at the moment with the NSCS recommendation that's out for consultation.

And I think, you know, we still got to watch this space because I think there's going to be more coming in the year's ed.

Yeah, really super topical, Sam.

And just to point out, we did record a podcast talking about this paper in more detail, if anyone wants to listen to that.

Tom, you work a lot in cancer diagnosis in that sort of world.

I mean, obviously brilliant work from Sam and his colleagues, but I just wanted to know what your thoughts were.

Just reflecting on this paper in terms of sort of the wider policy discussions and discussions around the future of prostate cancer screening.

Yeah, yeah.

So I think it's very topical, isn't it?

There's lots of.

In the press around, you know, should we be doing PSA testing?

So we currently got a slightly fudged position or a patient asks for it and they've made that informed choice.

We.

We can offer it.

I think most.

She's probably.

Most gps probably would if a patient is, you know, making that informed choice.

But given that a lot of prostate cancer can be missed with a.

With a normal PSA.

So not a great test.

I think most GPs would probably, on balance say probably not for a national screening program as yet, but we need to do more research and Sam's involved in the Transform trial, which is using probably PSA with other risk factors.

So it might be about that high risk case finding, a bit like we're doing other trials in other.

In other cancer areas.

So I think psa, not a great test, not good for asymptomatic screening, has a role in symptomatic testing.

We need to be more nuanced and have further research evidence before we, you know, there is a need obviously to have early diagnosis and having good quality RCTs in this area.

Yeah, brilliant.

Thanks, Tom.

We're going to move on to number seven in our Top ten countdown.

And this paper was about what patients want from access to UK general practice.

And Sam, I think you're going to introduce this paper.

So tell us a bit more about this one.

Yeah, so this was paper by Helen Atherton and colleagues and really simple question and, you know, review methodology.

Not complicated but actually really important, you know, and access is such a topic thing that keeps coming up again and again and we know we have issues with not quite getting access right for a range of reasons.

So, yeah, I think this was.

I'm not surprised this isn't the top 10 because it generates a lot of interesting insights and sort of the published literature around that is right across the board in terms of, you know, the speed of access and the type of access and what's quality in terms of access to care.

So, you know, there's a clear message from patients about continuity and clinician choice.

You know, they, and often patients who valued having continuity with a clinician in the practice would wait for an appointment unless it was a, you know, they were very ill and they needed an urgent appointment.

But those that value continuity would be willing to wait.

You know, obviously they still wanted a timely appointment.

But, you know, this, this constant sort of perception of demand for same day appointments didn't hold so strongly when continuity was valued.

Communication from practices to patients about how to access appointments, who's available, when are they available, all very important.

And in this modern sort of setup of multiple routes to access, which I'm sure we'll talk about with some of the other papers, that seems to be really critical from a patient perspective.

So that's really interesting insight.

We talked before about the non GP roles in primary care, so patients weren't opposed to that, but for specific relevant things.

So the clinical pharmacist we talked about earlier, if it's a medication query, seems like from the literature, the patients are happy to discuss that kind of thing with a pharmacist as much as they're any other member of the team, but for other issues, they very much want to see a GP or things like that, and ease of booking as well.

The 8am rush gets talked about a lot and there's been various policy initiatives to try and create alternative routes and manage that, but patients do find that stressful if they feel they can't easily access the point where they need it.

My main sort of thinking from this paper was that a lot of the findings seem to make complete common sense, so people wanted to make an appointment in a timely fashion, they wanted to choose what kind of clinician they were seeing.

But some really sort of simple things, just about sort of, you know, wanting to have a practice that was near to Them and have a practice that was, you know, that had a simple appointment booking system.

So this isn't necessarily complicated, but it's complex to implement on the ground.

Any thoughts, Tom, about that one?

Oh, no, totally agree.

And I think it will link to our next one talking about techno stress, about sort of computers booking systems.

You know, obviously we're very keen to reduce health inequality so, you know, thinking about one side doesn't fit all.

We have some patients that can really utilize some of these systems well and easily, but we, we've got to be very aware of the inverse care law and digital inverse care and help certain patients with access continuity.

Yes, certainly matters for our patients with long term complex conditions.

You know that 30, 40% of people who have multiple long term conditions.

And you've alluded to our paper number six in the top 10 countdown, which I think probably has my favorite title of any paper.

I love that.

Yes, few years.

Yeah.

And as you mentioned, it's about techno stress and techno suffering.

Sam, tell us a bit more about this paper because this sort of of, you know, goes along the same thread really about access and also about remote and digital work.

Yeah.

And so this paper focuses on how it impacts on staff in UK general practice and Shout out to the Remote By Default team led by Trish Greenhalgh, who have three papers from the same study in between this year's top 10 and last year.

So well done to that team from Oxford.

This particular paper was led by Francesca Dakin and yeah, they sort of, in the context of the Remote By Default study, they looked at these, yeah, a number of different theories around impact of technology on healthcare staff and systems, how that interacted with relationships as well as individuals.

And in the.

Obviously the project was initiated in the context of the COVID pandemic, but it's still so massively relevant.

We've just had the change towards, you know, online access for requests for appointments at any time to try and mitigate that 8am rush we were just talking about.

So knowing on how these systems affect staff is really crucial and it's fascinating.

Like, obviously there's the context around the workforce and workload challenges that primary care faces, but even within that, the staff talked about clear positives and negatives to the technology.

You know, so it gave staff some ability to be more flexible in managing the demand.

They could look across, you know, the appointments were coming in and triage a bit more better.

Some staff felt that actually giving themselves a little bit more distance had positives and negatives in terms of managing their Own well being and things.

And reception staff talking a lot about sometimes the abuse or aggression they get when appointment demands can't be met.

And they're the ones that bear a brunt of it a lot in our practices.

So having alternative more remote ways help protect them a little bit.

But then also, you know, patients and clinicians have highlighted, you know, there's this level of dissatisfaction with remote working compared to a traditional face to face appointment.

Yeah.

And so many fascinating insights that they're kind of new avenues for access, creating demand and you know, inducing more supply, demand and stuff.

It just that vicious circle of you give more routes to access, suddenly there's more demand, there's more ways of people accessing primary care that has an impact on the workplace, suffering and challenges around the staff.

And I guess also the GPS talk a lot about don't sit in your consultation room all day, come and sit back even if it's for five minutes, come and chat to someone else, a member of the team and that.

But if you're doing more and more remote working, there are opportunities for interacting.

You're not walking to the waiting room anymore.

You don't need to leave your desk to deal with phone calls or emails or online consultation requests.

So staff in the SIS study talked about the importance of even if you just walk to reception to have a chat to someone, you know, stretch your legs five minutes, that's so important for your own well being.

So yeah, really important study, really well thoroughly done.

I highly recommend these papers because they generate a lot of really good insights that are so relevant to how primary care is evolving.

Brilliant.

Okay, right, so we're going to move on to the next paper that is number five in the top 10.

And I'm going to come to to you Tom, to talk about this one.

And this paper looked at antidepressants and risk of postural hypertension.

So yeah, tell us a bit more about this one.

So really interesting and topical I think.

I think really interesting, a really interesting paper.

So postural hypertension, just those reminders, definitions.

So that would be if you're dropping more than 20 milligrams of mercury for systolic or more than 10 for your diastolic reading within three minutes.

Do you remember that from our medical school we do it in practice.

Isn't it interesting about one in three older adults has postural hypertension increase increased risk in Parkinson's and other long term conditions.

Also those on multiple medications, we might do that medication review.

So an interesting kind of area, one that I wasn't really aware of that.

It was this increased risk with the antidepressants.

So this was a large primary care database study with over 41,000 patients for nearly 20 years looking at obviously coded data for postural hypertension and looking at were patients on SSRIs on tricyclics, on other antidepressants.

And they did find a fourfold increased risk if you are within the first 28 days, but not after that.

So I think that's a kind of crucial bit for us into clinical practices to warn patients around if we're starting in older adults antidepressants.

As a note in this paper, it was well known and it's in the side effect list for SSRIs, but less so for tricyclics and for the newer SNRIs, mirtazapine and duloxetine.

So yeah, the rate was about fourfold increase in that 28 days for SSRIs, but also two factor increase for tricyclics like amitriptyline and also for the newer SNRIs.

So yeah, I think the take home message is particularly in your older patients who are on blood pressure medications, careful prescribing, start low, go slow.

That's almost my analogy in medications.

Of course we should treat depression appropriately in older adults, but really, can they check blood pressure at home?

Is there a carer they can check with them if they're feeling dizzy?

Can they check the blood pressure?

Can do it sitting and standing.

Do they need to have a health checkup with a HCA or pharmacist within the practice when you start in these meds?

Actually quite some a good clinical topic, I think and some key take home messages for clinical practice.

I think that's exactly it.

And we did a podcast with Cinniboni, the lead author of this study and my takeaway from that was that actually this is going to change my practice as well when I'm prescribing these medications to older adults aged above 60.

So really clear clinical message which I think is fascinating coming out of this sort of big database study as well.

Sam, any thoughts about this paper?

Anything you wanted to note?

Yeah, no, I guess it's just helpful to think about the physical effects as well.

As you're treating someone for their mental health, you want to arrange follow up and make sure that they're experiencing the treatment well and you do warn them about the short and long term effects from treatment.

But just having a more sterilized awareness of what you need to talk about and when and what you need to check in on is really, really helpful.

Absolutely.

And I think that's it.

Just having something to mention to patients when you're starting these medications, just to look out for it.

So that's really, really great as a clear clinical message.

And we're going to come back to number four in the top 10 countdown.

And we're going to back to the team from Oxford that was led by Rebecca Payne for this study.

And this paper looked at challenges to quality in contemporary hybrid general practice.

So another paper along this theme of sort of access and quality and I guess again, you know, we're talking about care that's increasingly delivered through a blend of different types of consultation approaches.

So remote, digital or face to face.

And this sort of multi site, mostly qualitative study looking at UK practices just really looked at sort of quality improvement efforts that were happening in these practices, just in this context of all the strain that's happening in practice as well.

And a key result of this paper was that the human elements of general practice, like relationship based care, compassion and continuity, I know we've mentioned that a few times, were really difficult and sometimes impossible to sustain in practice, especially when you're working sort of in an asynchronous model of working as well.

Interestingly, some of the systems that the practices had introduced to try to improve efficiency have created new forms of inefficiency.

So things around accessibility or patient centeredness or equity, these sorts of things don't really compensate for structural disadvantage and digital exclusion.

And Euan Lawson, our editor in chief at the journal, actually felt that this was one of the most important papers we published last year.

But yeah, I wonder what you both think about what high quality general practice actually looks like in this hybrid era and how are we going to build these systems that improve access without fragmenting care?

I guess.

Yeah, that's a tough one.

Big question, sorry.

I kind of feel like we're all trying to feel our way and, you know, there are definitely practices out there that do this well and there are definitely practices out there that are struggling with it, you know, and it's sad, like I don't think we have enough experience and know enough about to say, is it, is it just the context?

Is it just the demographics of patients we serve and the geography and the way the practice is set up, or are there just good ways of doing this that should become standard practice to make this better for all?

Because, you know, workload that increases with total triage and things like that, you know, there's no, it's clearly Ways that don't help.

But yeah, I don't know if we know enough and we need stuff like this.

This data generated from this project is so valuable because it does create insights, but it's in depth data from a few practices and obviously there's thousands of practices around.

So yeah, I don't have the right answer and it's a huge question but.

But yeah, I feel like I'm still feeling our way with this.

I think really good.

You know, this is obviously the 12 general practices.

I think it's very linked to the, I think probably the same practices with the techno stress one with the ethnographic work.

So it's really good that we've got some high quality qualitative research in our top 10, you know, that can really help us in terms of like some of the key things we want to have a look at in our practice.

Like everyone else, we're onto the full triage model till half six, you know, that's pretty exhausting.

We split the day so you either do the morning or the afternoon.

We try to embed continuity sort.

So we started doing continuity project that for our high risk patients.

We flag in the notes who their regular doctor is.

So then you kind of reduce that, that taskification, the fragmentation and then if someone's triaged, I pretty much say you are coming and see me face to face, which we know from research actually most patients probably want a face to face appointment, some don't, but majority do.

So I think it's actually shifting that narrative to say look, digital triage, but majority of people maybe need a face to face appointment appointment and maybe longer appointments.

Actually we need to be a bit more flexible in general practice.

Some patients, you know, can't fit in that 10, 15 minute model.

We might need a 20, 30 minute appointment.

A bit like the Nordic countries where that's more proactive and we get better outcomes.

So I think we're all in a bit of flux at the moment, but fundamentally we need more staff and more resourcing.

Yeah, fair enough.

I think that's really the bottom line.

And a new contract for GPS for sure.

A 20 year old contract.

Absolutely right.

So we're going to move on to number three in the top ten.

And that's right.

So Tom, tell us about this.

So it's about a really super common condition that we see in general practice and also a really super common medication that we prescribe.

But tell us more about that.

Yeah, so this is actually another great qualitative piece of work.

So this is around low dose amitriptyline for IBS.

So IBS very common.

15 to 20% of people have inortal bowel syndrome.

Amitryptone we've had for a long time trypsychic antidepressants.

So this was a nested qualitative study in the Atlantis trial which was an RCT published in the Lancet, the RCT behind this and it won research paper the year last year, remember seeing it at the RCGP conference.

This has really changed my clinical practice.

So this clearly demonstrated that for second or third line treatment, amitriptyline is very effective for ibs.

And they found, actually this is a placebo controlled trial and they found that the side effect rate was no different between intervention and control.

So often we're very worried about amitriptyline side effects.

They found no difference between amitriptylate and to 30 milligrams compared to the control group.

Also super interesting for that Atlantis trial they had a PPI co produced self titration document.

So I'd actually change my practice and enable patients to self titrate using this protocol up to 30 milligrams of triptyline.

So that's really useful.

Have a look at the trial resources.

So yes, so this was an embedded qualitative study both with patients and with GPs looking about the sort of concerns and how we address those concerns.

So some of the key concerns are about, you know, I'm taking antidepressant but I'm not depressed.

What are, what are the potential adverse reactions I could have?

Is this over medicalizing IBS and then how we might address that about, you know, saying amitriptyline is well used, we use it, we don't really use it for depression anymore.

We use low dose for chronic pain, it's a low dosage, it's not for depression.

It's well established, it's safe and easy and flexible to use and has benefits beyond ibs.

So I thought, yeah, a great, a great, great demonstration of how you can embed a qualitative study to a great rct.

Yeah, and what's really great here is that, you know, the patients and GPs in this study felt that actually it's worth a go really because of the lower side effect profile and maybe some of the benefits might actually outweigh the concerns as well.

So again, like the postural hypertension study, this is a paper with, you know, some quite good, good ideas for how we can roll these sorts of interventions out in clinical practice as well.

So that, that was really great to hear about.

So we'll move on to paper number two in the top 10.

And this paper was focusing on artificial intelligence techniques for early detection of early detection of lung cancer.

And this was a paper that was led by Martin Schutt and colleagues, based in Amsterdam.

And it's really fascinating, this one.

Basically, the team looked at, they analyzed free text GP clinical notes and looked at whether AI techniques could help detect lung cancer earlier.

The question here is, yeah, are there any signals in that electronic record that might contain the signals that we don't really recognize in real time?

And the team used sort of natural language processing techniques to build this prediction model from GP consultation notes and then test it fit with both internal and external validation in this study.

Interestingly, the model that they developed predicted lung cancer about five months before diagnosis and about four months before referral.

And the model performed generally fairly well, actually, in both the internal and external validation.

But the trade off is that the performance of the model really depends heavily on the risk threshold that you see, which means that in day to day practice it could generate a lot of alerts if it was, you know, rolled out in an electronic health record.

So patients might get a lot of clinical workup for relatively fewer true cancer diagnoses.

So I guess really, I mean, both of you work in sort of the cancer diagnosis area and I wonder what your thoughts are like, do you think that this kind of work or AI techniques could genuinely shift the stage at which cancer is diagnosed and improve outcomes, or do you think it's just going to risk creating more noise and anxiety and extra workload without really clear benefit?

Yeah, it's a huge question and there's obviously a lot of excitement and interest about the role of AI in healthcare.

Obviously this is a retrospective study looking back at notes that are available.

So, you know, if you're going to, why?

And I guess people have to think about where does it fit in with clinical practice?

Is it an aid?

Is it like something that prompts you?

Is it something that goes back over your consultation notes after the patient's left and say, oh, we picked up on this and that, and they've mentioned this three times.

So have you thought about this kind of thing?

So in other spheres, so in screening, for instance, they're starting to look at the role of AI as like a second reader for screening imaging test.

So breast mammography or lung CTS for a lung cancer screen.

So there's still a pair of human eyes on it and stuff, but the AI can either sort of be a second pair of eyes, can sort of Start to guide people towards areas of interest.

So in the messy world of undifferentiated people turning up to primary care, that's obviously going to have to be really sophisticated and fit into how we deliver care.

So feels like it's way off still and the thinking is moving ahead.

So it's not just about what's happening in the consultation, what people are coming with.

It's patients entering their own symptoms into an AI chatbot type setup and it tells them what to do, whether it's triaging or other things.

So, yeah, there's a lot of different thoughts about how it can be applied and there's a lot of modeling going on.

I feel like there's a lot more thought still needed on how to implement it, where it fits in and how it can actually benefit patients.

Yeah.

So early days, really, I guess.

Yeah, absolutely agree with that.

Needs, obviously, more studies like this, but more prospective studies.

More.

More RCTs as well.

But certainly this shift is coming so quickly.

I mean, recent research shows that one in four GPs are already using AI in some form in their consultations.

You know, the large language models, it could be the, the ambient scribe technologies which have been highly promoted.

So I think we're going to see a lot more papers in the AI space.

Space.

And we'll be, you know, when we come back next year, I'm sure we're going to have more in this.

Absolutely more in this space, for sure.

Yeah, I definitely can predict that.

So that's it.

That's something that's definitely coming up as a increasing topic in terms of the research papers that are submitted to us.

And.

Right, we're going to come to the number one paper.

So number one of the top read and published research papers in the bjgp.

And Tom, we're coming back to you.

And it's another discussion around a clinical topic, really.

And we're going back to amitriptyline as well.

So we are.

I feel like I've done a bit on amitriptyline.

So.

Yeah, so, yeah, this was a randomized control trial in the Netherlands looking at amitripty and mirtazapine for insomnia.

So insomnia is obviously a really big issue for us in primary care.

First line treatment, cbt.

But we know there's long waiting list for cbt.

This was then looking at these common drugs which we have in primary care, low dose amitriptyline or metazapine.

So amitriptyline, 10 to 20 milligrams, mirtazapine, that's half a dose 7.5 milligrams up to 15 milligrams.

Patients were 18 to 85 with insomnia.

They recruited 80 patients and they were randomized over 16 weeks to look at outcomes.

And they looked at the insomnia severity index and found significant impact, particularly for mirtazapine.

A bit of a signal for amitriptyline, but more of a signal for mirtazapine at that low dose over six weeks with a quicker improvement and increased recovery rates.

So they showed recovery rate of around 50% on mirtazapine compared to 14% on placebo and amitriptyline, about 36%.

So I think this is really good showing actually this is a real world primary care issue.

We were thinking we probably could use some of these drugs, but actually bringing really good RCT quality evidence to show actually for six weeks of a trial of therapy, this can be quite good.

And obviously it's nuanced.

Depending if a patient's got chronic pain, you might be down the more the amitryptone route like we talked about, you know, headaches, ibs, et cetera.

So I think it's that really that holistic general practice.

But mirtazapine, if it's purely insomnia, not depression, you know, mirtazapine could be an option for that short term, six weeks.

And I think, you know, that could change my clinical practice and already has.

I'm using it as potentially as a, as a second or third line treatment.

Once you've tried CBT and other measures for insomnia.

Yeah, absolutely.

And it's not uncommon for us to talk about sort of the side effects from mirtazapine and amitriptyline as well.

But I thought it was really interesting that the use of low dose amitriptyline didn't really lead to a clinically relevant outcome.

No.

And I think that same like the Atlantis, that was a large scale RCT and that showed no increased adverse reactions compared to placebo.

So I think actually we probably over egg some of the side effects and actually at low doses these are pretty safe medications actually.

So great.

That's it.

That's been the top 10.

Do you guys have any thoughts about the papers that we've discussed here today or anything that you wanted to wrap up?

Well, I'm certainly going to change some of my clinical practice.

Already have with the sort of utilizing amitreptal and mirtazapine in practice, but also thinking about in older adults thinking about that postural hypertension risk.

Just another great advert for the spectrum of primary care research.

We've got trials, reviews, qualitative, quantitative range of topics.

Love it.

Just the brilliance of academic primary care.

Absolutely.

Yeah.

Real showcase, I think, for some of the great research that we've published here in the last year.

But, yeah, Tom and Sam, I just wanted to say thank you very much for coming here today.

Sam, you can go back to your puppy duties if you wish as well.

Thanks.

Yeah.

And congratulations again, Sam, for your paper that's made the top top 10.

It's given Tom and I something to aspire to for next year, so.

Oh, thank you, but no, congrats to all the authors.

You know, we're thrilled that we get such good, quality research through the bdogp.

And thanks to the readers for accessing all this great research.

Great, yeah.

So thanks very much for your time here and thanks for everyone who's listened to our podcast today.

As I said, I'll put the links to all the papers that we discussed here on the podcast on BJGP Lifetime.com on the podcast show Notes, but just wanted to highlight that if you're interested in hearing more about current research that is being done in the UK and in primary care, please do come and join us at the BJGP Research Conference, which is being held on the 20th of March this year in Bristol.

I think all three of us will be there this year and the conference website is up@bjgp.org conference, so hopefully we'll look forward to meeting some of you there as well.

But yeah, thanks again.

Thank you.

Bye.

Bye.