BJGP Interviews

Today, we’re speaking to Dr Jamie Scuffell, GP and NIHR In Practice Fellow at King’s College London.

Title of paper: Patterns in GP Appointment Systems: a cluster analysis of 3480 English practices

Available at: https://doi.org/10.3399/BJGP.2024.0556

GP practices in the UK are using a wide range of different appointment systems to meet patient demand and improve access. This cluster analysis of NHS appointment data from 56 million appointments and 3480 English practices demonstrates two predominant models of primary care delivery. ‘Same day’ practices tend to fulfil appointments on the same day using GP telephone consultations. ‘Routine care’ practices tend to employ non-GP staff members offering face-to-face appointments and longer appointment wait times. ‘Same day’ care practices had younger and more urban populations.

Episode transcript

This transcript was generated using AI and has not been reviewed for accuracy. Please be aware it may contain errors or omissions.

Speaker A

00:00:00.640 - 00:00:54.360

Hello and welcome to BJGP Interviews. I'm Nada Khan and I'm one of the Associate Editors of the bjgp. Thanks for listening to this podcast today. In today's episode, we're speaking to Dr.

Jamie Scuffle, who is a GP in South London and an NIHR In Practice Fellow at King's College London. We're here to talk about a really topical issue in his new paper here in the bjgp.

The paper is called Paper Patterns in GP Appointment A cluster analysis of 3,480 English practices.

So, hi, Jamie, it's really great to meet you and talk about this work, I guess, really just to start, as you point out in this paper, each practice has their own systems and strategies to manage appointment booking. But how do you think that this impacts on access and patient appointment booking in each practice?

Speaker B

00:00:55.000 - 00:02:17.300

Yeah, it's interesting because I think, as you say, appointment systems have developed even further, really, since COVID and we've ended up with this a quite interesting diverse range of implementing appointments across the country, across England at least.

And I suppose the things that have changed are, you know, if you phone up a practice now, actually, you might not even phone them up, you might submit an online consult, you might be triaged, you might see not a gp, but a range of other professionals as well. And also it might not be done face to face, it might be done by telephone or online.

In fact, there's a new appointment system range of things that have happened across England, and actually there's some evidence that that might relate to access in some ways. So we know lots of people who don't speak English struggle to navigate that system of getting an appointment, for example.

And we also know from the qualitative evidence that there's some digital exclusion as well with appointments. So, yeah, so I think there's lots of issues with access and how that relates to appointment systems.

And so far what we've done is looked at components of the appointment system and how that then affects access. But what we haven't really done much of is looking at the appointment system as a whole and how that might affect access.

Speaker A

00:02:18.180 - 00:02:33.200

Yeah.

So in this study, you wanted to look at patterns of primary care delivery in English GP practices, and you used this Appointments in General Practice data set. Can you tell us just briefly what's available in this data and what you were looking at here?

Speaker B

00:02:33.360 - 00:03:20.700

It's a tremendous data set and I think could be very useful. So every English GP practice for every day of the week publishes up to NHS England.

The appointments that have Happened for the day across lots of different categories, actually, so across whether they've been attended or not, or not attended, whether they're face to face or telephone or home visits or online also if they're GP or non GP appointments. And the sort of. The real clincher that is brand new is looking at appointment lead times as well.

Whether these appointments were booked same day or whether they were booked more in advance when the data is good enough, which is an. If it's a very, very useful data set.

Speaker A

00:03:20.780 - 00:03:43.720

Yeah, so you looked at data from about three and a half thousand practices this year. So as you mentioned, you weren't able to look at all practices due to quality issues.

But just talk us through what you were looking at here and I guess, yeah, just start us off with telling us what you found and perhaps we'll get into how you group the practices as well into clusters. But yeah, tell us a bit more about what you found here.

Speaker B

00:03:43.720 - 00:05:23.440

What we wanted to do was take a set of measures, I suppose, of an appointment book.

So for every practice we said, well, actually we might define their appointment book by the proportion of people who see a GP or the proportion of people who have a telephone consult, or the proportion of people who are booked same day and seen same day. And we kind of came up with actually about 12, in the end, 12 measures of an appointment book.

What we then wanted to do was kind of group together practices with similar characteristics and we picked two different types. I suppose these were the two poles at either end of the spectrum.

And the two poles are that there's a more traditional, what we've called a routine care group of practices, and this is about two thirds of the English practices that we included.

And the appointment characteristics they had were they are more likely to book in advance appointments rather than booked on the same day, more likely to be face to face appointments, interestingly, more likely to use not just GPs but also non GP appointments for delivering care. So they're the big categorization of the routine ones.

So longer wait times, more likely to use non gps and more likely to have face to face appointments.

And then I suppose there's the sort of the more the newer style of appointment system, which we've called the same day appointment system, and that's more likely to be led by GP telephone consults that happen and are booked on the same day. And in this case, same day appointment availability is quite substantially higher than the routine care practices.

Speaker A

00:05:23.600 - 00:05:34.560

And you found that actually there was quite a lot of difference within the practice population and where the practice was based on these two sort of clusters. So the routine and the same day practices as well.

Speaker B

00:05:34.960 - 00:06:59.980

Yeah.

It's absolutely fascinating that it's not just differences in the appointment systems, but actually there are underlying differences, not just in the practice populations, but also in the workforce associated with each of those practices. So the practices that were same day, much more likely to be in urban serving urban populations rather than rural populations.

Also, this might just be an account of. Because they're more likely to serve urban populations, they also serve more ethnically diverse populations.

Interestingly, no big differences between deprivation. And then also the same day practices have a very slightly younger population overall than the routine care practices.

List size is also slightly bigger with those same day practices compared to those that have a more routine approach. Also, differences in workforce.

The number of direct patient care staff, full time equivalents per 10,000 people on the practice books is a bit higher in the routine care cluster than the ones who are delivering same day care.

And the differences are that actually GP levels are pretty much the same, but the routine care cluster employs more nurses and more nurse practitioners and also employs slightly more administrative staff.

Speaker A

00:07:00.540 - 00:07:19.490

Yeah.

So you work as a GP and I just wonder from your own experiences whether what you found in this data reflects what you know or understand sort of on the ground and whether you had any insights from your own work, which might sort of explain why there are these differences, or if you had any thoughts about that.

Speaker B

00:07:19.650 - 00:08:44.140

I did a whole load of locoming when I finished training, actually. It was fascinating to go to a dozen practices over a period of time and look at how the appointment system was set up.

And I think when you're an individual GP behind a door, seeing patients, it's sometimes quite hard to anchor yourself in the wider picture of what's happening at other practices, even the one just down the road.

So I think it's quite interesting to think about how especially the slightly more bigger practices where I work in South London do tend to have had a more of a same day approach to delivering appointments and have also been a little bit more telephone, triage, telephone first in their approach for a longer period of time.

So I think the, I mean, one of the challenges of when you cluster these data is you can cluster into lots and lots of different clusters, but what you're trying to do is make it meaningful to and interpretable to people and practitioners. And I think those two very much.

I can, I can picture practices that operate in those two different ways, even though they actually have quite close in geography together. Quite close in geography.

And I suppose then it's interesting to think about not just how the appointment system is set up, but then kind of patient outcomes as well that might be associated with those two approaches.

Speaker A

00:08:44.620 - 00:09:13.990

One sort of interesting area that some papers have looked at is this sort of balance between quicker access and other outcomes. So, yeah, as you point out, same day access might allow quicker access, but might tend to be on the phone.

And whether that has impacts on other things in the practice, such as continuity of care. And what are your thoughts on that? And the balance shown in this data between sort of maybe quick access versus more routine type care.

Speaker B

00:09:14.310 - 00:10:12.150

I think it's so interesting with the potential opportunities to increase continuity of care are actually potentially higher with more of a same day approach, or at least a same day triaged approach.

I think depending on who you talk to, some people might say, well, actually if you are triaging patients, you could very well increase continuity of care.

But I think our general feeling is that we know same day appointment availability isn't necessarily associated with increased patient satisfaction in the GP patient surveys. And then we also know that patient satisfaction is in some way associated for some people with continuity of care.

So I think there's an argument for saying it could go either way with the same day approach either encouraging or discouraging continuity of care, depending on the context, which I have to say is difficult to get at with these data.

Speaker A

00:10:12.790 - 00:10:32.940

But yeah, another area I was interested in in your data was that the same day access was associated with more urban practices and also a younger population. And I suppose it's just sort of thinking about the patterns of why this might be occurring.

And do you think it's sort of patient driven or do you think it's practice driven?

Speaker B

00:10:33.500 - 00:12:18.490

That's really interesting. I mean, we know. I think it's probably a bit of both. Let's start with practice driven.

So we know that some of the ethnographic work demonstrates that practices organize appointment systems not just around clinical need, but also around demand. And they might not be the same.

And also there are other components, as we've said, you know, there are workforce differences between these two practices as well, which may have come about as a result of the differences in appointment systems, or it may just be necessity that there's lower employment of GPs in these more routine traditional type approaches and therefore there's employment differences there. So the appointment system may have come about as a result of those practice factors.

I think patient level factors are really interesting and I think that's where the work should go next. Although we know that the same day care type Practices do have a slightly younger population.

What we don't know from these data is who is consulting more, who is consulting less.

It might be that actually these same day practices do just as good a job or even better job at responding to clinical need once we take into account the consultation rates between two groups.

So I think it'll be really interesting to try and look at these patient level data and look at the experience of a single patient with a particular characteristic who's, who's subject to different types of appointment systems and then see if that does really affect outcomes. I think that's, that would be a really interesting thing to do next.

Speaker A

00:12:18.970 - 00:12:47.610

Yeah, so I guess that's sort of thinking about the impact of these different practice systems because we know that practices may decide to adjust their systems based, as you said, on their staffing or their patient population. But I guess as you mentioned, we don't really know what the impact is on for patient satisfaction.

Or do you think there's some way you could match this to the GP patient survey or the GPPS survey to sort of look at satisfaction as well?

Speaker B

00:12:48.250 - 00:14:04.400

Absolutely. Next step, I think is to try and do that. Yes. And if anyone's interested in any of this, I'm always really happy to speak to people, to collaborate.

We've had some, few really good papers recently looking at the association between appointment data and patient experience from single components. So looking at same day appointments particularly and demonstrating that increasing.

This is Patrick Birch, I think, and teams work in Manchester looking at the fact that if people have increased same day appointments that there might be a reduction in patient satisfaction and also scores of access and continuity in the patient survey.

What would be really interesting is to try and look at this a little bit more causally, I think, and ideally identify where practices have changed deployment system and then look at corresponding changes in satisfaction.

So I think one of the challenges is this is all ecological data and it's easy to draw strong conclusions from those data without really understanding the underlying mechanisms of what's happening. And as we've spoken about, there are lots of things going on in here.

Speaker A

00:14:04.400 - 00:14:07.800

Any other key findings you want to highlight from this, this paper?

Speaker B

00:14:08.120 - 00:14:47.090

Well, I suppose only just briefly to talk about the administrative differences between the two groups.

I think it's really interesting that even though there might be a whole mechanical change in the same day approach practices and you know, it demonstrates that these practices are, they've got a bigger list size but they've also got potentially some administrative economies of scale as well in managing, in managing the appointment load that's coming through.

There's a bit more work to understand what really is happening with the mechanics of these practices compared to the ones that have a more traditional approach.

Speaker A

00:14:47.650 - 00:15:07.460

So it sounds like, yeah, there's a lot of more that we could do in the future in terms of research to understand what's going on. But from this paper, do you think that there's any nuggets that practices on the ground could pull from this research?

And I think you've mentioned some of it just about sort of understanding their own appointment book and things. But do you have any thoughts about how this paper and the results could be used?

Speaker B

00:15:07.940 - 00:16:25.240

I think it's really interesting to see the diversity and the grounding of your practice versus other people's other practices. And so I think that's one piece is just really understanding where you sit compared to others.

I mean, the other real thing that would be great to understand a bit more is we spoke about the appointments in general practice data quality, and the data quality is limited really by the fact that when we look at our appointment books in gp, they're multicolored and all of the colours correspond to a particular category and those categories aren't well matched up to NHS England categories.

And although we did do some work to show that the practices we did include and didn't include looked much the same in terms of demographic characteristics and practice characteristics, it would be nice to be able to do a full census of appointment systems and if there is a spare five minutes for any practice manager listening, to be able to just categorize those few appointment types will just mean that we can then, you know, help feedback some of these data that are going up to NHS England back down to gps and help to understand a little bit more about, you know, what's happening on the ground and be able to understand these effects a bit more, I think.

Speaker A

00:16:25.320 - 00:16:50.240

And as you mentioned, I think you're absolutely right that at a national level and in terms of guiding policy, decision making and what the best sort of systems are, it's important to have a big picture look at the data. So I think that's really fascinating work, work that you've done here.

But yeah, I think that's been a really interesting conversation around this area and good luck with your future work, but it's been great to chat to you about it.

Speaker B

00:16:50.640 - 00:16:52.040

Thank you. It's been great to chat to you.

Speaker A

00:16:52.040 - 00:17:17.799

Too, and thank you all very much for your time here and thanks for listening to this BJGP podcast. Jamie's original research article can be found on bjgp.org and the show notes and podcast audio can be found at bjgplife. Com.

It's been great to talk again about some of the balancing between access, continuity and patient satisfaction, so do go back and take a read of Jamie's paper. Thanks again for listening and bye.

Today, we’re speaking to Dr Steve Bradley, GP and Senior Clinical Lecturer based within the School of Medicine and Population Health at the University of Sheffield.

Title of paper: General practice chest X-ray rate is associated with earlier lung cancer diagnosis and reduced all-cause mortality: a retrospective observational study

Available at: https://doi.org/10.3399/BJGP.2024.0466

It is known that there is wide variation in the use of chest X-ray (CXR) by general practices, but previous studies have provided conflicting evidence as to whether greater utilisation of them leads to lung cancer being diagnosed at an earlier stage and improves survival. This observational study analysed data from the English national cancer registry on CXR rates for individual general practices, along with stage and survival outcomes; it found earlier stage at diagnosis and improved survival for patients diagnosed with cancer at practices that used the test more frequently. Increasing use of CXR by GPs for symptomatic patients, particularly by focusing on practices that use the test infrequently, could improve lung cancer outcomes.

Transcript

This transcript was generated using AI and has not been reviewed for accuracy. Please be aware it may contain errors or omissions.

Speaker A

00:00:00.640 - 00:01:06.820

Hello and welcome to BJGP Interviews. I'm Nada Khan and I'm one of the Associate Editors at the Journal. Thanks for taking the time today to listen to this podcast.

In today's episode, we're talking to Dr. Steve Bradley. Steve is a GP and senior clinical Lecturer based within the School of Medicine and Population Health at the University of Sheffield.

Early diagnosis of cancer has been an area of research that is Steve's real strength. And we're here to discuss his recent paper here in the BJJP titled General Practice.

Chest X Ray Rate is Associated with Earlier Lung Cancer Diagnosis and Reduced All Cause Mortality A Retrospective Observational Study. Hi, Steve, Great to speak again and to talk through this paper.

I suppose I want to start by saying that, yes, we know that earlier diagnosis of cancer is a good thing because it can lead to earlier stages of diagnosis and treatment. And you start the paper with a short discussion about screening for lung cancer.

But talk us through why this, this alone won't solve delays in lung cancer diagnosis and what else we need to be doing.

Speaker B

00:01:07.540 - 00:02:14.620

So, yeah, this context is really important because screening is a hugely important development and the UK has led in many ways on lung cancer screening using low dose ct. And this, we hope is going to be very, very beneficial for patients.

But it would be a mistake to think that this is going to solve the problem of lung cancer. And there's a few reasons for that.

One is that only about half of people who get lung cancer would have been eligible for screening because screening concentrates on the highest risk population. And also we know that only about half of people who are invited for screening actually choose to participate in screening.

So the upshot for general practice really is that most patients are still going to be coming through by symptoms and in the same way.

So screening is good news in terms of lung cancer detection, but we still need to do as well as we can in terms of picking these patients up through symptomatic pathways. And actually, this is something we touched on in an editorial for BJGP about a year or 18 months ago, I think.

Speaker A

00:02:15.020 - 00:02:20.300

Yeah. So talk us through that. What was that editorial focusing on? Just for people who may not have had a chance to read it.

Speaker B

00:02:20.620 - 00:03:10.660

So it really was really discussing the situation where we are now in terms of awaiting for a national screening program for lung cancer screening and also considering the role of general practice.

So we set out that, just as I've said, that the role of gps is still going to be very important for lung cancer detection, but also that there are certain considerations that are important for GPs in terms of understanding what the program is, because a lot of patients might come to us to talk about lung cancer screening.

So it's good for us to have a basic understanding of what's involved and also some issues around the data that lung cancer screening uses, particularly smoking status.

So it becomes particularly important for our smoking records to be as accurate as possible because a lot of decisions around eligibility for lung cancer screening may. May hinge on that.

Speaker A

00:03:10.740 - 00:03:20.340

And just talk us through. So what were you trying to do in this paper?

So in this paper you were looking at people sent for chest X rays in different practices, but talk us through why you wanted to look at this.

Speaker B

00:03:20.980 - 00:04:06.250

Yeah, so this, this study was really inspired by earlier work which looked at rates of endoscopy requested from general practices and how that might affect outcomes for upper gastrointestinal cancers in terms of. Of when they are detected, what stage they are detected at.

So One of my PhD supervisors, Matt Callister, had had this idea for this project, I think, going back around 15 years or longer, as to whether we could look at practices in terms of how much they use chest X ray, and then look at what happens to patients who are diagnosed with lung cancer, in terms of what stage of lung cancer they are diagnosed with, when they are diagnosed, and also with their survival as well. So that's really what we aim to do in this paper.

Speaker A

00:04:06.490 - 00:04:17.050

Talk us through just briefly what you did and just. Yeah, it was quite a big study. But yeah, just briefly, how did you go about doing this?

Because you looked at quite a lot of data, didn't you, to try to look at these different associations?

Speaker B

00:04:17.849 - 00:05:13.860

So we took data on general practices from 2013 to 2017. So this is general practices in England. And we used the kinds of data that's available on general practice profiles.

That website is also known as fingertips. And we got information on how often different general practices were requesting chest X ray in a year from the Diagnostic Imaging Data set.

And then we also got data on lung cancer outcomes from the National Cancer Registry from the year after. So 2014 to 2018.

So we put those together and we had Data on around 160,000 patients diagnosed with lung cancer in that period and information on general practices. Around 7,000 general practices.

Speaker A

00:05:14.500 - 00:05:23.780

Let's go to what you found here. So what was that association between the rate of practice chest X rays and stage of cancer diagnosis? What did you find here?

Speaker B

00:05:24.520 - 00:07:23.330

So what we did was we broke up practices in terms of how often they were requesting chest X rays, and we did that in two ways. One was in five groups into quintiles and that was adjusted based on factors like demography of the practice, smoking status, et cetera.

And then we had another set of categories which was just based on what we call natural frequency. So just numbers that weren't into three categories that weren't adjusted.

And the purpose for that was we wanted to be able to have a way that people in practices or who are working in the health system could just eyeball figures and get a sense of where practices were and how this might affect outcomes. So we had those different categories.

And for the quintiles we found that practices in the top quintile of chest X ray requesting had both improved stage of diagnosis. So we find an odds ratio of 0.87 favoring early stage diagnosis. So that's stage one or two compared to late stage, stage three or four.

So an odds ratio of 0.87. So that's, that's a really quite substantial improvement. And also improvements in survival.

So hazards ratio of 0.92 favoring one year survival for that top quintile, 0.95 for five year survival as well. And that five year survival that's using only patients who survived to at least one year.

So that's, that improvement isn't just a reflection of the improved one year survival. So we feel this is really quite important.

The other categories with the three different groups that what we call the natural frequencies, we didn't see the quite the same scale effect in the top grip, the top third group, but that's, that's really probably a dilutional effect because they're broader categories. So the top group isn't showing us the same scale of effect.

Speaker A

00:07:23.650 - 00:07:39.410

And you've sort of alluded to this, but you know, each practice will have its own specific population and demographics. Was there anything at a practice level that influenced the rate of chest X ray requests or stage of cancer diagnosis or survival?

Speaker B

00:07:40.740 - 00:08:39.100

So in terms of how often practices request chest X rays. So we've looked at this previously in a paper published in bjgp.

It was called something like association of chest X ray rate and general practices and populations. And what was surprising just was really how minimal the effect of any differences at all are and recorded characteristics between general practices.

So I think in its entirety what we looked at, all of the factors, including differences in populations and practices, accounted for less than 20% of the variation. So most of the variation that's happening is not from things that we can record or understand.

Probably most of this variation is to do with human beings and cultures and what we believe about chest X ray and how valuable we think the test is and adjust our habits and things like that. And that's important because those things can be changed and we can influence those things.

Speaker A

00:08:39.900 - 00:09:03.270

Yeah. And I think that's sort of where I was going to go next, really. And I guess the question is why?

So why would practice level, sort of rates of chest ray, chest X ray ordering impact on lung cancer diagnosis and survival? And I know that the data here might not have answered that question, but what are your best guesses about this?

And you've alluded to this a bit in terms of human factors.

Speaker B

00:09:04.150 - 00:10:40.920

Well, I mean, I think the mechanism this would be working is that if people are doing more, they're taking the opportunity to organize more chest X rays for these very common symptoms. And if you look at the NICE criteria, which they say we should consider an urgent chest X ray, they're really very broad, common symptoms.

Things like cough, shortness of breath, weight loss, chest pain, also raised platelet count, tiredness.

So really symptoms that people mention all the time, People might mention this as an aside, or they might mention it during a chronic disease review or something else.

So there is probably flexibility in terms of what primary care teams do, in terms of what they do with those kinds of disclosures, whether they organize tests like chest X ray or not. So lung cancer is challenging because it usually presents with symptoms which are very common, very non specific.

For example, a cough is the most common symptom, but cough is a very common symptom in general.

So our thinking is really that if teams are more vigilant about how they investigate these common symptoms with chest X ray, they'll be picking up disease earlier.

It's important to say there are limitations with chest X ray, but I think this evidence really gives us some grounds to say we should should use the test, even understanding that there are limitations in terms of accuracy. And although it isn't always successful in picking up lung cancer, it does do it a fair amount of the time and we can use it effectively.

Speaker A

00:10:41.560 - 00:11:07.350

So we both used to work in Leeds where there used to be an open access chest X ray clinic or a self request chest X ray service. So this is where people aged 40 and over could, with symptoms potentially suggestive of lung cancer, could just walk in and request a chest X ray.

Do you think that services like that should be made more widely available if more chest X rays potentially could lead to earlier diagnosis?

Speaker B

00:11:07.750 - 00:12:58.690

Yeah. So this is a self request chest X ray service. So not to be confused with open access which tends to be used for the way that we request chest X rays.

You know, you request, the GP requests it on the computer and then the patient turns up within two weeks, say, and they're able to just get it at their convenience. So, yes, self request services have been used in Leeds now for well over a decade and also are being used in Manchester and elsewhere.

So, yes, I do think these could be used more widely and we know that they are successful in reaching the right patients, patients who have a history of smoking and patients from less affluent communities as well.

And we know also that the proportion of these, these chest X rays that are leading to cancer diagnosis is around equivalent of what gps request as well. I think these services are a good thing, really, because there are patients who find it hard to access general practice to get appointments.

There are patients who also don't want to talk about their symptoms and are worried that they're going to be given a lecture about smoking if they come with respiratory symptoms. And so it just suits some patients.

I think in principle it's a sensible thing to do, but I think, particularly at the current time, where access to general practice is so difficult, or even where it isn't, even where it isn't that much of a problem, patients still have a perception that it is going to be very difficult to get a general practice appointment. So I do think it's a valuable thing to do. And we published a paper in BJGP at the start of this year.

It was recommendations from the Roy Castle Lung Cancer Foundation Group on symptomatic diagnosis. And that was one of the points made in there about expanding these services.

Speaker A

00:12:59.090 - 00:13:19.030

And you've mentioned about some of the limitations of chest X rays, and I know that you've done a lot of work around chest CT as well. And what do you think the role is of a chest ct? And do you think that patients with symptoms suggestive of lung cancer.

Is there a balance between requesting a chest X ray or chest CT in general practice? What are your thoughts about that?

Speaker B

00:13:20.310 - 00:15:30.850

It's difficult.

The guidelines internationally almost all say that for most potential lung cancer symptoms, except for hemoptysis, coughing up blood, the chest X ray should be the first line test. But we know that there are problems in terms of accuracy.

It's missing around about a fifth of cases of lung cancer, which is not something we should be complacent about because this is such a devastating disease and we need to pick it up as soon as possible.

But there are really practical limitations around ct, and particularly in a country like the uk where we just have a lot less access to CT than other high income countries like Australia and the us. So I do think it's a balance actually.

I think it would be a mistake to just give into a council of despair and that we think chest X rays rubbish and isn't worthwhile, particularly when it's going to be difficult for us to get CTs for our patients. But at the same time CT's kind of drawbacks as well, even if we did have perfect access in terms of over diagnosis as well.

So in terms of what we should do practically, this is a kind of classic problem for gps, particularly in countries like the uk, where we have limited access to ct. And in theory all English gps now have access to urgent direct access ct.

I think it's probably more complicated on the ground and I'm not sure if that theory has translated into practical reality for a lot of GPs working in England. So I think GPs really just need to use their intuition quite often. I said just use. It's actually a really difficult thing to do.

But it depends, in short, it depends on how worried you are, you are about your patient and how concerned you are. And also it is the case that a lot of these symptoms overlap with other serious conditions, not just cancer.

So even if you do get the perfect test that rules out lung cancer, the job isn't over there. You, you probably do need to think about other serious conditions as well.

Speaker A

00:15:31.650 - 00:15:48.630

Fair enough.

And yeah, we could sidestep into whole discussion here about so called gut feelings and when clinicians feel inclined to make certain decisions based on that intuition, that clinical intuition as you describe, which I think is a better way of conceptualizing gut feelings. Really.

Speaker B

00:15:48.870 - 00:16:14.720

Yeah, I mean I think we could be, we could be frustrated by this and, and, and want clearer guidance and clearer evidence at the same time. This is really our job as clinicians and it's something we should take pride in and how we think through these problems.

And this is why, this is why we're, we're here. So it is, it is one of the difficult aspects of the job, but it's also an aspect of the job we should take pride in as well, I think.

Speaker A

00:16:14.800 - 00:16:26.080

And Steve, this is sort of, you know, your, really your area of focus and research and knowledge, but is there anything else you want to add here about chest X rays in general practice? Just before we wrap up?

Speaker B

00:16:26.240 - 00:17:23.730

I think the take home here really is the chest X ray is a useful tool. The radiation dose is negligible. It's equivalent to a few days of natural exposure to radiation and the test is useful.

So if the possibility of lung cancer is crossing your mind, I think a good first step is doing a chest X ray.

And it's worthwhile knowing what just having the odd glance at what the NICE NG12 symptoms for possible lung cancer are because it's really surprising how broad these are. And a lot of our patients will come to us with these, with these symptoms.

The other thing is that an increasing...

Today, we’re speaking to Professor Martijn Schut, Professor of Translational AI in Laboratory Medicine and Professor Henk CPM van Weert, GP and Emeritus Professor of General Practice, both based at Amsterdam University Medical Center.

Title of paper: Artificial intelligence for early detection of lung cancer in GPs’ clinical notes: a retrospective observational cohort study

Available at: https://doi.org/10.3399/BJGP.2023.0489

In most cancers, the prognosis depends substantially on the stage at the start of therapy. Therefore, many methods have been developed to enhance earlier diagnosis, for example, logistic regression models, biomarkers, and electronic-nose technology (exhaled volatile organic compounds). However, as most patients are referred by their GP, who keeps life-long histories of enlisted patients, general practice files might contain hidden information that could be used for earlier case finding. An algorithm was developed to identify patients with lung cancer 4 months earlier, just by analysing their files. Contrary to other methods, all medical information available in general practice was used.

Transcript

This transcript was generated using AI and has not been reviewed for accuracy. Please be aware it may contain errors or omissions.

Speaker A

00:00:01.600 - 00:00:55.370

Hello and welcome to BJGP Interviews. I'm Nada Khan and I'm one of the associate editors of the journal. Thanks for taking the time today to listen to this podcast.

Today we're speaking to Professor Martin Schutt, who is a professor in translational AI and Laboratory medicine, and Professor Hank Vanwort, GP and Emeritus professor in General Practice, who are both based at Amsterdam University Medical Center. We're here to discuss their paper, which is titled Artificial Intelligence for Early Detection of lung cancer in GP's clinical notes.

So, yeah, it's great to see you both here today. And Martin, I'll come to you first.

I suppose we know that it's important to try and diagnose cancer early, but could you talk us through what's the potential for artificial intelligence here in terms of identifying cancer earlier based on patient records?

Speaker B

00:00:55.810 - 00:01:52.220

Yeah, that's a very interesting question because the potential kind of like goes hand in hand with the huge amount of interest in AI. And I think there are great opportunities. There are also great challenges.

But talking about the opportunities, especially in the context of the article that we wrote, is on the data side. So on the data side, the digitalization of electronic health records gives great opportunities.

A lot more is digitalized, and that means that we also, in our case, have access to free text, and that we, with the advent of the large language models, with also new developments in AI, we also have better ways of making use of those data. So those two combined creates a really interesting formula for big opportunities for AI in the general practice and healthcare in general.

Speaker A

00:01:52.300 - 00:02:05.960

And you mentioned access to free text records. So what GPs are typing into the record records?

But before we get into the study, can you just briefly describe what is natural language processing and how that can be used in free text records?

Speaker B

00:02:06.760 - 00:03:10.100

So we know that a lot of clinical risk scores, they work with features of patients, so their age and their gender or sex. And. But of course, a lot of information is also written up in unstructured way. And in our case that is text.

But we can also think of images and audio, and in that sense we have access to that data by different ways, which natural language processing is one of them. And it means that we give AI access to this text through, for example, advanced models like we now have, like ChatGPT users.

But that's only one extreme of the spectrum that we can talk about, because you could also imagine that we just simply look with keywords through the text, and then if certain keywords were mentioned, that you include that in the information that is available to your Docu to your, to your model.

Speaker A

00:03:10.260 - 00:03:18.820

And Hank, I don't know if you want to comment on just what we know already about clinical scoring systems for early diagnosis of cancer.

Speaker C

00:03:19.140 - 00:04:21.310

The problem with what we already know is that we know things because they have been coded in the past. If, if you look at the ways to access data, the only way to access data was by using codes.

And the big jump forward is made by using not only codes, but also text, because codes will always be replicating themselves.

By which I mean that a GP who likes to, to have to make notes of what he has been speaking about with patients, he cannot code all the things that he will write down.

So codes will always form a very exquisite extraction of the content of a consultation and will never present us with new information because codes only exist when the information was already there. Otherwise there will be no codes. Just so implicitly there is be a replication of what we know when we have to code our things.

Speaker A

00:04:21.899 - 00:04:49.139

Yeah, absolutely.

And I work with a colleague called Sarah Price who's done some research around coding and she's shown in her research that clinical coding can be biased depending on the outcome. So people who have bladder cancer, they're more likely to have codes for hematuria or blood in the urine.

So, yeah, there could be a discrepancy in how clinicians code things rather than write it in the free text.

Speaker C

00:04:49.139 - 00:05:09.160

Yeah, because in the past there has been done some marvelous research by Willie Hamilton, Hamilton and for example, and Judy Hippisley Cox is well known, but they had to use codes. So there was never a jump forward. And I think that now with the aid of natural language, we can make a jump forward.

Speaker A

00:05:09.559 - 00:05:36.620

And the methods that you use here are quite complex, but I'll try to summarize it briefly.

So essentially you analyze the electronic health records of over half a million Dutch patients and used these natural language processing techniques and machine learning to look back in the records of people diagnosed with cancer. And then you look to see what data in those records could be used to predict lung cancer.

But is there anything you want to add to that, just for a lay audience? Martin?

Speaker B

00:05:36.700 - 00:06:20.170

Yeah, one nuance, a small correction on that is that we don't only look at the patient with cancer, but we look at the cases and controls. So we both look at that because the AI needs to be able to distinguish the case from the controls.

I think that's one important distinction because in healthcare, fortunately, we always have to do with low prevalences. We don't have too many patients compared to the healthy patients. That is Something of what the complexity of these kinds of models is.

I think that is also important to realize when you develop these kinds of models.

Speaker C

00:06:20.810 - 00:06:22.250

May I add something because.

Speaker A

00:06:22.250 - 00:06:22.730

Yes, please.

Speaker C

00:06:22.730 - 00:07:09.230

Because if you look at the, the scientific side of it, then if you develop a prediction model for, for a cancer, for example, then you have to do that with a logistic regression method. And logistic regressions can, can contain many variables, but not as many as you can use when you, when you can use new large language models.

So you can also analyze many more variables. But you can. That's one point. And the second point is that you can analyze those variables in connection to each other.

Great advantage compared to the past. So if you look at the model that we are, we used for this research, I think we use two layers of 100 variables in different relations to each other.

So that gives you 100 times, hundred possibilities.

Speaker A

00:07:09.470 - 00:07:14.630

Talk us through what you did develop here. So what? Talk us through that. Maybe Martin, you can try to explain.

Speaker B

00:07:14.630 - 00:08:21.700

Yeah, Can I start with. So we picked up a signal.

So we develop prediction models taking into all of these, what you said, over half a million patients, all the clinical notes, the consultations that they had, put it in a prediction model. We pick up a signal, we can make a prediction model that can. That performs well. So that's one.

But the second step is that ideally we would also like to get some information from that model. It's like, what do you use to predict what does contribute to a prediction for lung cancer?

And then we come to the nature of the complex methods that we use is that they are black box. We are not able to open them up and see what is in them.

And that is actually, I say, planning forward that we would like to peek into those boxes to see like, what triggers these predictions for lung cancer, which can then be again used in clinical knowledge and independent of the algorithm or the model that we developed.

Speaker A

00:08:21.780 - 00:08:34.980

And the model that you developed actually performed quite well in terms of the sensitivity of the model in terms of distinguishing which patients should be referred for potential lung cancer symptoms.

Speaker B

00:08:36.669 - 00:09:05.149

Correct. I'm going to end that off to Henk.

Maybe just say in between that, when I mentioned predicted performance, I'm talking about the C statistic or the area under the curve, which is the first, how you say, performance criteria. If that doesn't go well, then we should try other things. But that performed well.

And then we translate those indeed into clinically relevant specificity sensitivity. And that's where Hank played a big role.

Speaker A

00:09:05.490 - 00:09:06.450

Yeah, go ahead, Hank.

Speaker C

00:09:06.770 - 00:10:48.000

Yeah. First I'd like to say something about the content of what we found because we did a small exercise to, to discover what was inside the black box.

But that's. Therefore we need much more money to do a good project to, to come up with that. But we found some predictions which were quite astonishing.

The, the thing, two things I, I always tell as an example, and the first thing is that when a GP starts to prescribe incontinence material to a man, then he has a risk for lung cancer, which you can, you can of course explain, because if you have lung cancer, you start coughing and then you start coughing, there is, there is a small chance that you, you wet yourself. And the other thing we found is that the number of slashes which was in the file was related to the, to the risk on lung cancer.

And that was quite a big question for us what that would mean. And at the end we came up with the explanation that there is a connection between lung cancer and cardiovascular diseases.

And that connection is, of course, smoking and GPS always use a slash to note blood pressures. So if you have a lot of slashes in your file, you have a lot of blood pressures noted.

And if you have a lot of blood pressure noted, then you probably will have a high blood pressure, which is related to lung cancer. That are two small explanations of what you find inside the black box, as we now used.

And if you see what's in, you can always think of an explanation, which is the funny thing, of course.

Speaker A

00:10:48.720 - 00:10:58.240

Yeah. So do you think models like this could help clinicians target investigations like chest x rays or CTs in people who might be at risk of lung cancer?

Speaker C

00:10:58.480 - 00:11:59.590

Of course. And why we did this, so is that you can of course use a model like this for a number of applications.

If you use it for a diagnostic, in a diagnostic way, you will have other concerns about your sensitivity and specificity than when you used in, for example, a screening way. If you look at screening, the number of positives will be much lower than when you used in a diagnostic sense.

So it is the way you want to use this algorithm which gives you the decision about what thresholds you will use. We worked out the 3% threshold because that is the referral threshold, which is defined by nice a few years ago.

And if you want to have 3%, then you have to. You need to investigate 33 people to find one with lung cancer.

Speaker A

00:11:59.830 - 00:12:26.810

Yeah. I'm also thinking about sort of the potential practical application of something like this in a practice.

So if you were bringing this sort of tool to a general practice, would you be able to Then suggest sort of what thresholds they would be interested in or what the availability was of certain tools like chest X ray or how do you think that this could be applied in practice? And are there more 10 or. Hank?

Speaker C

00:12:26.970 - 00:12:35.450

Yeah. For example, now if, if I would make, would have to make the choice now I would go for the 3% because that is the advice threshold by Nice.

Speaker A

00:12:36.250 - 00:12:38.010

Martin, do you want to add anything to that?

Speaker B

00:12:38.090 - 00:13:14.510

Yeah.

It's interesting that talking about thresholds, that it is important to realize that these models are not fixed in the sense of you can configure them with a different threshold depending on the evasiveness of a follow up action, the costs of a follow up action, the severity of the disease. So extending this to other diagnosis, to other conditions.

But it's important to realize that these models are kind of like moldable to still use one model in different situations.

Speaker A

00:13:15.030 - 00:13:33.270

And just in terms of applying something like this, how do you imagine it might work at a practice level, at that GP's level? So might it suggest an alert or something if a patient was above a certain threshold to trigger an investigation?

Or how do you envisage this being used in practice?

Speaker B

00:13:34.070 - 00:14:47.990

Could very well manifest as a flagging system. But still looking at bringing a model from theory or from research into practice has a number of steps which in this case still need to be done.

So we took data from three big cities in the Netherlands on which we externally validated models that we used. So we developed the model in one city and then externally validated in the other two.

So that's one big step is external validation, but then also the clinical uptake, setting the thresholds, the technological infrastructure in different GP systems and connections to other systems.

And when you do the updating, that's, that's another big challenge and also the step to maintaining the model afterwards because it's not something that we set and then it's fixed in time.

Of course we have to be open, we have to be aware of the fact that these models need to be maintained and we have problems of drift and the setting might be changed and say that might have different application, how things are registered, which all has implications as to how useful this model remains in practice.

Speaker A

00:14:48.150 - 00:15:41.470

And one thing I wanted to touch on is that you mentioned that these sorts of models will use hundreds of different variables.

And I think the way that a lot of GPS practice when they're thinking about cancer is they're thinking about maybe five to 10 alarm symptoms or red flag symptoms that they're attuned to.

So when their Patient presents with that, they kind of are already thinking, right, I need to be doing something, maybe doing, making referral or ordering more tests.

But in this sort of model, because there could potentially be hundreds of variables, it's more that the system is learning or as Martin says, flagging which patients might need anything further alongside the clinician's intuition or concern about a patient's symptoms as well. So it's in addition to the clinical intuition and thinking, thought processes as well.

Speaker C

00:15:41.790 - 00:16:49.340

Of course, this is very, this will be very disrupting in a GP's mind because he will have to refer patients who, who are not in his mind as at risk. And that's not what we, what we used to do.

I mean, the GP is somebody who would, who calculates the risks for patients and if the risks are low, are low, he will not refer in his mind. And if you don't know how a risk is, is made up, then of course the mind of a GP will be, will be in problems.

Because one thing you have to say, if you speed up the process of diagnosing cancer with four weeks, until now, what we see is that if you speed up surgery for four weeks, there will be a 6% decrease in mortality, which is a huge gain.

So I think that in the end gps will be prepared to accept that the system might be better than themselves, because that's the step you have to accept.

Speaker A

00:16:51.500 - 00:17:10.119

It's really fascinating work and obviously, as Martin has mentioned, there's a lot more work to be done for these AI driven and natural language processing driven models. But it's very exciting and I can already see the application potentially for lots of different cancers and not, not just lung cancer.

So is that where you're heading now with this?

Speaker C

00:17:10.599 - 00:17:25.319

Of course, this project is almost 10 years old now, so we saw, we saw in the, in the start, we saw the potential for, I mean, it's not only for cancer, even also for many other disease.

Speaker B

00:17:26.039 - 00:19:01.850

So in addition to that, indeed, what ankle also just mentioned, there's lots of variety in, in different words. Of course, let it be said, the different languages is also a challenge.

If you challenge, if you look at texts, which is also something we have to tackle technically for the different models and approaches that we have, but also clinically that these words have different meanings.

And then also what you say is like, yes, this was for lung cancer, we did similar work for

Today, we’re speaking to Dr Afrodita Marcu, a Research Fellow in Cancer Care at the University of Surrey.

Title of paper: Symptom appraisal and help- seeking before a cancer diagnosis during pregnancy: a qualitative study

Available at: https://doi.org/10.3399/BJGP.2024.0208

There is a gap in current understanding about the experiences of women diagnosed with cancer during or around pregnancy including how they appraise and seek help for cancer-related symptoms. This qualitative study found that women and healthcare professionals often interpreted symptoms through the lens of pregnancy, particularly when symptoms were vague. Health professionals need to ensure full assessment of symptoms, timely referral, and effective safety-netting for these women.

Transcript

This transcript was generated using AI and has not been reviewed for accuracy. Please be aware it may contain errors or omissions.

Speaker A

00:00:01.040 - 00:01:04.650

Hello and welcome to BJJP Interviews. I'm Nada Khan and I'm one of the associate editors of the bjgp. Thanks for listening to this podcast today. In today's episode, we're speaking to Dr.

Aphrodite Marcoux, a research fellow in Cancer care at the University of Surrey.

We're here to talk about a paper she's recently published here in the bjgp, which is titled Symptom Appraisal and Help Seeking Before a Cancer Diagnosis during Pregnancy, A Qualitative Study. So it's really lovely to meet you, Aphrodita, to talk about your research.

And I think this is a really important area and I wanted to get your thoughts on just why this area is so important to research.

But I think that most people will probably appreciate that during pregnancy, women's bodies are going through lots of changes, so it can sometimes be difficult to know what's normal and what's not. But talk us through why you wanted to do this study.

What are the challenges faced by patients and by doctors around cancer diagnosis in women who are pregnant?

Speaker B

00:01:05.050 - 00:02:32.190

It's an important area to research because the symptoms of pregnancy, the bodily changes that naturally occur during pregnancy, can mask the symptoms of cancer, both for the women experiencing them, but also for the healthcare professionals with whom they come into contact and with whom they share the symptoms.

So it's an important area to research from that point of view in terms of understanding the potential causes for delay in receiving a cancer diagnosis.

And one of the areas which we discovered was less researched was early diagnosis or timeliness of diagnosis of cancer diagnosis in the context of pregnancy.

So we conducted this research because there was a lack of research, especially in the uk, on women's pathway or other pathways to a cancer diagnosis and pregnancy.

And we wanted to get a more detailed understanding knowledge of how women make sense of their symptoms during pregnancy, how they seek help and why to whom they present, midwife, gp, other healthcare professionals and how they receive a diagnosis. What is the pathway to a cancer diagnosis and pregnancy?

We wanted to get more clarity about that, more detail, and we were also interested to find out whether delays characterize this pathway to a cancer diagnosis in pregnancy, be they patient related delays or healthcare system related delays.

Speaker A

00:02:32.830 - 00:02:44.750

And I guess you've touched on this. So does do we know if there are delays in cancer diagnosis amongst women who are pregnant?

So do they tend to have a later diagnosis than women who aren't pregnant?

Speaker B

00:02:45.540 - 00:03:15.770

International research suggests so. For example, in relation to breast cancer, which is the most frequent cancer diagnosed during pregnancy?

Women are typically diagnosed later or later stage breast cancer than women who are not pregnant of similar age groups.

So there is some indication that in the body of literature around cancer in pregnancy that women might receive a later diagnosis compared to women who are not pregnant.

Speaker A

00:03:16.570 - 00:03:56.710

And this was a qualitative research study. So your team talked to women diagnosed with cancer during or soon after pregnancy and you talked to 20 women.

And I really wanted to focus on what you found here. And you used an interesting model here.

So you map the findings of your research onto a commonly used framework in cancer diagnosis, the models of pathway to treatment. So you've talked about sort of these pathways that women experience during their cancer diagnosis journey during pregnant.

And I wanted to just talk through some of the findings. So when women were pregnant, how did they interpret their symptoms that may or may not have been related to cancer?

Speaker B

00:03:58.390 - 00:06:22.420

Well, they interpreted these symptoms in various ways and that depended on the type of cancer they had, but also the type of symptom or symptoms that they experienced.

So, for example, we had a couple of women with colon cancer, bowel cancer, one of them had abdominal pain and bloatedness, bloating, whilst the other one had blood in the stool. So obviously the symptoms were interpreted differently by the two women.

The one with blood in the stool saw this as incongruent with pregnancy because it's very unusual to have blood in stool during pregnancy, and looked up her symptoms online, did some reading and decided that these symptoms warrant medical attention. So she contacted her GP fairly promptly.

The other lady with more diffuse symptoms with abdominal pain and bloating repeatedly mentioned these symptoms to her midwife, but these were put down to pregnancy. And she always, when she actually engaged in information seeking online, she always added in pregnancy.

So she saw these symptoms as being part of the pregnancy, was her first child, and thought it's normal to have abdominal pain or to feel bloated during pregnancy. So she was actually diagnosed quite late with stage 4 colon cancer.

Of course, we are not drawing cause effect conclusions because this was a qualitative study, but we do find that the symptoms, which are vague, are more likely to be interpreted as being signs or symptoms of pregnancy, as being pregnancy related bodily changes.

So it depends on the type of symptom that people symptoms that people experience, but also depends on the type of cancer they have, because some cancers, such as breast cancer, are more people are more familiar with breast cancer, let's say, than with lymphoma or other types of cancer, and they're more likely to see breast lumps which are typical symptom of breast cancer as being indicative or being suggestive of breast cancer, and it makes them more worried and more inclined, speak to a healthcare professional to see their GP promptly. So for women who experience breast lumps, the pathway was in a sense clearer.

They knew who to contact and also the healthcare professional in question, the GP was more likely to make referrals to tests and secondary care, refer them to the breast clinic.

Speaker A

00:06:22.980 - 00:06:30.180

I think it was interesting in the paper you described this as people viewing their symptoms through the lens of pregnancy.

Speaker B

00:06:30.620 - 00:07:20.420

Yes.

So the pregnancy offered for the women sort of heuristic and also in competitive lens, they understood their symptoms in the context of the pregnancy.

Of course, people, women expected their bodies to change during pregnancy, such as breasts becoming larger or various lumps appearing, perhaps being related to milk ducts and things like that. So they always thought of pregnancy as being the reason for their bodily changes rather than cancer.

So again, we have different patient histories, different types of symptoms, and this leads to different ways or different pathways to cancer.

Diagnosis in pregnancy for some people can be a very straightforward pathway, for others, it's more prolonged, one more complicated one because of the types of symptoms they experience.

Speaker A

00:07:21.380 - 00:07:36.880

And I guess the flip side to this is what the healthcare professionals did with women coming to them with symptoms during their pregnancy.

And this is all from the perspective of the women you talked to in the study, but what did they feel that the healthcare professionals thought of their symptoms when they presented to them?

Speaker B

00:07:37.120 - 00:09:06.970

Well, I suppose it varied from individual to individual.

In some cases, the healthcare professionals thought that the symptoms were due to pregnancy, saying it's hormonal, just hormonal changes, nothing to worry about.

In some cases they thought that the baby was lying on a nerve, which explained why people had symptoms such as back pain or breathlessness or other symptoms, in some cases, the gps thought it is hormonal change, it's probably due to pregnancy, but I will refer the patient to a breast clinic because the system is there in place, it's better to rule it out quickly. So some women had prompt referrals to breast clinics for breast related changes, such as breast lumps.

In some cases, there were some other contextual factors which led the GPs or other healthcare professionals to interpret the symptoms.

For example, one lady had bleeding from the nipple and that was explained as being caused by her child jumping on her or just causing some sort of physical damage. So sometimes there were some other factors which helped explain the symptoms and why they occurred.

But in terms of how women reported their interactions with the, with the gps or sometimes with the midwives, I think this actually varied, but according to the symptoms and to the person they were talking to and also their personal histories as well in terms of having previous breast related changes or having a previous diagnosis of breast cancer.

Speaker A

00:09:07.210 - 00:09:43.030

So I guess I'm hearing two different things. One is that it depended if the symptoms were congruent with pregnancy or not, so that sort of affected the patient and clinician side.

And then another thing you've mentioned a few times is the ease of referrals. And for instance, like you were talking about breast lumps, there's a very recognized easy to use system for that.

So I'd imagine that that's an easier pathway to refer to with clear guidelines on what to do. So it feels like those two different areas might have affected patient presentation and also what the healthcare professionals did.

Speaker B

00:09:43.430 - 00:11:43.590

Yes, exactly, yes. So, I mean, our sample was quite heterogeneous.

So the majority of our sample, 13 women, had the breast cancer diagnosis, but the other participants in our study had different types of cancer, such thyroid Hodgkin lymphoma, non Hodgkin lymphoma, colon cancer, malignant melanoma. So because our sample was heterogeneous, it's difficult to think, you know, there's actually one pathway to a cancer diagnosis in pregnancy.

There are many pathways and that depends on the type of cancer, personal history, previous interaction with healthcare professionals, previous presentations for similar symptoms in the past. So there are many factors which actually influence women's and healthcare professionals interpretation of these symptoms during pregnancy.

Yeah, I would say there are many pathways rather than a pathway to cancer diagnosis.

But it's fair to say that across the sample what we saw was that women were inclined, they tended to see their symptoms as being caused by pregnancy or pregnancy was the first thing they thought of, whilst at the same time considering other potential explanations such as cancer.

So for example, women with breast lumps realized that this could be breast cancer, but at the same time wanted to believe they were linked to pregnancy.

So with some sort of kind of dual thinking or dilemmatic thinking, it's not that people thought of one cause, people thought of many causes, but also there was some sort of hope or refusal to believe this could be cancer, because it's obviously frightening time and experience for this woman, especially during pregnancy, whilst they're pregnant.

And it's not necessarily something they wanted to believe in the first place, but they ruled out that the potential explanation, seeing that the symptoms persisted or got worse, such as the breast lump getting bigger or the pain getting worse, I.

Speaker A

00:11:43.590 - 00:11:53.810

Thought it was an interesting pointer that you had in the paper for clinicians, which was the to assess the symptom, not the pregnancy. So I wonder if you could talk through that a little bit.

Speaker B

00:11:53.890 - 00:12:55.420

Yes.

I think what we recommend, based on our findings, is that GPs or other healthcare professionals that women report their symptoms to should see the think of the symptoms and not actually the pregnancy.

So should isolate the symptoms from the pregnancy and think in a person that is not pregnant, what would I think the symptom might be or what tests would I recommend?

So it's about that thinking about it outside the context of pregnancy, so that the focus should be on how that symptom should warrant further investigations and not explain everything as being pregnancy related.

So one recommendation would be for healthcare professionals, GPs to perform physical examinations, unusual blood tests and take into account the clinical history of the women presenting with these symptoms and not describe these symptoms to pregnancy. Just consider that those symptoms in their own right.

Speaker A

00:12:56.060 - 00:13:10.460

And you also talk about the importance of safety netting, I think, and you touched upon this even from the women's perspective about when things were getting worse, that acted as a bit of a nudge for them to come back or to seek medical attention in the first place as well.

Speaker B

00:13:11.350 - 00:14:05.560

Yeah.

So if GP suspect that the symptoms are not cancer, they should agree sort of management plan with the women in questions and give them an idea about how long they should expect those symptoms to last and what they should do if those symptoms persist and encourage them to return if the symptoms not go away, and also tell them what to do if the symptoms worsen.

So they should have safety netting in place for the management of the symptoms and for future help seeking on the part of these women if the symptoms persist. And women should be reassured and encouraged to return if the symptoms do not go away.

So in our study we found that some women represented multiple times to their GP and some of them actually were diagnosed only after emergency presentation.

Speaker A

00:14:06.270 - 00:14:24.110

So that's been a really interesting discussion around this paper and it's been just really interesting, I think, hearing about this research and some really clear messages for clinicians as well. But I think that's a great place to wrap things up. So I just really wanted to say thank you very much for your time. Aperita.

Speaker B

00:14:24.190 - 00:14:24.830

Thank you.

Speaker A

00:14:26.190 - 00:14:51.810

And thank you all very much for your time here and for listening to this BJGP podcast.

I hope you found today's research discussion helpful, not just for your practice, but also in thinking about other research questions coming out from this area. Aphrodita's original research article can be found on bjgp.org and the show notes and podcast audio can be [email protected] thanks again. And bye.

Today, we’re speaking to Dr Stephen Gibbons, Consultant Clinical Biochemist at Leeds Teaching Hospitals NHS Trust, and Dr Clare Spencer, GP Partner and Menopause Specialist at the Meanwood Group Practice in Leeds.

Title of paper: Optimising testosterone therapy in patients with hypoactive sexual desire disorder

Available at: https://doi.org/10.3399/bjgp25X741321

Transcript

This transcript was generated using AI and has not been reviewed for accuracy. Please be aware it may contain errors or omissions.

Speaker A

00:00:00.400 - 00:01:08.824

Hello and welcome to BJJP interviews and welcome to our new season of the podcast. Hope you all had a great break over Easter and thanks again for listening to this podcast today.

My name is Nada Khan and I'm one of the associate editors of the BJTP. In today's episode, we're speaking to Dr. Stephen Gibbons, consultant clinical biochemist at Leeds Teaching Hospital NHS Trust, and Dr.

Claire Spencer, a GP partner and menopause specialist at the Meanwood Group Practice in Leeds. We're here to talk about the recent clinical practice paper published here in the bjgp.

The paper is titled Optimizing Testosterone Therapy in Patients with Hypoactive Sexual Desire Disorder. So thanks, Stephen and Claire, for joining me here today.

It's great to talk to you about this paper, especially because it's in an area of a lot of interest to patients and clinicians in general practice wondering what to do about testosterone prescribing.

I guess I wanted to kick things off, Stephen, really, by asking, what made you start investigating testosterone replacement in patients with hypoactive sexual desire disorder?

Speaker B

00:01:08.952 - 00:03:09.662

So it was actually a conversation with a colleague at work over coffee and she mentioned to me that she'd noted quite a lot of high testosterone in females of a particular age and she was asking why that might be. So I explained it's probably because of TRT in this condition called hsdd, but that was kind of quite anecdotal at that point.

So we thought we'd do a clinical audit. So myself and two colleagues, Kia and eloise, we audited 100 patients from Leeds.

So we looked at a sample of 100 patients on TRT for HSDD and we audited them against the British Menopause Society guidance, which state that you should do a pre testosterone measurement and then you should check at at six to eight weeks, I believe. And what we found is that actually there was quite poor compliance with the BMS guidance. And at this point we felt a little bit out of our depth.

But we thought, well, this is quite alarming. Probably the most alarming thing was the number of patients with a really high testosterone that weren't adequately followed up.

So we thought, right, let's bring some clinical experts in at this point. So that's when we got in touch with Dr. Spencer and Dr. Jasim and Dr. Wal Ford, who's also on the paper.

She's a consultant endocrinologist at Leeds, and we kind of had a look at the data and we all agreed that, you know, there were significant findings. And the question was why?

Because there are quite comprehensive guidance out there from the bms, but I think we all felt that potentially they lacked some of the finer detail. Potentially in some areas they were a little vague. So that's when we came up with these additional recommendations.

And they're certainly not supposed to replace the BMS guidance, but it's a supplementary kind of recommendations to support the BMS guidance. So that's where we started, really.

Speaker A

00:03:09.766 - 00:03:18.014

And I guess if we just dial this back a bit. Can you or Claire talk us through what is hypoactive sexual desire disorder and how common is it?

Speaker B

00:03:18.102 - 00:04:38.868

So hsdd, essentially, it's a condition where they get persistent absence of sexual dis. Desires or fantasies. So you might.

Some people might term it low libido, I suppose, but the difference between low libido and HSDD is that in HSDD there's an emotional component, so emotional distress. And it doesn't just affect women, of course. This affects both males and females. But the prevalence seems to be much higher in females between.

Between about 15 and 20% of females will experience HSDD. In males, it's probably slightly lower, around 5%. And I mean, Claire may expand on this, but we don't actually fully understand the causes, really.

Probably multifactorial. There's certainly associations with physical conditions, things like diabetes and thyroid disorders.

There is an association with hormonal imbalances, estradiol and testosterone, although the evidence is not as strong as one might think for testosterone. Certain medications can be associated with hsdd, things like antidepressants and then psychological issues.

So anxiety, depression and current or previous relationship problems.

Speaker A

00:04:38.964 - 00:04:51.368

And Claire, you are a menopause specialist, and I think the question that lots of people are probably wondering about is, is this an issue amongst women who are going through perimenopause or menopause as well?

Speaker C

00:04:51.564 - 00:06:31.562

Yes, it's an incredibly common condition or symptom of the perimenopause and menopause. And as Stephen said so brilliantly, there are so many reasons behind that. So HSDD is obviously the far more severe end of the spectrum.

But depending on which study you read, anywhere between 40 and 60% will complain of.

Women will complain of low libido in the menopause, and obviously that then needs unpicking as to whether that's the more severe end of the spectrum or incredibly common. And this does happen to men as well as women, I think it's worth calling out. But in the menopause, a very common cause would be final symptoms.

So in the menopause, with the loss of estrogen. Up to two thirds of women will develop vaginal dryness, soreness, irritation, lack of lubrication, painful or discomfort during intercourse.

And that can have a really significant impact then on libido. And so there are some very specific causes related to the menopause.

Also, if we think about all of the myriad of symptoms of the menopause, so including hot flushes, night sweats, lower mood, low motivation, many women gain weight in the menopause. Again, you can see how that then impacts and add to that anxiety, loss of resilience, you know, and the sort of more psychosocial factors.

Plus layer on top of that, often women have been in a relationship for many, many years. You can see that there are additional challenges also. So it's a really common and distressing issue.

Speaker A

00:06:31.746 - 00:06:46.874

And I think the question that maybe lots of gps will have, I think, is what are the current guidelines around using testosterone? And Stephen, you mentioned the BMS guidelines, the British Menopause Society. So what are the current guidelines telling us about using testosterone?

Speaker C

00:06:47.002 - 00:08:47.795

So if we think about the NICE guidance for menopause first, that's NG23 and that has been recently updated, the new Update published in November 2024. And so nice say that testosterone can be used for low libido in the menopause in adequately estrogenized women.

So basically women on hrt, because actually HRT containing estrogen plus or minus progesterogen can be helpful in managing libido. Libido and estrogen definitely has a really important part to play.

But NICE say that testosterone can be added if you've managed the vaginal symptoms, if you've managed menopause symptoms. If women are taking hrt, then you can add testosterone.

On top of that, the British Menopause Society have really helpfully published guidance also, which goes into a little more detail on the practicalities of prescribing and monitoring.

And so the British Menopause Society would recommend that total testosterone is checked as a baseline and then pragmatically at around three months and then six to 12 months after that, again highlighting that this is predominantly prescribed for women on HRT and highlighting the importance of managing as much as you can the other symptoms that might be having an impact on libido.

Also, it can be really difficult because, as we both said, there are so many factors that can impact and sometimes you do have to take more of a pragmatic approach and manage symptoms as best you can. Plus there may be a psychological aspect that needs to be approached through talking therapies plus testosterone on top of it.

So complex issues, complex answers, often multifactorial approach is needed.

Speaker A

00:08:47.907 - 00:09:11.660

And I guess what you've done here, as you mentioned, Stephen, was to develop local guidelines to help clinicians to guide testosterone testing.

And I'd recommend to people listening to take a look at the full paper, which will be linked in the show notes that give the specific guidelines that you've described and developed. But can you give us a bit of a summary of what GP should be thinking about in terms of testosterone measurement?

Speaker B

00:09:11.820 - 00:11:31.210

Yes, I think probably the main point really is we'd certainly seen an increase in the number of advice and guidances from for secondary care about persistently raised testosterone in these individuals and what level should they be aiming for? And the guidance is not quite clear currently, the BMS guidance, what the actual target values are.

Obviously, these will be lab dependent, which adds another layer of complexity to this.

But essentially what we thought we would do is try and look at general levels and say, well, if it's less than 75% of the lab reference range, then TRT could be trialled.

The other issue we get is, once the patient's on trt, what should they do if the level is persistently elevated, we feel that anything above 110% of whatever the lab range is would be too high and they should lower the dose and repeat two to four weeks after and continue that until you've got a level that's within the normal range or just above the normal range for your lab. I think the other thing that we've definitely seen at Leeds is some very, very high levels.

So the normal range that we quote at Leeds is less than 1.8 nanomoles per liter. Most of the patients that we're talking about here with, with higher levels between 2 and 4 nanomoles per litre, which is too high. But we.

We get the odd1 that's 10, 11, 12 nanomoles per litre and we get phone calls or advice and guidance about this. Now, at that level, it's. It's potentially contamination from venipuncture site, although the advice is to put the gel on below the waist.

Often patients will apply it to the arm and if they apply it to the arm and then have a blood sample collected from the arm, you can get contamination from the venipuncture site and that's when you see levels of 10, 11, 12, 13.

Now, if you do see that, the advice, of course, is not to reduce the dose because the dose might be correct, it's to repeat it without the contamination to confirm the dose.

Speaker A

00:11:31.630 - 00:11:42.294

And I guess, as you've mentioned in your clinical practice paper, patients whose testosterone levels are perhaps too high would be experiencing significant side effects, I'd imagine.

Speaker C

00:11:42.422 - 00:13:54.108

I think it's really interesting because, as Steven said, sometimes the level is very high due to contamination and it is quite interesting. You can get.

Obviously, if the level is truly very high, you are more likely to have side effects such as acne, like skin changes, hirsutism, additional unwanted hair, greasy scalp are the most common, although at the more severe end of the spectrum there can be virilisation and voice changes, but because it's frequently contamination and when you recheck it, it can be normal, there may not be side effects and you can actually get side effects from quite small increments of increase. Everybody's very different in their sensitivity.

We know that the blood test for testosterone, which maybe will come on, isn't a perfect reflection of testosterone activity in the body.

The blood test measures the total testosterone, which is the sum total of the inactive protein bound, plus the very, very small free fraction of free testosterone.

And so the blood level of testosterone may not actually reflect activity of testosterone, which is why there's very poor correlation between an actual total testosterone level and clinical symptoms. And it doesn't. The blood level doesn't predict who will respond to testosterone or not.

And I think that the British Menopause Society have been pragmatic in that they've said, yes, measure a baseline just to make sure that the testosterone isn't on the high side before you add more in. And then in monitoring, the aim is always to keep it within physiological limits.

For women, there doesn't seem to be a level that we have to aim for for a clinical response, but we do need to be safe and we do want to minimize the risk of side effects. So it is very difficult because the evidence around this is very poor. Because of the complexity of testosterone activity in the body, I wonder.

Speaker A

00:13:54.204 - 00:14:16.432

I think that quite a few gps might be. Well, I. I know that a lot of GPs are quite hesitant about prescribing testosterone, especially in women during the perimenopause or the menopause.

Claire, do you have any advice for GPs wondering if they should or can be prescribing testosterone to their patients?

Speaker C

00:14:16.576 - 00:15:48.260

Yeah, so that's a really good question. In most parts of the country, and certainly in Leeds, testosterone is an amber drug. So what that means is it's for specialist initiation only.

Now, some GPs are interested in the menopause and they've done additional training and are comfortable initiating testosterone and certainly once testosterone has been initiated by somebody, for example, like myself in the specialist menopause clinic or Dr. Ward in endocrinology, they may then feel comfortable prescribing ongoing.

But if you're prescribing, you're taking responsibility for that monitoring. And I think that's where the difficulty lies and that often GPs aren't confident because they haven't had training in their specialty.

Which is a reason why this paper's so good, because it's so clear of what to do. They're not sure what to do when they get different levels of testosterone. They're not sure to answer patient queries on it.

So the advice for GPs would be prescribe testosterone. If you're comfortable, prescribe. If you're comfortable monitoring.

There has to be a system in place for monitoring and reminding patients that they need these blood tests. And we have a system in place with the specialist menopause clinic.

And if in doubt, ask, you know, always, always prescribe and practice within your comfort zone.

Speaker A

00:15:49.240 - 00:15:56.976

Really clear advice there, Claire. Thank you. Anything else either of you want to add about this area of prescribing or monitoring?

Speaker B

00:15:57.088 - 00:17:03.464

I mean, the only other thing I'd like to touch on really, is the. The SHBG comment from the British Menopause Society. So the British Menopause Society does reference SH measurement.

So, as Claire alluded to earlier, SHBG is the binding protein for testosterone. And what we found in the audit data was that out of 100 patients we looked at, SHBG was only measured about 11 times.

But in only one case did it add any clinical value. And I think there's a significant lack of understanding about how SHBG will actually add any value to the measurement of a total testosterone.

So our advice really is that at the minute, there's probably not enough evidence to routinely measure SHBG in these patients.

Potentially in ones that are difficult to manage or where there's a poor correlation between testosterone concentration and clinical effect, SHBG might be worth measurement measuring. After discussing with kind of the local experts or the duty biochemistry.

Speaker A

00:17:03.622 - 00:17:47.312

I think just hearing that is really bringing to the forefront why this collaboration between yourself as a clinical biochemist and GPs is why this paper is really valuable, because it's bringing that expertise about measurement of testosterone and also the clinical use of it. So that's been really interesting to hear, but I guess, yeah, that's been a really great chat around this area.

And as I said, I hope people will go back to read the paper just for the full details of the guidelines that you have suggested in in that. But as an area that's probably of increasing importance in general practice prescribing, I think it's been a really useful paper to talk about.

So yeah, I just wanted to say thank you very much for your time.

Speaker C

00:17:47.496 - 00:17:49.420

Thank you, thank you.

Speaker A

00:17:50.040 - 00:18:20.220

And thank you all very much for your time here and for listening to this BJDP podcast.

Stephen and Claire's original clinical practice article can be found on bjgp.org and the show notes and podcast audio can be [email protected] I hope you found today's podcast helpful.

I certainly know that it will help me in the future in terms of guiding my decisions around initiating and monitoring testosterone in women, especially around the perimenopause or menopause. Thanks again for listening and bye.

Today, we’re going to do something a bit different and take a look back at the recent BJGP Research Conference, which was held on the 21st of March 2025 in Manchester. I’m going to discuss some of the highlights and really focus on what the conference is about and how to get involved in the future.

Here are some of the links I discussed in the podcast:

Links

https://journals.sagepub.com/doi/full/10.1177/1609406918797475

https://bjgplife.com/write-for-bjgp-life/

Today, we’re speaking to Dr Louise Clarke, a GP and researcher based at the University of Nottingham.

Title of paper: Barriers to diagnosing and treating vulval lichen sclerosus: a survey study

Available at: https://doi.org/10.3399/BJGP.2024.0360

Previous research has identified a significant diagnostic delay and misdiagnosis of vulval lichen sclerosus (VLS), a condition most commonly presenting to primary care. Health care professionals (HCPs) in primary care share the concerns of women with VLS citing frequent misdiagnosis, embarrassment and lack of knowledge as barriers to diagnosis. In this survey, 92.6% of HCPs felt further education would be useful with 37.7% never having participated in learning on vulval skin disease, self-directed or otherwise. Key enablers identified to facilitate timely VLS diagnosis and treatment include: a comprehensive education programme for HCPs, implementation of standardised pathways of care and development of a VLS diagnostic criteria to be implemented in primary care workflow.

Today, we’re speaking to Dr Victoria Tzortziou Brown, a GP and Reader in Primary Healthcare and Health Policy at Queen Mary University of London, and Vice Chair for External Affairs at the Royal College of General Practitioners.

Title of paper: Language of primary medical qualification and differential MRCGP exam attainment: an observational study

Available at: https://doi.org/10.3399/BJGP.2024.0296

To the authors’ knowledge, this is the first study on the association between the language of the primary medical qualification and attainment in the Membership of the Royal College of General Practitioners (MRCGP) examination. It shows that undertaking undergraduate clinical training in a country where the native language is not English can statistically significantly and negatively affect examination performance in MRCGP exams. The study also shows statistically significant positive correlations between Multi- Specialty Recruitment Assessment, International English Language Testing System, and Professional and Linguistic Assessments Board scores and the MRCGP exam scores; this suggests that past performance in these assessments can help with the identification of those international medical graduate registrars who may find tailored support beneficial. 

Today, we’re speaking to Professor Helen Atherton, Professor of Primary Care Research at the University of Southampton.

Title of paper: Supporting patients to use online services in general practice: focused ethnographic case study

Available at: https://doi.org/10.3399/BJGP.2024.0137

Use of, and access to, online services are increasing within general practice in England. Current approaches to digital facilitation as observed in this study, appeared to be ad hoc and fitted around multiple services. Reception staff were key to supporting patients to use these platforms, but training, resources and support for such staff were not readily available. Enabling patients to have the best chance of using online services requires vision, strategy and investment of time and money. As practices and patients increasingly use online approaches to healthcare provision, practices should be mindful of patient groups who may find accessing services online to be a challenge and who thus require targeted help and support.

Today, we’re speaking to Dr Sam Merriel, a GP, and NIHR Academic Clinical Lecturer in General Practice based at the University of Manchester.

Title of paper: Factors affecting prostate cancer detection through asymptomatic PSA testing in primary care in England: Evidence from the 2018 National Cancer Diagnosis Audit

Available at: https://doi.org/10.3399/BJGP.2024.0376

Asymptomatic, informed choice prostate specific antigen (PSA) testing occurs in primary care in the UK in the absence of a national prostate cancer screening programme. This study shows that four fifths of prostate cancers are diagnosed following symptomatic presentation rather than from asymptomatic PSA testing. There is a 13-fold variation in asymptomatic PSA test detected prostate cancer between English GP practices, without clear explanatory practice-level factors. Patient factors amongst men diagnosed with prostate cancer, including ethnicity, age, deprivation, and multi-morbidity, have a significant impact on the likelihood of being diagnosed following asymptomatic PSA testing.