SBIDER presents

Investigating symptom propagation: What are the implications for infectious disease modelling and health economic evaluations? - Phoebe Asplin | SBIDER Presents - Episode 11

Phoebe Asplin discusses recent publications associated with her PhD project, using mathematical models as a tool to explore the extent to which symptom propagation exists and its subsequent impact on infectious disease and health economic modelling assessments. Have feedback? Complete our Google Form: https://bit.ly/3IIgPKH!

Duration:: 13m
Broadcast on:: 31 Jul 2024
Audio Format:: mp3

What is symptom propagation? Does symptom propagation occur for respiratory pathogens? And if so, which ones? When there is symptom propagation, what effects might this have on studies that predict the spread of the disease and its impact on healthcare costs? We explore these questions and more in this episode of Spider Presents. This is a Spider Presents, a series produced by the Spider Podcast Hub. My name is Laura Guzman. And mine is Ed Hill. In this episode, we are joined by Phoebe Asplin, a PhD student from the Amatix for Rural Water Systems Centre for doctoral training, which has the acronym MASSIS, based at the University of Warwick. We will be discussing recent work done as part of Phoebe's Phoebe's Phoebe project, investigating symptom propagation and its potential implications for modelling assessments. This project is a collaboration between members of Spider at the University of Warwick, with myself and Matt Keating as co-supervisors, a University of Glasgow supervision by Rebecca Mansey, and external partners to UK Health Security Agency, which has the acronym UKHSA. Two relevant studies that our discussion covers are epidemiological and health economic implications of symptom propagation in respiratory pathogens and mathematical modelling investigation, which was published in Plus Computational Biology in May 2024, and a scoping review titled "Sentinel propagation in respiratory pathogens of public health concern, "a review of the evidence," which was published in the Journal of the Rural Society Interface in July 2024. Hi, Phoebe. Thanks for joining us today. Hi. Thanks for having me. Welcome to Swider Presents. To begin, Phoebe, we heard you from you in a spider careers podcast episode about PhD study, where you gave your perspective about carrying out a PhD as part of the MASSIS programme. So, would you like to share with the audience how you got a general interest in research in epidemiology and mathematical modelling? Yeah, so I did my undergrad at Warwick studying maths, and I always found that I was more interested in applied maths, where I could really see the real-world applications. I was quite lucky that at Warwick, there were quite a lot of mathematical biology modules, so I basically took every single one that I could, and through that, I gained an interest in epidemiology. And then I did my MASSIS project on epidemiology, and then that kind of continued into my PhD. That sounds a great career path, and one we will share probably a lot of interesting applications in mathematics. For your PhD project, there is a focus on symptom propagation. What is symptom propagation, and what are the questions you are hoping to help address? So, symptom propagation is where your symptom severity depends on the symptom severity of the person who infected you. So, for example, if you were infected by someone with really severe influenza, you would be more likely to develop severe influenza yourself than if you were infected by someone who was, say, asymptomatic. So, what we're looking at in my PhD is two main questions. The first is, does symptom propagation occur, and for which pathogens does it occur? And then the second question is, if it does occur, what are the effects on epidemiological outcomes and the effectiveness of intervention strategies? And in the last few months, we've had these couple of publications. Across these two studies, what was the thought process? What was the particular design behind them to then help address these key questions, the focus of your PhD? The first question, which was, does symptom propagation occur, was the focus of our scoping literature review. So, this was looking at evidence for symptom propagation for 14 different respiratory pathogens. And this was also trying to understand what are the biological mechanisms underpinning symptom propagation. Our second study was more modeling focus. So, we designed a model to include symptom propagation, and we have a single parameter that encapsulates the strength of symptom propagation, which we call alpha. Our study is then looking at the effect of different values of alpha on epidemiological outcomes, and also on the effectiveness of three different types of vaccine, which have varying effects on how much they prevent infection or how much they reduce symptom severity. So, these two studies are different in format and style. Are there any reflections on the writing process, the similarities and differences to convey to the listeners? Yeah, I think writing these two papers was kind of two completely different experiences. I think maybe our modeling paper was more standard in that we sort of had a research question. We went away and did some research, came up with a set of results and a set of figures, and then after that we wrote it up into a paper. But I think for the literature review, it was much more that the writing up and the research kind of happened at the same time. And as I went through the research, I would write part of it up and then the write up would then inform the research itself in terms of finding the gaps and what extra parts need to be addressed to have a complete literature review. Yeah, I can imagine how many papers you have to go through in the field there. That's it. Trying to remember, is more than a couple of hundred? I think something like 215 papers in total across all 14 pathogens. Wow. As I say, this was a scoping review, so yeah, very much trying to get just to capture a breadth of information across, say, focus on 14 specific respiratory pathogens. And so going into a bit more detail on the scoping review article. So what were the main findings for evidence to support or not support symptom propagation occurring for the different respiratory pathogens that we considered in that study? For almost all the pathogens we included in this study, we found that there was support for symptom propagation, so this includes pathogens like influenza and SARS-CoV-2. However, this was not actually the case for all pathogens. We found that there was variation in this and we also found that there was variation in the mechanisms through which symptom propagation occurs. So for some pathogens there would be really strong evidence for one mechanism and then no evidence for another mechanism, but then for a different pathogen it might be the other way around. So our main finding was really that there is evidence for symptom propagation, but there's also a lot of heterogeneity between pathogens in what evidence is and what that evidence looks like. So it's really important to be considering all pathogens independently. And that's especially important if there were to be a novel pathogen of public health concern and we wanted to see if symptom propagation was an important factor. And from the modeling focus study, what are the potential implications for epidemiological and health economic modeling assessments? Were there any results that you found surprising? So we found that stronger symptom propagation led to an increase in the proportion of cases that were severe. And we also found that it led to intervention strategies which reduce symptom severity. So vaccines that reduce symptom severity being more effective whilst having no effect on interventions that were purely infection blocking. I think maybe the most surprising result was that we found under strong symptom propagation vaccines that only act to reduce symptom severity were more effective at reducing the total number of severe cases than vaccines that prevent infection. So it was actually better to just act to reduce symptom severity than reduce case numbers. And I think that's quite interesting. We can find these combination of parameters where you can potentially get a cancer intuitive type finding. So at this point having carried out these two very different types of study, at this current point in the projects, what from in your view I'd say the current take home key messages of the current findings and what are the limitations with the work done so far that it's important that we keep in mind. So I'd say the two main takeaways are that there exists evidence for symptom propagation and that symptom propagation is something important to consider. And then also that symptom propagation acts to affect the outbreak and also how effective our intervention strategies are. I think this is really important given how little symptom propagation is actually typically considered. It's almost never considered when we're evaluating how effective intervention strategies are. But I think at these findings demonstrate that it should be and it can in fact completely change our results. And for example, what type of vaccine would be more effective in terms of limitations. I think the main limitation from our modeling study is the uncertainty in our parameters. So this study just uses estimates from the literature where there can be a lot of variation between studies. In particular, our results are somewhat sensitive to the variation in transmission rates between mild and severe cases. A key assumption we've made is that severe cases are more transmissible than mild cases. However, if this was not the case then our results would be qualitatively quite different. We also have an included behavioral considerations. For example, individuals with more severe disease might act to reduce the number of contacts and therefore act to reduce their number of secondary infections. So I think this would be something really important to address in future studies. That's very interesting and especially, for me, for example, the knowledge of symptom propagation is very little and I didn't know much about this. So when if you spread this sort of information, it helps other researchers to make a stronger consideration of this sort of topic in their modeling. Yeah, I think that was kind of the purpose of our scoping reviews. Like symptom propagation is kind of so unheard of and so unknown that we wanted to put out a review that kind of put symptom propagation, I guess, on the map a bit and get people thinking about it. Amazing. So to conclude, what knowledge gaps on symptom propagation would you like to see further investigations focus on? Relating to the literature review, I think one of the main areas that could use more research is looking at symptom propagation for specific strains. In our literature review, we found some evidence that there would be variation in how much symptom propagation occurs between different strains, for example, for COVID-19. But there's really not enough biological evidence looking specifically at specific symptoms to be able to perform this sort of analysis. But in future, if there's more of these studies coming about, then being able to do a review for a specific pathogen looking across strains, I think that would be really valuable. And then in terms of on the modeling side, as I said, one of the limitations is the uncertainty in parameters. So being able to actually fit this model to data would be really valuable. And in particular, trying to estimate this value of alpha, which is the strength of symptom propagation from data, in order to determine to what extent symptom propagation actually occurs. Very much so. And that is the current focus of ongoing work at this time. So I think watch this space for further outputs in the future. Yeah, definitely. Yeah, I'm looking forward to that. This was very clear and complete. I don't have any more questions. I will have in the future, though, because again, this is all new and I'm very exciting. But I still, we need to learn how to use it in our own models. Thank you very much for your review. Thank you very much. And thanks for having me. Thank you again, Phoebe, for joining us. And thank you all for listening. We hope you will join us again for our next episode of Spider Presents. You