Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

Welcome everybody on behalf of the Royal College of Surgeons of England. It gives me enormous pleasure to welcome our 2024 Hunterian Professor James Glasby James is an T four registrar in general surgery in Birmingham and also an NHR clinical fellow in Global Surgery and clinical trials. If I were to list all the accomplishments of this young man, my talk would be longer than his, but I have to say that two of his accomplishments would suit a very senior research at consultant level. James has published more than 200 peer reviewed papers and is a co applicant on more than 15 million lbs of successful grants. More than a lot of people who are full time in research and retiring. What I find the most attractive thing about your CV. James is the paper on global surgery COVID collaboration for which James was a co leader. This paper is the world's largest collaboration project in global surgery. Today's day is all about partnership and global surgery. And I think very fitting that we have here. One of the co leaders of the world's largest collaboration in global surgery and the double for that paper is published in the Lancet but also in the Guinness Book of Records. So James gives me enormous pleasure to welcome you to the podium to speak to us on transforming post discharge, follow up pathways in global surgery using telemedicine. Thank you. Mhm. Thank you very much for the honor of being here today. I'm hugely grateful to the college professor Lavi. Thank you all for listening to this talk which will describe my doctorate studies around transforming post discharge follow up with partners in low middle income countries. Professor John Hunter was a member of the college here and a fellow of the Royal Society in the latter part of the 18th century. He worked between Chelsea and Westminster and ST Barts Hospital before training as an army surgeon for a number of years, he predated Pasteur's germ theory about 100 years but made some significant progress in the fields of inflammation and infectious disease. So hopefully, it's fitting that today's talk will focus mainly around surgical wound infections which are a problem that affect many patients around the world and a key research priority for our networks. I don't have any conflicts of interest. Delighted to say that this research was funded by one of the RCs research fellowships and by doctoral research fellowship between 2019 and 2022. My main conflict of interest is this, I'm honored to be presenting this work today, but really none of it would be impossible without the hard work of many 100s of collaborators around the world. So I'm hugely grateful to each and every person that made this work possible. So a spoiler alert, I'm going to run through today. The problem with telemedicine for post discharge, follow up, perhaps present one or more potential solutions and then discuss perhaps their remaining uncertainties. I would like to speak to anyone interested in running SSI research in their context after today, if you're looking at developing an infection monitoring program out with the research setting, please do get in touch because some really exciting opportunities for collaboration and new partnerships from this work. So starting from the beginning, why do we think that research is fundamental to advancing global surgery? What we've heard today how surgery is increasingly being recognized as a global health priority and a focus in the NSPS about increasing surgical capacity, about task shifting members, the workforce building capacity. But if we increase the surgical volume around the world without an eye on data driven practice and improving and enhancing surgical quality, so will the rate of absolute rate of surgical morbidity increase? And this could be a real problem for our patients. And why is it particularly important in the context of global surgery? We know from our monitoring studies, Global search 12 and three, that the risk of serious complications after surgery, death, wound infection are much more common in low middle income than high income countries. And there's a pretty linear relationship and that's often due to case mix patients present later also to do with capacity to rescue patients from serious complications of surgery. Our paper in the Lancet global research three described that about 25 per cent of the variability and the risk of death after cancer surgery was related to center level factors. If you put this all together, then actually surgery, if you considered it to be a cause of death, morbidity would be one of the leading causes of death and morbidity around the world. So what is surgical wound infection? You forgive me for the gory images. We've all seen patients like this before that develop a bit of oozy serous fluid from the wound, then maybe a little bit of blood pass and it can lead to full thickness. The essence, the whole spectrum of and wound infection is damaging for patients. The symptoms are very unpleasant, it's smelly, it's uncomfortable, but it can prolong patients getting back to their normal lives and it can cause serious morbidity and even be associated with death. And that's particularly relevant to patients in low and middle income countries who are already at risk of catastrophic expenditure just as a result of their index surgical episode. So the increased financing that would be required to treat these complications could be life changing for them. So, the NHR Unit on global Surgery, why we focused on SSI this has been driven by our partners across the network. All of our work is grounded in co prioritization. And with our first exercise in 2017 SSI floated to the top as the key research priority across all contexts and settings. Appreciating that between 3050 per cent of patients in some low resource environments may develop a wound infection. After abdominal surgery, you can see why that might be the case. So the home for this work was ther global Surgery Research Unit, which is collaboration between seven higher education institutes in seven different countries across three continents. We've grown together and shared partnerships now since 2014 and we have deep relationships with both the hubs acts as effectively leaders of the unit with South South collaborations, South North collaborations and have their own infrastructure for health economics, statistical analysis, data collection and database management. This is a a map of the the networks so far. So be talking today about work we've done within our SSI trial portfolio and this has evolved since 2016. Each of our studies has a very tenuous animal name. So you have to forgive us in advance. The two studies that we've completed so far are Falcon and Cheetah, which we're looking at different ways to prevent wound infection through skin preparation. The type of sutures that we use and whether they should be coated in antimicrobial material. And Chito is about changing gloves and instruments before you change. Before you close the fascia, an operation, we're currently recruiting to Penguin which is looking at different oxygen concentrations. We're just about to open a MAMS platform trial, looking at pre operative washing intraoperative wound irrigation, different type of dressing. So the studies have got more complex over time and we've had to push the boundaries of SSI methodology in order to be able to achieve this. The work is all grounded in the idea that it should be policy changing at the point of the research results. And we've been working with colleagues from the who's SSI prevention guidelines in order to try and bring that to patient benefit as quickly as possible. So, in the context of this SSI research, what is the problem with follow up? The main problem is with SSI, the more you look, the more you see. And if you look at surveillance studies in the UK, for example, they often report 2 to 5 per cent wound infection rates and that's just not correct. But we've been fighting with thehr for many years about our sample size calculations in grant applications because they do not believe that in the UK, the wound infection rate is 22 per cent after main, the elective and emergency abdominal surgery and it's not that different in elective and emergency settings. If you look at the control arms of randomized trials, including which we just completed the first phase of in the UK, and this has been reproduced there. And the rates in low middle income countries across a similar portion of operations between 2040 per cent, rising up to 50 per cent in the dirtiest contamination groups. So this is a really common problem affects many of our patients and we should all be thinking about. So what's the answer? So you may say, well, we just need to look really, really hard and make sure we always identify patients with wound infections when they develop them. But if we follow the US CDC guidelines are currently accepted as the gold standard. That would mean bringing every patient back to the hospital at 30 days after their operation for an in person assessment. And the CDC guidelines not to be too critical are rather imperfect to start off with one of the diagnostic criteria is have you been diagnosed with a wound infection by a clinician would automatically give you a yes diagnosis of SSI even if that was a community healthcare worker with limited training in identifying wound infection. The absolute cost of bringing patients back to be reviewed by a clinician expensive for a patient and the healthcare service just back of an envelope perhaps if it costs $20 for the healthcare service, 10 for a patient in the trial, the size of Falcon that would be 200,000 lbs, 200 $1000 worth of funding. There's a huge opportunity cost where we say that there is insufficient. We've heard a lot today about the workforce problems around the world and insufficient surgeons, obstetricians underneath this, deliver the healthcare that's required. There's an opportunity cost to getting trained CS to follow these people up. Huge implications of this for equity access and representation of patients within trials. This is a photo of Manna, which is one of our key collaborating centers at the foothill of the Himalayas, Stephen Alexander. That's the only surgeon in this hospital says that when patients have an operation with him during the winter months, they can only come back nine months later when the season is correct for them to visit back, but he always sees them back at nine months. So you can imagine how inconceivable it would be to get people to return early for a wound review, other forces at bay of course. And during COVID, it was really important to adapt and to deliver a robust follow up pathway was able to deliver si follow up without having to get people back in person. So I guess the other end of the spectrum would be why we just not worry about looking. Well, this has consequences to patients health systems and research systems for patients. If we're not proactively trying to identify complications such as wound infection that can lead to prolonged presentation. And that's particularly important where perhaps people have slightly variable health literacy for health systems, you can't accurately measure or monitor morbidity and for research systems that introduces the risk of bias and you end up with under powered trials, do not allow us to change practice. So neither of those I would say is the right answer. So I'm going to share with you in the next 20 minutes or so, what the work that we've done that? I've led up part of my phd on transforming this pathway to make it robust, reliable and valid for patients in low middle income countries. I'm focusing on wound infection here and give 10 questions with some data driven answers. So I guess the first question would be, can we not just follow patients up in hospital and after they're discharged, it's OK for a research purpose. If we can reliably pick up wound infections before the patient leaves, then we don't need to worry about what happens afterwards. And this is actually what was used in many of the checklist wound infection studies and many published trials were perceived to be high quality at the time. You may have a lower event rate. So in Falcon, which I've done this analysis on which is published in Lancet Global Health. The in hospital rate of wound infection was 12 per cent was 20 per cent, 30 days. So if we were considering in hospital wound infection as a surrogate, perhaps for 30 day SSI, then if it was to be a good surrogate, it would need to meet four criteria. The first is that the treatment must have an effect on the surrogate. The in hospital SSI. The second is that the treatment, the thing that was tested in must have a similar effect on the true outcome. The 30 day SSI that the surrogate must affect the true outcomes. They must influence one another and that the full effect of the treatment outcome must be described by the surrogate. So we analyzed this using meta analyst method that was developed by ORC in Belgium and effectively what you do is do a fixed effects meta analysis and look at the sleep of the linear aggression. And there's two things that we're looking at. If you forgive the slightly small diagrams, the first is can the effect of the treatment, you're testing the trial be fully described by the surrogates when you compare it to the true endpoint. So that's the first two graphs that you can see. You want it to be over naught 0.8 for the trial to be considered for the for the. Um so to be considered trial valid, and you can see all of the estimates fall well below naught 0.8 on this diagram on the right is individual level surrogacy. So to what extent does the in hospital measure correlate to the 30 day measure? And again, this should be above 0.8 to be considered individual level valid, which none of the effect estimates across any of the countries that were participating reached. The key take away really was that you've got to follow people up after they leave hospital for wound infection in high quality research. And this made it into one of our key criteria for assessing SSI trials in this lancet infectious disease metra analysis. So I guess question number two is, well, if you have to follow people up after they leave hospital, can we just give them a call? We trained clinicians, we know what the CDC criteria are. We've seen lots of wound infections. So we looked at this in global Search T and we had 15,000 and discharge patients. We looked at the rates of wound infection detected when people had a unstructured telephone call. So whatever, we didn't say that this is what you had to do, it was whatever the local clinicians were already delivering versus in person rates. And what we saw was a drop out of about 25 per cent in the SSI rates detected when people gave unstructured telephone call versus their in person review. So 25 per cent of wound infections missed. I then did a systematic review and Metro analyzed all of the available data in the literature alongside the Global Search two data and that confirmed that one in three wound infections are missed with the unstructured telephone call perhaps would be going on at present. So my takeaway from this really was that that doesn't really work either even with really well trained clinicians. So one potential solution is the bluebell wound healing questionnaire. This was initially developed by Professor Jane Blais Bee's theme in Bristol. And we were a key collaborator in the validation of that study across the UK. This is a questionnaire that's got 19 items. It's patient reported, it can be delivered by clinicians and non clinicians and the patient can fill it in themselves, it can be delivered in person and over the telephone and in the UK, it was really accurate in detecting wound infection. So this potentially provides a very useful answer. I part of a wound research group in the UK. And we think this is a really good tool for diagnosing wound infection in the UK. But it was developed by country researchers in English first language contexts. And we had no idea whether it would be relevant, cross or cross culturally valid around the world. So there's a few problems with this questionnaire. Firstly, we also need to deliver it in lots of different languages, some of which may only be spoken and not written languages. There may be cultural and contextual differences into the sort of questions and symptoms that people might experience when they have a wound infection. And we have no idea whether ringing people after a trial in the seven low malignant countries would be possible, let alone whether we can accurately detect wound infection. So, what we did is I took the Falcon trial and I built into the protocol, a study within a trial. This has become very popular in trial methodology at the moment. Often it focuses on tweaking the actual mechanism of the trial to make it more effective or efficient. For example, how you get more people back to follow up here. I pre built a diagnostic test accuracy study into the Falcon trial protocol which allowed me to adapt and validate the study. So if anyone's adapted a questionnaire and validated before I wouldn't necessarily recommend it if you want an easy life, it was very, very, very hard work. The patient reported outcome experts are very, very meticulous in this aspect and it has to be right if you're going to use it as a tool in high quality trials. But for each country, each of the seven countries and in each language, we had to do a cross cultural and cross language adaptation. And the way to get around it across the 22 languages that we had to do within the seven country network was to do a period of cross cultural adaptation and then to translate it afterwards that needs forwards translation. So from your home language to the language you want to translate to backwards again, led by a clinical champion with two in each direction and then with several checkpoint interviews throughout including a check in process. At the end, this was hard work. And then what we did is take the responses to 1000 questionnaires in a pilot phase and do something called rash uni dimensional measurement modeling. This is a psychometric model allows you to investigate the internal properties of a questionnaire. So if you're measuring wound infection as a common trait, you can work out which items best describe wound infection and they will rank them. And you can look at whether there differences in the way that people respond to questions based on their own characteristics. So do men describe wound infection differently to women? People from rural contexts describe it differently to people from urban contexts. Triangulated all those findings together to come up with some recommendations for using this questionnaire across these seven countries. This is an overview of that process which I won't go into in detail, but I promise you was that rather complex. And the first thing we picked up was that the response levels just didn't make sense. And they were for a question say like was there any redness in your wound? People could say not at all a little bit, quite a lot and a lot. And the middle two categories do not translate well across different contexts a little bit and quite a lot rather similar to most people around the world. And this was very nicely described both in the qualitative interviews and in this rash mentioned measurement modeling. So allowed us to use both a qualitative and quantitative approach to make this recommendation, which is quite a big thing in patient report and outcome terms. There was also several examples where wording was considered to be too complex, not relevant. So, redness is a concept actually doesn't work very well in many types of skin tones, people fed back that actually shining of the skin was much more relevant for some patients. There was an item asked about the exact temperature that you'd recorded, which people may not have access to a thermometer. And there were some questions around antibiotic use which people said some patients may not recognize. So after that point, we had an adapted questionnaire which we could use potentially to detect wound infection over the telephone. So the next point is it actually going to be possible. So what we did is before patients came back for their 30 day follow up in the Falcon trial, which was mandated in that trial because we were working by our historical troublesome CDC rules in the two or three days before. At some point, a local researcher gave the patient a telephone call and completed their questionnaire. So as near as possible to 30 days and they all had to have both. So we were able to compare it directly to the gold standard and that person completing the question over the telephone, didn't have to be a trained clinician about half the time it was a data manager or non clinician, which she thought was really important in terms of scaling this up across a complex research network and shifting capacity. So over 1200 patients, this was delivered in 25 languages and 50 per cent of the time, we weren't able to deliver it in a formal written language. So it was just a verbally spoken language and ad hoc translated by the person doing the questionnaire. This was very troublesome for the patient reported outcomes, measure experts that we think represents the diversity of the delivery network and would be true to real life. In a protic randomized controlled trial, we managed to reach 90 per cent of patients for follow up. So this is possible. We had a comment earlier about whether this would be possible or not. There was lots of innovative solutions on how to do so so often be just one telephone per household, for example. So people would either wait to that person, have the telephone within the house or go to the community telephone. Some people traveled out the village in order to be able to get the telephone call at a specific time. But with these inventive solutions derived by the community, they were able to complete the follow up. People really liked having this extra contact with the research team. And 99.5 per cent of people said that they felt it added to their clinical experience. So my takeaway from this was I think most patients anywhere in the world, even in the most rural and austro environments can be reached by telephone for follow up. And if it's accurate, it should be used, the standard for our future research work. So next question was, is it accurate in detecting si? So these are two graphs where as the X axis goes along, you get more points in our adapted wound healing questionnaire score. And then the why a access was on this side. On the left was the proportion of patients with a wound infection. So effectively, the higher the score was, the more people had a wound infection, you might anticipate when we looked at the discrimination of the questionnaire at a statistically calibrated cut point and we could discuss the length whether that's the correct thing or not. And then the the receiver operating was about 0.87. So this is pretty brilliant discriminating patients with or without SSI maybe not 100 per cent of the time. But with all of the efficiencies gained, I'd say that this was sufficient for high quality research study and it would never be give misclassification by arm in a trial. It would always be the same for each arm. So it would never introduce research bias even if it missed the occasional patient with wound infections. And this was robust across urban or rural settings, across men and women across the people that just ad hoc translated the questionnaire versus the form of translation, which I think again, speaks to important pragmatic answer. So I take away from this was that when you wanted to text SSI over the telephone, you can do so with this wound healing questionnaire, it can be done with or without a trained clinician being present and it can be performed across a variety of languages. So this is a really useful tool for the future. Our next step was what would happen if you added in a live video. And there's lots of different advances in this that have been described, the selfie photo, you ask a patient to take a picture of their abdomen. There's the video recorded and then sent in we like the live video for a number of reasons. Firstly, almost everyone was using whatsapp in any case. So something that was accessible, community engagement, involvement work so that people did have access to whatsapp. The second thing is it requires no data storage. And it's just the only thing that you write down is the evaluation of the person on the telephone at that time. So no problems with transferring data or any issues that the last point was that you can do the assessment as if the person was in clinic with you. So if you would like the patient to move or to push on something or to give you a better look, then you can do all of these things over the telephone. But this is a total evidence gap in the literature. We looked at different methods of telemedicine in the systematic review I mentioned earlier and 37 studies described and any one of them had ever looked at video follow up to diagnosed re infection. This is totally fresh. So what we did in the cheer trial, which just a telephone follow up as an evolution of this was to get people to do the questionnaire and then to have a video wound assessment immediately afterwards and then reassess the wound based on your video findings that allowed us to compare what you got over the telephone with what you found when you did the video assessment and we asked people to emulate as much as possible, what you would do in a clinical setting. When the patient was in front of you, our colleagues in India and Nigeria felt that there might be sufficient availability of mobile phones within their communities to do this study. It was only within selected context and I'll show you maybe some difference between urban and rural settings in a second. Again, most people were able to complete video follow up in these two countries across seven different centers. So meals, the interesting one because again, that's a very, very rural hospital. And even there two thirds of patients were able to do this live video wound assessment as high as 100 per cent, 95 per cent in some of the main cities in Nigeria where there was problems it was to do with image quality and connection. As you might imagine, we found that the inte reliability was pretty good between the video assessment and the telephone questionnaire, but it did actually change the assessment in 5.6 per cent of patients. And really, it was able to detect more of the mild symptoms when it was scored by the rash measurement concept identified earlier. So probably you pick up a few extra wound infections, which you wouldn't do without the questionnaire. So probably where live video is feasible, it may be a useful adjunct to detect a mild exercise. So what should we do now? Going forwards, I think a robust pre discharge follow up using those CDC criteria is really, really important. And then probably it's OK just to call people and complete this wound healing questionnaire at 30 days. If it's possible feasible to supplement that with a live video, that would probably be a benefit where the capacity was available. I think one important takeaway from our community engagement involvement work is that should always be clear pathways to triage people back to care after they leave hospital. And often those were not particularly clear, which was fed back when people said that they really liked that additional telephone call. And this is an example that we used in the UK of a pathway rather like that where patients, this is from. My Kenneth mclean is a colleague in Edinburgh part of the unit where patients basically were triaged regularly after they left hospital via text message. If they had any problems, they were able to in the questionnaire and come back and see a doctor. So this sort of system is integrated and low burden is probably the answer for the future. So we're now able to implement this wound healing questionnaire in trials and clinical practice. We're using it now in Dragon Emmalin, which are two funded SSI trials and be used for 25,000 patients within the next few years within this trial. How do you make sure that we can get back to patients to pick up telephone call and make that even better? What we've come up with, with our CEI representatives across the seven countries. We came up with a co produced behaviorally informed toolkit with some simple things you can use in everyday practice. And they stuck this up around the hospitals for the duration of the study which basically made that people had the correct phone number before they went home. The mobile phone was definitely in use and you checked it with them before they left, for example. So this is part of the toolkit that we can apply when using this in trials in the future. It's all really simple things you may do every day, but certainly many things that I hadn't thought about. And maybe just one last reflection if you indulge me, I think that telemedicine can help underpin health system preparedness in the future. And this was some work that I helped lead during my phd around a surgical preparedness indicator. We found that surgery was very fragile during the pandemic, we hear a lot in the news about the elective waiting list and the backlogs and certainly when there was extreme stress on the health system, elective surgery was the first thing to drop. So we think that by improving the preparedness of emergency surgery, which includes emergency elective surgery, which improves critical care, contribute to whole systems to the lesions to improving preparedness as such as part of this co prioritization exercise. Having remote telemedicine outpatient pathways was one of the key indicators and it was actually the second worst performing indicator, particularly low performing in less well resourced environments. So hopefully, by improving the telemedicine capacity and capability of hospitals around the world that may help to keep providing surgery and elective healthcare in the future. If there's further pandemics, if there's natural disasters, et cetera. So where the remaining uncertainties, I think there's some more work to do on video follow up. And we've just got funded to run a trial in the UK, which will include about 25,000 UK patients across six different clinical pillars, each of which will have some kind of digital photo or video follow up. And we will have data from the Marlin and Drug and trials. I think together that's going to be a very rich data set for a and deep learning, which may be a useful adjunct, which is going to be simple and cheap to apply across very wide set of contexts, we have only examined this in abdominal wounds so far. And some colleagues in plastic surgery in York are looking at this for other wound types. There's a big opportunity to validate this sort of questionnaire across other contexts because the symptomology is likely to be really different. And we have yet to prove that by introducing high quality post operative surveillance in the community that, that translates to clinical benefit beyond the intuitive things that I spoken about. In summary, hopefully, I've shown that we've developed a validated optimized digital remote follow up pathway to detect wound infections in labor income countries. And last call for collaboration. If you're running a trial in this area, if you're setting up a monitoring program, I'd love to talk to you over coffee after this. If you're looking for a phd program or IBSC or MSC, this is wonderful opportunities in the unit. I was incredibly well supported by my colleagues. I'm pleased to get in touch. Big thanks to my phd supervisors, Neil and Dion in particular, were incredible throughout this whole process. The phd fellows that I shared time with seven H leaders who were tremendously supportive and with very, very closely that time to my, my long suffering partner and my little son and to the wonderful team in the ITM that come to support me today and to our collaborators around the world. So many thanks again to the college and I like to take any questions Thank you, Professor Glasby. Professor Glasby, you've educated us and I want on behalf of the college to congratulate you, congratulate you on the elegance and efficacy of the study. You've not only done a great piece of research, but you've set in motion a tool that is going to be used so widely in global surgery. And many congratulations. And it gives me great pleasure to present you with the Hunterian medal. Thank you, James. Traditionally, we don't have questions after the Hunterian, but people are very, very welcome to catch you afterwards. So a word on protocol now we will now process out and Dorothy and I will return and as quickly as we can without our gowns and end end of the day. Thank you. Good. What a fantastic day we've all had. I'm sure I speak for you all when I say we've had a wonderful day and it's now left for me to say thank you all today for making time out. And thank you to everyone who has joined us virtually. I hope you feel as inspired as those of us who are in the room feel. It's been amazing. It's covered many areas and for those who are wondering, how do I get involved? I hope that you got the answers. Please do contact the Global Surgery Unit at the college. If you have any remaining questions and we'll take those forward for you. I'd like to invite you all now to the ground floor meeting room where we'll have a few drinks and some networking. Thank you all very much and I look to look forward to seeing you at our next Global Surgery frontiers conference. Be on the look out because he will be bigger and better. Thank you all very much.