UK Stroke Forum 2025 | Interim results from the PHOTONIC study: prehospital video triage for suspected stroke

Jessica Weng

Hello everyone, I’m Jessica. I’m a research fellow in health economics at University College London and I’m working on the quantitative analysis of PHOTONIC.

Angus Ramsay

Hi everyone, I’m Angus Ramsay. I’m a principal research fellow at University College London and I’m the chief investigator of the PHOTONIC study. The key issue with pre-hospital video triage and the need that it addresses is that a lot of the techniques used to identify stroke are quite over-inclusive. So things like FAST and so on identify far more potential stroke patients than there are in reality. That means of the patients who go to stroke units, almost 50% of them aren’t really having a stroke and they might benefit from being transferred directly to an emergency department or not being transferred at all, staying home or being referred to their GP for a follow-up case. So PHOTONIC was set up based on some earlier evaluation work that I led looking at two of the pilot PVT sites that were set up in the wake of the beginning of the pandemic. And based on those findings, there was an indication that pre-hospital triage services were acceptable and seen as safe by staff. We decided to look at this in more depth and got funding from the NIHR to study four pre-hospital video triage services across the southeast of England. In the study we are using qualitative methods to look at how the services were put into action, how they’re being delivered and used, how different stakeholders including staff, patients and carers experience them and the impact on care delivery, patient outcomes and cost-effectiveness. I’ll hand over to Jessica to talk a bit more about the quantitative analyses.

Jessica Weng

For the quantitative work stream of PHOTONIC we use large observational data sets from multiple sources. So to study patient outcomes, we use inpatient admissions data from hospital episode statistics linked with mortality data from Office for National Statistics. And to study the impact on stroke care delivery, we used data from a stroke audit database. Unfortunately, we could not link the two datasets together, so we analyzed them separately. The main take-home message from these two datasets is that we did not see any statistically significant effect at the overall group level. We did see significant effect within the individual cohort. This is largely due to the variation across the four case sites, and the contextual factors were very site-specific, and these really influenced the outcomes. And there were some limitations to analyses due to the nature of the data we were working with. One limitation is the classification of the exposure variable. We were not able to directly identify patients who actually received PVT in our data sets as this was not recorded. So we were looking at hospital or regions that were using PVT, and that’s how we define our exposure group. So within this exposure group, we actually had patients who were treated and who were not treated, and by mixing them in exposure group, this dilutes the actual treatment effect, so making the overall group effect seem closer to zero than it actually is. And another limitation that we were working with is the time trend. So we model the single COVID variable, which is a period from March 2020 to April 2022, that there were multiple waves of the pandemic and each wave was affecting the healthcare systems differently. And another limitation is the outcomes we were looking at they were heavily influenced by factors other than PVT or the patient’s health. For example if a patient was not discharged from the hospital this could be due to a lack of capacity in the rehabilitation team, so they could not take over the patient. And this could increase the length of stay in the hospital for the patient. Also, there was service reconfiguration that was happening around the same time when PVT was implemented in two of our case sites. So this could also influence our outcomes. Although we were not able to demonstrate a significant overall effect with the quantitative data, we do aim to capture a lot of these contextual factors qualitatively.

Angus Ramsay

And just thinking about the wider lessons from the work, we saw a number of benefits from the qualitative research. So, for example, in terms of patient and carer experience, patients who were assessed using the video technology found it enormously reassuring to speak directly with a specialist stroke physician to get an understanding of what was going on and what the decision was, whether it was, we think you have a stroke, you should come to us, or we don’t think you’re having a stroke, this is why you should be going somewhere else. People find that valuable. Patients also found it valuable to be communicated with clearly by ambulance clinicians who explained what was going on and very often facilitated the assessment processes that were run through pre-hospital video triage. In terms of impacts on staff and services, we generally found a high level of acceptance and enthusiasm for pre-hospital video triage amongst both ambulance and stroke clinicians. There was a view that it did what it was trying to do. So it was effective at influencing conveyance decisions, whether a patient should go to the stroke units or elsewhere, and that it helped stroke teams be more ready for the stroke patients who were coming in. So they had a sense of what was going on and they knew what to do when the patient arrived. So it was valuable in those ways. Additional sources of value were it helped smooth the referral pathways for TIA clinics. It also had an important effect on how different staff groups interacted. So there was a great increase in the degree of respect, shared sense of mission and understanding of the different types of expertise that stroke and ambulance clinicians had. So a greater sense of mutual understanding and respect. So there were a number of real benefits, perceived benefits at least, to the work. I should emphasize that all of this is emerging findings rather than the absolutely finalized papers. We’re still in the process of putting those together. We’ve just come out of the session where we were presenting our findings to a wide range of clinical leaders from stroke services across the country and I think internationally. And the view in there was very much that pre-hospital triage is coming. It makes sense, and people who do it, the first time they do it, they understand the value of it. Examples that were identified were some of the ones that I outlined. Another really important point is just general service sustainability. The pressures that stroke and other services are under. Anything that can streamline things so that people are getting where they need to go first time is seen as enormously valuable. Anything that reduces the pressure on stroke services and other services appropriately is going to be of value so that people can dedicate their time to dealing with the people who really should be receiving care at that point.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.