ESOC 2022 | ParaNASPP: prehospital NIHSS and paramedic training for improved stroke triage
Stroke is well recognized as a highly time sensitive condition. Early identification and acute treatment of stroke is crucial to optimize outcomes for patients and yet, a lack of clear workflow and differing systems between the pre- and in-hospital settings can often lead to time delays. Mona Guterud, MSc, Paramedic, PhD Candidate, and Helge Fagerheim Bugge, MD, PhD Candidate, Oslo University Hospital, Oslo, Norway & Norwegian Air Ambulance Foundation, introduce the Paramedic Norwegian Acute Stroke Prehospital Project (ParaNASPP; NCT04137874): a project aiming to improve triage of patients with acute stroke through structured paramedic training and improved communication with physicians at stroke centers. The study used a stepped wedge cluster design, enrolling patients with suspected stroke within 24 hours of symptom onset. Paramedics undertook an e-learning program on stroke and the NIH stroke scale (NIHSS) and utilized a mobile application for NIHSS scoring that transferred data directly to the on-call stroke physician at the hospital. The primary outcome of positive predictive value (PPV) of a final stroke related diagnosis was comparable between the control and intervention groups. While a significantly longer on-scene time was recorded, the intervention was shown to save time once the patient arrived at hospital. The intervention group also identified more subtle strokes. Implementation of prehospital NIHSS use and direct communication with in-hopsital services may enable smoother workflow and more correct triage. This interview took place at the ESOC 2022 congress in Lyon, France.
Transcript (edited for clarity)
Mona Guterud:
We have been focusing on the prehospital triage, as we see as one of the main issues as the identification of the stroke patients are primarily done in the prehospital context. So as it is today, there are so many prehospital scales that complicates the triage system for the hospital that uses only NIH stroke scale. That’s what our project have been dealing with, mainly...
Mona Guterud:
We have been focusing on the prehospital triage, as we see as one of the main issues as the identification of the stroke patients are primarily done in the prehospital context. So as it is today, there are so many prehospital scales that complicates the triage system for the hospital that uses only NIH stroke scale. That’s what our project have been dealing with, mainly.
Helge Fagerheim Bugge:
We’ve also seen that like prehospital delay is one of the biggest delays for treatment and the fact that patients don’t arrive at hospital in time for treatment. So it’s really important that we have good paramedics that can actually identify symptoms of stroke and bring the patients into a stroke ward and give them the right level of care and treatment.
Mona Guterud:
The design of the trial was a stepped wedge randomized trial. We had five ambulance stations in Oslo that recruited patients over a total of 77 weeks, we had a COVID-19 pandemic break and the intervention, you can tell us about that, Helge?
Helge Fagerheim Bugge:
Yeah. Our intervention is three parts, so it’s all about pre-hospital NIH stroke scale. But our sort of first step of the intervention is teaching paramedics about stroke and the NIH stroke scale. And the second part was that we made the NIH stroke scale into a mobile application that guided paramedics through the examination using pictograms and ensured the registration of symptom onset, anti-thrombotic medication, and vital parameters. And the third part was the communication, so trying to get a common language between the pre- and in-hospital stroke services. So the results of the pre-hospital examination was sent directly to the on call stroke physician at Oslo University Hospital. So they could sort of, live as they were talking to the ambulance, see the results of the NIH stroke scale and all of the parameters done by the paramedics, so that was our intervention.
Mona Guterud:
Well, the main results were that the primary endpoint, we had a neutral result. We looked at the positive predictive value for a final diagnosis of stroke. And we had in the intervention group a PPV of 48.1 and in the control group 45.8, so a neutral endpoint. We also had a lot of secondary outcomes. And the main outcomes we presented yesterday was time variables. And for the pre-hospital time variables, we didn’t gain any time. In fact, in the intervention group, there was a significantly longer on scene time, but something happened in hospital.
Helge Fagerheim Bugge:
Yeah, so in hospital, we saved time. So the time from sort of arrival at hospital to first CT had a median of 18 minutes in the intervention group and 20 minutes in the control group. And for patients with symptom onset within four hours, so that arrived to hospital within four hours of symptom onset, the medium time was 17 versus 19 minutes, also in favor of the intervention. And door to first CT for patients treated with thrombolysis was a minute shorter also significantly in the intervention group. The door to needle time for patients that got thrombolysis was 25.5 in the intervention and 27 in the control, but that was not significant due to the number of people that actually got treated. So we didn’t have the statistical sort of power to say that the difference was significant.
Mona Guterud:
And we also saw that we are most likely to identify more subtle cases of stroke cases in the intervention group, because there was a median of two in both the control and the intervention at admission, but it was more spread out. So there was a significantly lower NIH stroke scale score in the intervention group. Most likely indication that we are finding more of the subtle cases. And we also had a better functional outcome for the intervention group, a significantly better outcome. So that was very positive results in our population.
Helge Fagerheim Bugge:
I think it’s sort of two part answer to that. And the functional outcome part is of course, related to the smaller strokes we identified and FAST is made to detect large vessel occlusions and sort of the ‘easy’ stroke cases where you have very clear symptoms, while the NIH stroke scale picks up a lot more subtle cases. So this is sort of in line with what we thought would happen.
Mona Guterud:
We anticipated this. Yeah.
Helge Fagerheim Bugge:
Yeah. So that was sort of good news. And we also think we have a bit of a selection bias there with people in the control group. They’re asked to include patients if they suspect stroke, but like everyone else you want to be correct when you do something, don’t you? And the fact that they included a lot more of the large vessel occlusions and the big strokes sort of, that’s when you can be sure and it’s quite easy to include them, but when you’re a bit uncertain, you have not very clear subtle symptoms, then they have chosen not to include. And even though we asked them to and we really much would’ve wanted them to do it. So that gives us a bit of a selection bias, which also means that yeah, we have a lot more of the big strokes in the control group and smaller ones in the intervention group.
Mona Guterud:
Yeah. And we also think that the COVID-19 pandemic sort of tripped up the time variables for us because due to the study design with a stepped wedged randomization, almost every one of the intervention group was during the pandemic. So we know that in the prehospital phase, it took longer on scene. We haven’t seen it exactly in this study, but we have to look closer into that. But you have to put on your personal protective equipment before entering the patient. And when you’re arriving at the hospital, you’re stopped at the door for screening. So we know that there are several steps that are most likely to have influenced the intervention group to a much larger degree than in the control group. So we have to look closely into that.
Helge Fagerheim Bugge:
Still a lot of work to do, looking at the data set and figure out really what has happened and what’s behind all the data we have. Well, we we won’t include more patients in this trial. This trial is finished. So we’re just sort of, kind of do up our results. But the plan is to start training more paramedics in the area around Oslo University Hospital.
Mona Guterud:
We believe that this has been a success and the stories from both paramedics and the stroke physicians are quite clear. We want to use this system, this model. It enables a more fluid communication and there are concrete plans to implement this in the Oslo University Hospital this Autumn.
Read more...
