Embargo: 00.01 hrs CEST on Wednesday 20 September 2023

Barcelona, Spain: Two out of the four internationally-recommended screening tools used by emergency medical services are inadequate for recognising sepsis, according to new research presented at the European Emergency Medicine Congress today (Wednesday).

Mrs Silke Piedmont, a health scientist at the Department of Emergency Medicine Campus Benjamin Franklin Charité – Universitätsmedizin Berlin (Germany), and her colleagues from the University of Magdeburg and Jena (Germany), analysed data on 221,429 patients who were seen by emergency medical services (EMS) in Germany in 2016 outside of the hospital setting. They found that only one out of four screening tools had a reasonably accurate prediction rate for sepsis – NEWS-2 (National Early Warning Score). It was able to correctly predict 72.2% of all sepsis cases and correctly identified 81.4% of negative, non-septic, cases.

A second screening tool, qSOFA (quick Sequential Organ Failure Assessment), correctly predicted 96.6% of patients who did not have sepsis.

Mrs Piedmont said: “We found that paramedics never documented a suspicion of sepsis, and emergency services physicians rarely did so, only documenting a suspicion in 0.1% of cases. The screening tools recommended in the Surviving Sepsis Campaign guidelines differed greatly in terms of which and how many patients were identified as possibly having sepsis.”  

Sepsis, often referred to as blood poisoning, is a life-threatening condition that arises when the body’s immune system goes into overdrive in response to an infection and injures the body’s tissues and organs. It is vital to recognise it early, otherwise it can lead to shock, multiple organ failure and even death.

Speaking before the Congress, her colleague Dr Wolfgang Bauer, senior physician at the Charité, said: “In emergency care, there are good and long-established standards for the detection and treatment of heart attacks and stroke that have improved patients’ chances of survival. Unfortunately, a lot less attention is paid to sepsis and standards to improve early sepsis recognition and survival. Our study found there was a similar incidence for sepsis, 1.6%, as for heart attacks, 2.6%, and stroke, 2.7%, in cases seen by emergency medical services. However, in terms of both percentages and absolute numbers, more patients died from sepsis than from heart attacks or stroke. Out of all cases with sepsis, 31.4% died within 30 days after being seen by emergency services, versus 13.4% and 11.8% respectively for heart attacks and stroke. These findings emphasise the need for better sepsis awareness and more frequent use of effective screening tools.”

Mrs Piedmont and colleagues say the Surviving Sepsis Campaign guidelines about which sepsis screening tool is the most reliable for use by EMS are not specific enough and are based on little, useful evidence. The aim of the current study was to assess which of four screening tools was best for EMS to predict sepsis – NEWS-2, MEWS (Modified Early Warning Score), SIRS (Systemic Inflammatory Response Syndrome) or qSOFA.

The study linked data on 221,429 cases with follow-up between 2016 and 2017 from ten health insurance companies with information from documentation by paramedics and emergency doctors on 110,419 cases in 2016. This enabled the researchers to calculate the four screening tools’ ability to predict that a patient had sepsis. The predictions were confirmed or rejected during subsequent hospital investigations after contact with EMS. The researchers also looked at incidence and death rates for sepsis compared to heart attack and stroke, how much was recorded about any suspicions of sepsis, and how often EMS staff would have documented sepsis if they had used screening tools.

When they compared the performance of the four screening tools, the researchers found that NEWS-2 was the best for identifying patients with sepsis as it had a sensitivity (correctly predicted sepsis) of 72.2%, followed by MEWS, which had a sensitivity of 46.8% and a specificity (correctly predicted no sepsis) of 88.4%, SIRS (30.4% sensitivity, 93.8% specificity), and qSOFA (24% sensitivity, 96.6% specificity). Out of all EMS cases, 24.3% of cases were predicted to have sepsis by at least one of the screening tools, but only 0.9% were predicted to have sepsis by all four tools simultaneously.

“The incidence and death rates for sepsis and the low recognition of it, emphasise the need for better awareness and more frequent use of screening tools,” said Mrs Piedmont. “No screening tool provides ideal performance. NEWS-2 best supports emergency medical services in identifying most patients with sepsis. EMS patients that are NEWS-2 positive should be flagged up as potentially having sepsis and referred for special attention and assessment by emergency doctors who are expert in sepsis. If EMS insist on using the qSOFA, they should be aware that a positive qSOFA makes sepsis likely, but also, that a negative qSOFA cannot rule out sepsis conclusively.

“A rule of thumb for EMS staff could be that NEWS-2 negative patients are the most likely not to have sepsis, and qSOFA positive patients are the most likely to have sepsis – and also that qSOFA misses many patients with sepsis. Further clinical patient assessment and evaluation will always be needed for both tools. Future sepsis guidelines should be more precise and omit recommendations for MEWS and SIRS for emergency medical services since they were inferior in all the measures for accuracy.”

The researchers say these findings could also apply to other countries, especially as studies in Canada and the UK support aspects of their results [2,3]. “Ours is the first study comparing all four screening tools and showing the predictive usefulness of applying the screening tools to all adult patients independently of any presumptions or preliminary diagnoses by emergency medical services,” said Mrs Piedmont.

There were two things that could be improved, say the researchers: better and complete assessments of vital signs, such as body temperature and breathing rate; and “translating” alarming vital signs into a suspicion of sepsis, which could be helped by using a good sepsis screening tool.

“Sepsis causes approximately 20% of all global deaths,” said Mrs Piedmont [4]. “There is great potential to save lives and maintain patients’ quality of life if sepsis is recognised and treated earlier. As most sepsis cases start outside of hospital, emergency medical services play a vital role. They can shorten the length of time until sepsis treatments can be provided quickly in hospitals and reduce the risk of dying if they suspect sepsis.

“In addition, it’s crucial that the public’s awareness of sepsis is increased, by integrating it into education systems and through media campaigns. Patients ought to know the variety of sepsis symptoms, how urgent it is to call for prompt help and to ask the question ‘Could it be sepsis?’. As sepsis prevention is even better than early detection, they should also be aware of preventive measures such as hygiene and vaccination.”

The researchers hope their findings will inform new guidelines on sepsis that are being drawn up in some countries so as to give more specific recommendations for sepsis screening. More research is also needed to improve screening tools and to assess the real-world effects of screening tools on patients with and without sepsis.

Professor Youri Yordanov from the St Antoine Hospital emergency department (APHP Paris), France, is Chair of the EUSEM 2023 abstract committee and was not involved in the research. He said: “Sadly, we too often see patients being brought into hospital emergency departments with advanced sepsis. Some of them we can save, but unfortunately some will die who could have been saved if they had received treatment at an earlier stage. Emergency services need tools that can help them quickly and accurately predict that a patient may have sepsis and should be investigated further in hospital. This study shows that more needs to be done to develop such tools and improve the existing ones. Only then will we be able to bring down the death rates from sepsis.”

(ends)

[1] Abstract no: OA97, “Sepsis screening by Emergency Medical Services - Why recommendations of the Surviving Sepsis Campaign are too unspecific. Results from a cohort study,” by Silke Piedmont, in the oral abstract session “Pre-hospital and EMS”, 10.35-12.00 hrs CEST, Wednesday 20 September in the VIP room.  https://shorturl.at/guwJR

[2] Lane DJ, Wunsch H, Saskin R, Cheskes S, Lin S, Morrison LJ, Scales DC. Screening strategies to identify sepsis in the prehospital setting: a validation study. CMAJ. 2020a;192:E230-E239. doi:10.1503/cmaj.190966.

[3] Scott LJ, Redmond NM, Garrett J, Whiting P, Northstone K, Pullyblank A. Distributions of the National Early Warning Score (NEWS) across a healthcare system following a large-scale roll-out. Emerg Med J. 2019;36:287–92. doi:10.1136/emermed-2018-208140.

[4] Rudd KE et al. Global, regional, and national sepsis incidence and mortality, 1990-2017: analysis for the Global Burden of Disease. Lancet 2020; 395: 200-11. doi: https://doi.org/10.1016/ S0140-6736(19)32989-7.

Note: When obtaining outside comment, journalists are requested to ensure that their contacts are aware of the embargo on this release.

Funding: The data were acquired as part of the project “Integrated emergency care: A focus on emergency medical services” (http://rettungsdienst-im-fokus.ovgu.de) which was funded by the German Innovation Fund of the Joint Federal Committee (G-BA) (funding identifier: 01VSF17032). The funder had no influence on the published results. All authors have no conflict of interest to declare.

About European Society of Emergency Medicine (EUSEM)

Embargo: 00.01 hrs CEST on Tuesday 19 September 2023

Barcelona, Spain: People who come to emergency departments with alcohol-related diseases or conditions are more likely to make return visits and to die in the following 20 years, than people who come to emergency departments for other reasons, according to new research. For many, this means they may die in their 40s or 50s.

Professor Drew Richardson told the European Emergency Medicine Congress that he and his colleagues had followed 194 patients who had alcohol-related diagnoses when they arrived in the emergency department of The Canberra Hospital in 2002. They compared them with a control group of 194 patients who had diagnoses unrelated to alcohol, and they followed both groups until 2022.

“The group of patients with alcohol-related diagnoses made 44% more visits over the next decade and had a 138% higher death rate over the following 20 years, than the control group,” said Prof. Richardson, who is Professor of Emergency Medicine at Australian National University, Canberra, Australia. “The true death rate may be higher because we lost some patients during the follow-up period. After nearly 14 years, the percentage of patients lost to follow-up was similar for both groups: 40.2% for the alcohol-related cases and 39.2% for the controls.

“There was a wide variety of reasons these patients came to our emergency department, including alcohol withdrawal symptoms, trauma-related injuries, and acute alcohol intoxication.

“If these patients could be targeted by trained professionals while they are in the emergency department to educate them about the consequences of alcohol use, and to offer them assistance in moderating their alcohol consumption, it might be possible to reduce this significant health burden. Presentations related to alcohol consumption are a major burden in emergency departments.”

The study was initiated by one of Prof. Richardson’s medical students, Ms Regan Lim, who had personal family experience of the effects of alcohol and wanted to investigate further.

“Chronic and excessive consumption of alcohol has been a long-standing problem in our society. Emergency departments are the first point of contact for the many consequences of alcohol-related harm. Alcohol-related cases make up 9.5%-15.2% of presentations to emergency departments, and 8.3%-17.9% of emergency department occupancy in Australasia,” said Prof. Richardson.

The researchers analysed the number of patients who came to the emergency department between 1998-2002 (the period before the study started), 2003-2012 (the decade after) and 2013-2022 (the second decade after). The patients with alcohol-related diagnoses had made 522 presentations to the emergency department before the start of the study, compared to 389 for the patients in the control group. In 2002, the numbers were 437 compared to 399, respectively. In the first decade after, they were 1226 compared to 846 respectively. In the second decade, there was a smaller difference due to the numbers lost to follow-up: 820 compared to 673 presentations respectively.

Just over half (56%) of the patients in the alcohol group were male, the median (average) age was 28, and they were usually put into Triage Category 3, which meant they had potentially life-threatening conditions and needed treatment within 30 minutes. The majority (64%) presented in the emergency department in the late evening and over night, between 20:00 hours and 06:00 hours.

During the follow-up period, 44 patients died, of which 31 were patients with alcohol-related diagnoses who died a median of eight years after 2002, and 13 were from the control group, who died a median of 13 years after 2002.

Prof. Richardson said: “Society and policy-makers should recognise the major role that alcohol plays in illnesses and death in our community, and the need for preventative measures. Alcohol consumption is a significant part of Australasian society, but this study shows that consumption that leads to a visit to a hospital’s emergency department is extremely risky in the long term.

“We have been studying the effects of alcohol presentations in emergency departments for over a decade. The next logical step is a long-term trial of an alcohol intervention programme in the emergency department to see if this really does reduce burden of drink-related effects on patients and hard-pressed emergency staff.”

Professor Youri Yordanov from the St Antoine Hospital emergency department (APHP Paris), France, is Chair of the EUSEM 2023 abstract committee and was not involved in the research. He said: “This study shows a pattern that is familiar to many of us working in emergency departments around the world: alcohol abuse is responsible for a large proportion of patients vising emergency departments. Not only does this place significant burdens on emergency departments that are already over-stretched for a variety of reasons, including ageing populations and under-funding, but it shortens people’s lives too. Initiatives to intervene at an early stage to help prevent repeat visits to emergency departments and the problems associated with alcohol abuse would be very welcome and we look forward to seeing the results of further studies into this.”

(ends)

[1] Abstract no: OA110, “Long term outcomes after alcohol-related presentation to ED” by Drew Richardson, in the “Education, training and toxicology” oral session, Tuesday 19 September at 11.05-12.30 hrs CEST, Room 131. https://cm.eusem.org/cmPortal/Searchable/EXA/config/normal/redirectconfig/normal/redirectconference/EUSEM23#!sessiondetails/0000014290_0

Note: When obtaining outside comment, journalists are requested to ensure that their contacts are aware of the embargo on this release.

Funding: The research was supported by the Australian National University which provides time for all first-year medical students to undertake a research project. 

About European Society of Emergency Medicine (EUSEM)

Embargo: 00.01 hrs CEST on Monday 18 September 2023

Barcelona, Spain: If doctors in hospital emergency departments are trained to carry out ultrasound on patients with suspected deep vein thrombosis (DVT), they can nearly halve the time the patients spend in these departments.

Dr Ossi Hannula, an emergency medicine specialist at the Wellbeing Services County of Central Finland, Jyväskylä, Finland, who presented the findings at the European Emergency Medicine Congress today (Monday), said his findings could help to reduce overcrowding in emergency departments and improve death rates by enabling patients at greatest risk of dying, usually from non-DVT-related problems, to be treated more quickly by emergency staff.

“Prolonged stays in emergency departments are linked to emergency department crowding,” he said. “The longer a patient stays in an emergency department, the higher are the death rates and the risks of other complications, the longer their stay in a hospital ward, the lower the patient satisfaction, and the higher the financial costs and the burden on emergency department staff.”

DVT is a blood clot in a vein, normally in the leg, and it is a common condition in patients arriving in emergency departments, accounting for 1-2% of all such visits. An ultrasound scan, usually performed by radiographers or radiologists in the hospital’s imaging department, is the normal way to diagnose it, and treatments include anticoagulant medicines (or “blood thinners”) to stop the clot growing and to prevent it breaking off and traveling in the blood stream to other parts of the body, such as the lungs. If this happens, it can be fatal.

Dr Hannula’s earlier studies had shown that if general practitioners working in primary care were taught to perform ultrasound scans on patients with suspected DVT, they referred fewer patients to hospital emergency departments, resulting in less crowding and lower costs. He decided to see if ultrasound performed by emergency physicians instead of radiographers and radiologists could reduce the time patients spent in emergency departments.

Between October 2017 and October 2019, 93 patients with a suspected DVT were recruited to the prospective study carried out in two hospitals: Tampere University Hospital and Kuopio University Hospital. They were included in the study if an emergency doctor who had been trained to perform ultrasound scans examined them and performed the necessary ultrasound themselves. This is called “point-of-care ultrasound” (POCUS). POCUS can be done at the bedside in the emergency department, in the hospital ward, in an ambulance, or in the middle of a natural disaster. If the doctor thought a patient should also be referred to the imaging department, they could do this as well as performing POCUS themselves. The patients were aged over 18 years and able to give informed consent in Finnish.

“The aim of point-of-care ultrasound is to answer specific questions such as: ‘Is there a deep venous thrombosis that causes this leg to swell?’ or ‘Are there gallbladder stones present causing the abdominal pain?’” said Dr Hannula.

Eleven emergency medicine specialists and junior doctors in the two hospitals examined the patients in the study. Afterwards, Dr Hannula compared the results with a control group of 135 patients who arrived in the same emergency departments with suspected DVT on the same days but were sent for ultrasound scans in the hospitals’ imaging departments.

“We found that patients undergoing the standard ultrasound examination spent an average of 4.51 hours in the emergency departments, while the group receiving point-of-care ultrasound spent an average of 2.34 hours in the emergency departments – a difference of 2.16 hours,” said Dr Hannula.

“There have been mixed results from previous studies of point-of-care ultrasound that investigated how it affected the length of stay in emergency departments. It seems that the results can depend on the setting of the study. As this study was carried out in two different emergency departments in academic hospital, the results are convincing.

“The crowding in emergency departments is an increasing threat to patient safety as well as staff wellbeing. Using point-of-care ultrasound is one way of tackling this threat by reducing an unnecessary delay in decision making.”

Dr Hannula now plans to see if a similar reduction in length of stay in emergency departments can be achieved in other studies, for instance, for gallstones.

Professor Youri Yordanov from the St Antoine Hospital emergency department (APHP Paris), France, is Chair of the EUSEM 2023 abstract committee and was not involved in the research. He said: “This study shows that point-of-care-ultrasound is able to provide swift and precise diagnoses for patients who come to emergency departments with suspected deep vein thrombosis. An initiative like this that can reduce the time that patients have to wait in emergency departments is very welcome, especially as it has the potential to reduce the pressure on staff and improves the patients’ experience.”

(ends)

[1] Abstract no: OA89, “Emergency physician performed point-of-care ultrasound on patients suspected of deep venous thrombosis reduces length of stay in emergency department: a prospective multicentre study”, by Ossi Hannula, in the Best Abstracts session, Monday 18 September, 16.35-18.00 hrs CEST, room 131. https://shorturl.at/gmABV

Note: When obtaining outside comment, journalists are requested to ensure that their contacts are aware of the embargo on this release.

Funding: none.

About European Society of Emergency Medicine (EUSEM)

Embargo: 00.01 hrs CEST on Monday 18 September 2023

Barcelona, Spain: Bystanders are less likely to give cardiopulmonary resuscitation (CPR) to women than men, particularly if the emergency takes place in a public area, according to research presented at the European Emergency Medicine Congress today (Monday). The study also shows that in private locations older people, especially older men, are less likely to receive CPR.

The researchers say that CPR saves lives and urge people to learn how to perform CPR and to give it without hesitation to anyone who needs it, regardless of gender, age or location.

The research was presented by Dr Sylvie Cossette, a PhD nurse researcher at the Montreal Heart Institute research center, Canada. She conducted the research with Dr Alexis Cournoyer, an emergency medicine physician and researcher at the Hôpital du Sacré-Coeur de Montréal, Canada.

Dr Cournoyer said: “In an emergency when someone is unconscious and not breathing properly, in addition to calling an ambulance, bystanders should give CPR. This will give the patient a much better chance of survival and recovery.”

Dr Cossette added: “We carried out this study to try to uncover factors that might discourage people from delivering CPR, including any factors that might deter people from giving CPR to a woman.”

The researchers used data from records of cardiac arrests that happened outside of hospital in Canada and the US between 2005 and 2015, including a total of 39,391 patients with an average age of 67. They looked at whether or not a bystander performed CPR, where the emergency took place, and the age and gender of the patient.

They found that only around half of patients received CPR from a bystander (54%). Overall, women were slightly less likely to be given CPR (52% of women compared to 55% of men).

However, when the researchers looked only at cardiac arrests that happened in a public place, such as the street, the difference was greater (61% of women compared to 68% of men). These lower rates of CPR in public were found in women regardless of their age.

When the researchers looked at cardiac arrests that happened in a private setting, such as a home, the data indicated that with every ten-year increase in age, men were around 9% less likely to be given CPR during a cardiac arrest. For women having a cardiac arrest in a private setting the chances of receiving CPR were around 3% lower with every ten-year increase in age.

Dr Cournoyer said: “Our study shows that women experiencing a cardiac arrest are less likely to get the CPR they need compared to men, especially if the emergency happens in public. We don’t know why this is the case. It could be that people are worried about hurting or touching women, or that they think a woman is less likely to be having a cardiac arrest. We wondered if this imbalance would be even worse in younger women, because bystanders may worry even more about physical contact without consent, but this was not the case.”

Dr Cossette said: “We would like to study this issue in greater detail to understand what lies behind the difference. This could help us make sure that anyone who needs CPR gets it, regardless of gender, age or location.”

Professor Youri Yordanov from the St Antoine Hospital emergency department (APHP Paris), France, is Chair of the EUSEM 2023 abstract committee and was not involved in the research. He said:  “CPR saves lives, but sadly not everyone who suffers a cardiac arrest will get the CPR they need. This study gives us some clues about why that’s the case. A cardiac arrest can happen anytime and anywhere, so we all need to learn CPR and to be willing to perform it without hesitation.”

(ends)

[1] Abstract no: OA100, “Do age and location affect whether bystander cardiopulmonary resuscitation is provided to women experiencing an out-of-hospital cardiac arrest?” by Sylvie Cossette et al, in the Best Abstracts session, 17.31  hrs CEST, Monday 18 September, Room 131.

Note: When obtaining outside comment, journalists are requested to ensure that their contacts are aware of the embargo on this release.

Funding: Hôpital du Sacré-Coeur de Montréal: the chief of the lab Dr Alexis Cournoyer (research funding), and the Montreal Heart Institute research center (presentation at the EUSEM Congress funding)

Mandated disclaimer from the database registry: This work is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute or National Institutes of Health or any of the other stated funding agencies.

About European Society of Emergency Medicine (EUSEM)

Embargo: 00.01 hrs CEST on Sunday 17 September 2023

Barcelona, Spain: Half of patients discharged from the emergency department need only five tablets or fewer of morphine 5 mg or an equivalent opioid pain killer, according to new research presented at the European Emergency Medicine Congress today (Sunday).

The recent crisis in opioid abuse has been partly attributed to over-prescription, particularly for chronic pain, and doctors have become cautious about giving these drugs to patients.

However, researchers say it is vital that patients are given sufficient medication to help them recover from pain and injury, and the new study will help emergency medicine doctors to get the balance right.

The research was presented by Professor Raoul Daoust, from the University of Montreal, Canada. He said: “Opioids such as morphine can be very beneficial for patients suffering acute pain, for example when they have injured their neck or broken a bone. However, patients are often prescribed too many opioid tablets and that means unused tablets are available for misuse. On the other hand, since the opioid crisis, the tendency in the USA is to not prescribe opioids at all, leaving some patient in agonising pain.

“With this research I wanted to provide a tailored approach to prescribing opioids so that patients have enough to manage their pain but almost no unused tablets available for misuse.”

Professor Daoust and his colleagues recruited 2,240 adult patients who were treated at one of six hospital emergency departments in Canada for a condition that causes acute pain. All were discharged with an opioid prescription and were asked to complete a pain medication diary for the following two weeks.

Overall, half of patients took five morphine tablets (5mg) or fewer. However, the number of tablets that would be enough for most patients for two weeks varied greatly according to the patient’s painful condition. For example, patients suffering from renal colic or abdominal pain needed only eight tablets and patient with broken bones needed 24 tablets.

Professor Daoust said: “We found that, in general, patients consume few opioids, but this varies depending on the type of painful condition. Our findings make it possible to adapt the quantity of opioids we prescribe according to patient need. We could ask the pharmacist to also provide opioids in small portions, such as five tablets initially, because for half of patients that would be enough to last them for two weeks.”

The researchers now hope to apply their results in the clinic to evaluate whether they have an impact on long-term use and misuse of opioids.

Professor Youri Yordanov from the St Antoine Hospital emergency department (APHP Paris), France, is Chair of the EUSEM 2023 abstract committee and was not involved in the research. He said: “It’s estimated that millions of people around the world are struggling with opioid addiction and more than 100,000 people die of opioid overdose every year. These drugs play an important role in emergency medicine, but we need to ensure they are prescribed wisely.

“This study shows how opioid prescriptions could be adapted to specific acute pain conditions, and how they could be dispensed in relatively small numbers at the pharmacy to lower the chance of misuse. This research could provide a safer way to prescribe opioids that could be applied in emergency departments anywhere in the world.”

(ends)

[1] Abstract no: OA077, “Opioids for acute pain: how much to prescribe to minimize unused medication? (OPUM Study)” by Raoul Daoust et al, in the All sorts of pain session, 17:26 hrs CEST, Sunday 17 September, Room 131.

Note: When obtaining outside comment, journalists are requested to ensure that their contacts are aware of the embargo on this release.

Funding: The Canadian Institutes of Health Research Fund (CIHR)

About European Society of Emergency Medicine (EUSEM)

For immediate release on Wednesday 13 September 2023

Barcelona, Spain: The artificial intelligence chatbot ChatGPT performed as well as a trained doctor in suggesting likely diagnoses for patients being assessed in emergency medicine departments, in a pilot study to be presented at the European Emergency Medicine Congress, which starts on Saturday [1].

Researchers say a lot more work is needed, but their findings suggest the technology could one day support doctors working in emergency medicine, potentially leading to shorter waiting times for patients.

The study was by Dr Hidde ten Berg, from the department of emergency medicine and Dr Steef Kurstjens, from the department of clinical chemistry and haematology, both at Jeroen Bosch Hospital, 's-Hertogenbosch, The Netherlands.

Dr ten Berg told the Congress: “Like a lot of people, we have been trying out ChatGPT and we were intrigued to see how well it worked for examining some complex diagnostic cases. So, we set up a study to assess how well the chatbot worked compared to doctors with a collection of emergency medicine cases from daily practice.”

The research, which is also published this month in the Annals of Emergency Medicine [2], included anonymised details on 30 patients who were treated at Jeroen Bosch Hospital’s emergency department in 2022. The researchers entered physicians’ notes on patients’ signs, symptoms and physical examinations into two versions of ChatGPT (the free 3.5 version and the subscriber 4.0 version). They also provided the chatbot with results of lab tests, such as blood and urine analysis. For each case, they compared the shortlist of likely diagnoses generated by the chatbot to the shortlist made by emergency medicine doctors and to the patient’s correct diagnosis.

They found a large overlap (around 60%) between the shortlists generated by ChatGPT and the doctors. Doctors had the correct diagnosis within their top five likely diagnoses in 87% of the cases, compared to 97% for ChatGPT version 3.5 and 87% for version 4.0.

Dr ten Berg said: “We found that ChatGPT performed well in generating a list of likely diagnoses and suggesting the most likely option. We also found a lot of overlap with the doctors’ lists of likely diagnoses. Simply put, this indicates that ChatGPT was able suggest medical diagnoses much like a human doctor would.

“For example, we included a case of a patient presenting with joint pain that was alleviated with painkillers, but redness, joint pain and swelling always recurred. In the previous days, the patient had a fever and sore throat. A few times there was a discolouration of the fingertips. Based on the physical exam and additional tests, the doctors thought the most likely diagnosis was probably rheumatic fever, but ChatGPT was correct with its most likely diagnosis of vasculitis.

“It’s vital to remember that ChatGPT is not a medical device and there are concerns over privacy when using ChatGPT with medical data. However, there is potential here for saving time and reducing waiting times in the emergency department. The benefit of using artificial intelligence could be in supporting doctors with less experience, or it could help in spotting rare diseases.”

Professor Youri Yordanov from the St Antoine Hospital emergency department (APHP Paris), France, is Chair of the EUSEM 2023 abstract committee and was not involved in the research. He said: “We are a long way from using ChatGPT in the clinic, but it’s vital that we explore new technology and consider how it could be used to help doctors and their patients. People who need to go to the emergency department want to be seen as quickly as possible and to have their problem correctly diagnosed and treated. I look forward to more research in this area and hope that it might ultimately support the work of busy health professionals.”

(ends)

[1] Abstract no: OA66, “ChatGPT Effectively Generates Differential Diagnosis Using Emergency Department Physician Notes” by Hidde Ten Berg, in the Out-of-the-hospital and the newest technology session, 09:42 hrs CEST, Tuesday 19 September Room 131.

[2] “ChatGPT and Generating a Differential Diagnosis Early in an Emergency Department Presentation”, by H. ten Berg et al, In press, Annals of Emergency Medicine. DOI: https://doi.org/10.1016/j.annemergmed.2023.08.003

Note: When obtaining outside comment, journalists are requested to ensure that their contacts are aware of the embargo on this release.

Funding: No external funding

About European Society of Emergency Medicine (EUSEM)