The first impression is the one. The popular saying that expresses the human capacity to analyze and create conclusions quickly also valid for the hospital emergency room, suggests a study by 725 adults attended at the University of São Paulo School of Medicine Hospital of Medicine (FM-USP).
Supported by FAPESP and published in BMJ Supportive & Palliative Care, the work showed that, in comparison between the OPINION OF THE EMERGENCIAL PHYSICIAN on the probability of patient recovery and a Predictive score calculated by an artificial intelligence model (AI) human impression came closer to reality, with rate of 79.2%.
“In a world of predictive models based on artificial intelligence, the study sought to analyze the value of the medical eye and the importance of this professional’s first impression of the patient. I am a great AI enthusiast, but the results bring something that goes against what we are currently studying: the doctor’s eyes matters a lot,” says Júlio César Garcia de Alencar, professor at the Bauru Medical School (FMBRU-USP).
The investigation consisted of asking emergencist doctors if they would be surprised if the newly tied patient in the emergency room died in a year.
After this first analysis, the group of researchers followed the 725 patients during hospitalization to check if the outcome conditioned the opinion of professionals. All study participants were classified as serious cases.
“The interesting thing here is to note that the doctor had little information about the patient’s prior health conditions, only the report of the nurses who screened, with the main complaint of the individual and the risk classification based on the Manchester protocol, which guides the urgency of care through the use of bracelets with different colors [vermelho, risco de morte; laranja, casos urgentes; amarelo, risco não imediato; verde, casos menos graves; e azul, sem urgência]. It was just the look, his knowledge and the previous experiences that, in fact, told for this assessment, ”explains Alencar.
The work compared the results with a predictive clinical score potentially used in AI. “We confront the opinion of doctors with The Quick Sequential Organ Failure Assessment [qSOFA]a score validated since 2016 and that has been gaining space in medicine precisely because they have few clinical variables, such as respiratory rate, blood pressure and level of consciousness, ”he says.
Unfolding
Besides exploring the issue Humans versus machine the study has interesting developments, which can be implemented in the emergency room of hospitals. This is because the work is the first to use the surprise question (would you be surprised if the patient treated died in a year?) In the emergency department.
The surprise question is a strategy validated as a screening tool for patients who should be accompanied by palliative care teams – medical approach that improves the quality of life of patients and family in the face of terminal diseases through relief from suffering and pain treatment.
Previous studies had already shown the accuracy of the surprise question in specific groups, such as patients who do hemodialysis or Alzheimer’s disease.
In the study in the Emergency Room of the Hospital das Clínicas de São Paulo, doctors said in 20% of cases that they would not be surprised if the admitted patients died in a year. “This shows that there is some criterion, thinking of terminality. Therefore, there is a need to adapt the conduct of the Emergency Department to the patient prognosis, that is, to bring the possibility of palliative care to the emergency room, to provide assistance to the patient and his family,” Alencar tells Fapesp.
The researcher stressed, however, that more studies are needed before implementing some intervention in this regard. “This is phase two of our project. The next step will be to investigate the possibility of an intervention plan based on palliative care for this patient,” he said.
See also: Artificial Intelligence can generate “global annihilation,” experts say
This content was originally published in a medical look is worth more than AI assessment, says study on the CNN Brazil website.
Source: CNN Brasil
Charles Grill is a tech-savvy writer with over 3 years of experience in the field. He writes on a variety of technology-related topics and has a strong focus on the latest advancements in the industry. He is connected with several online news websites and is currently contributing to a technology-focused platform.
