California researchers have created an algorithm that uses just five data inputs to predict a patient’s COVID-19 prognosis — a big improvement compared to the bad old days of last year.
In the early days of the pandemic, medical professionals reasonably collected various categories of patient data to create a snapshot picture of how ill the patient was and estimate a prognosis. In today’s complicated world of this infectious disease, a bedside tool could help a physician assess a prognosis much faster.
COVID-19 has been full of bad surprises, sending a lot of people in the hospital and the ICU — well, you know the story. Hospital resources are stretched to the maximum, both in terms of equipment and people themselves. Running out of available ICU beds has become a hospital’s worth nightmare in the pandemic, so researchers wanted a tool that would help them forecast ICU usage.
Tools that offer a real-time prognosis of patients are few and far between. For doctors, it’s very useful to know, once a person is admitted to hospital, what are the chances the person will end up in the ICU. The new algorithm does just that: it’s designed to be give medical personnel some quick answers upon the patient’s initial hospitalization.
Assad Oberai, Hughes Professor in the Department of Aerospace and Mechanical Engineering, explained how the biomarkers can tell the degree to which COVID-19 is advancing in the patient. Oberai said this is a disease that “isn’t affecting just one system in the body, but is affecting many systems. Many things are going wrong at the same time, and by focusing on these five features, you can get a picture of how the disease is progressing.”
So based on this, if we measure what is going wrong with a patient, we can forecast how the disease is likely to progress. The indicators that researchers used to derive a patient’s prognosis are:
respiratory distress (breathing rate and blood oxygen level of patient);
immune response (likelihood of sepsis);
circulatory system (likelihood of blood clots);
Since the predictors rely only on quantitative data, they are less prone to errors and subjectivity, which make them scalable and applicable in all sorts of different scenarios. What is more, the model is simple and easy to use.
The team behind the algorithm are from the Viterbi School of Engineering and Keck School of Medicine at the University of Southern California. They presented their work in an article published in Scientific Reports in an article called “Machine learning based predictors for COVID‑19 disease severity.”
“Predictors of the need for intensive care and mechanical ventilation can help healthcare systems in planning for surge capacity for COVID-19,” the researchers explain in the study.
It is an old but imperative rule to be proactive rather than reactive in pandemics. In this instance, the authors called for a proactive approach towards resources such as ICU bends and ventilators.
“Given the urgency for resource allocation and optimization, we sought to identify patient-level clinical characteristics at the time of admission to predict the need for ICU care and mechanical ventilation in COVID-19 patients.”
Their study cohort comprised of 212 patients (123 males, 89 females) with an average age of 53 years, of which 74 required intensive care at some point during their stay, and 47 required mechanical ventilation. Neha Nanda, medical director of infection prevention and antimicrobial stewardship at Keck Medicine of USC, said that maybe, with time, “this proactive approach is something we can adopt universally to all emerging infectious diseases.”
However, in order for this to happen, the results would have to be validated on a much larger cohort. For now, only data obtained at the time of initial presentation was included as input to the predictive models. The need for ICU admission and mechanical ventilation at any time during hospitalization were selected as outcomes.
“The results presented in this study demonstrate that data acquired at or around the time of admission of a COVID-19 patient to a care facility can be used to make an accurate assessment of their need for critical care and mechanical ventilation,” the researchers conclude.