Agile Tests Done at Bedside Could Help Meet Challenges in Critical Care

This approach does not necessarily include understanding the definite causes behind those symptoms. However, even though some patients might present similar symptoms, the underlying biology of their illnesses can vary, which helps to explain why patients with the same signs and symptoms often respond differently to treatments. In recent years, technological advances in molecular sciences, big data, and machine learning have allowed researchers to dig deeper in understanding illnesses and symptoms. This new knowledge can radically change how health practitioners approach critical care at the individual patient level.

In a recent paper published in Nature Medicine, researchers at Queen's University (Kingston, ON, Canada) have argued that it is time for a paradigm shift in critical care. David Maslove (School of Medicine) who along with his colleagues published the paper spoke to the Gazette about how the COVID-19 pandemic advanced critical care research and what knowledge gaps we still need to address to put this new approach into practice. Over the last decade or two, as new tools in molecular medicine such as genomics and gene expression profiling came into wider use, health practitioners began to apply them in critical care. Those technologies allow a much deeper look at what is happening at the level of the cell and the genome when somebody is critically ill. They also generate massive quantities of data – data that traditional statistical methods do not always handle in intuitive ways. As a result, machine learning approaches perfectly complement the new data derived from these technologies: they reveal hidden patterns that can provide insights into the heterogeneity or diversity of critical illness syndromes, according to Maslove.

Since a group of critically ill patients were all affected by the same pathogen – the COVID19 virus, SARS-CoV-2, Maslove believes that there was a tremendous opportunity to test potential treatments in a group that was more homogeneous than usual, increasing the chances that a therapy would on average affect everyone the same way. Researchers were able to promptly design and perform randomized clinical trials to test potential therapies, and they identified some effective treatments. Prior to that, because clinical trials were treating heterogeneous groups, some patients benefitted from the treatment and others did not. It was harder to tell if the proposed treatment worked for anyone. On a practical level, researchers learned some key lessons about the importance of research infrastructure to be able to collect and analyze data quickly, and the importance of being ready to efficiently enroll patients into clinical trials, according to Maslove. They also reaffirmed that randomizing treatments is the best way to identify which ones work best. Critical care scientists from all over the world were working together towards a singular goal, reaffirming the importance of international collaboration.

Maslove believes that there is need to clarify what it means to respond to a treatment in the ICU context, where cases are so dynamic and complex that it can be difficult to understand whether a treatment has provided benefit in any given case. There is also a need for biomarkers that can be used to identify patients who are more or less likely to respond to a given treatment. In critical care, a particular challenge is that health practitioners cannot send a blood test off to a lab far away and wait for a complex testing like RNA sequencing to be completed. There is a need for agile tests that can be done at the bedside, according to Maslove. With this kind research, there is a shift towards precision medicine, an approach pioneered in cancer treatment, which offers tremendous potential benefits. By identifying which patients are most likely to respond to which treatments, health practitioners can increase the efficiency of the care that they provide. The precision medicine approach also spares patients from ineffective treatments.

Related Links:
Queen's University