These are Covid-19’s six symptom clusters, and here’s how they’ll help SA
Researchers say their work could help clinicians predict who will need respiratory support
One of the most challenging aspects of the Covid-19 pandemic is its unpredictability in those infected.
In many countries, including SA, this has made it difficult to ascertain, in time, who is going to need respiratory support, such as a ventilator.
SA also has other variables that make it difficult to determine the need and supply of respiratory support, such as provinces peaking at different times and the uneven strength of provincial health-care systems.
Now, a new study by an international team has outlined six different “symptom clusters” which could give clinicians an early indication of whether respiratory support may be needed.
It still needs to be peer-reviewed, but the study was based on data from 1,653 Covid-19 patients who were asked to regularly log their symptoms via an app. The scientists then used machine-learning algorithms to figure out if there were symptoms that could be clustered together.
The system is not flawless, but in the face of Covid-19 it could become a useful framework for SA and other countries now earmarked as being the worst hit by the pandemic.
“During the spread of Covid-19,” says lead researcher Dr Claire Steves, from King’s College London, “the strain on health-care systems has been felt globally and varying strategies for appropriate use of limited medical resources have been proposed.”
The strain on health-care systems has been felt globally and varying strategies for appropriate use of limited medical resources have been proposed.Dr Claire Steves
However, because the disease is so unpredictable and affects individuals differently, this has been difficult to achieve.
“Heterogeneity in disease and presentation is evident and the ability to predict required medical support ahead of time is limited. In this work, we sought to develop a clinical tool based on the time series of early development of Covid-19 that could be predictive of the need for high-level care in individuals more likely to seek medical help,” says Steves and the team.
So what are the six clusters?
Cluster one is mainly “upper respiratory tract symptoms”. This would include persistent cough and some muscle pain. Only 1.5% of patients in this group needed respiratory support and this was the most common cluster of symptoms.
Cluster two is also characterised by upper respiratory tract symptoms, but a fever and skipping meals are also in the picture. In this group, the percentage needing respiratory support rose to 4.4%.
Cluster three patients had diarrhoea and other gastrointestinal symptoms, but few other indicators, with only 3.7% requiring respiratory support.
Cluster four patients experienced severe fatigue soon after acquiring infection, then had continuous chest pains and coughs. Here, those needing respiratory support rose to 8.6%.
In cluster five, confusion, severe fatigue and skipped meals were grouped together. Here, almost 10% needed respiratory support.
The final cluster (marked respiratory distress, including early onset of breathlessness and chest pain, as well as confusion, fatigue and gastrointestinal symptoms) saw as many as 20% of patients needing respiratory support.
So how might this cluster or classification system make a difference in the real world?
“The ability to predict medical resource requirements days before they arise has significant clinical utility in this pandemic,” explain the researchers. “If widely utilised, health-care providers and managers could track large groups of patients and predict numbers requiring hospital care and respiratory support days ahead of these needs arising, allowing for staff, bed and intensive-care planning.”
Also, at a local level, the approach could allow patients to be monitored remotely by their primary health-care teams, “with alert systems triggered when individuals demonstrate symptomatology associated with a high-risk cluster”.
Prof Alastair Denniston, an artificial intelligence in health care expert at the UK’s University of Birmingham, was quoted in The Guardian as saying one limitation is that the “results were based on data from app users, meaning they may not hold across the wider population”.
He added, however, that it is “an important additional tool” in making sure those at highest risk get support on time.