Generalisable long COVID subtypes: Findings from the NIH N3C and RECOVER programmes 2022 Reese et al

Summary

Background
Stratification of patients with post-acute sequelae of SARS-CoV-2 infection (PASC, or long COVID) would allow precision clinical management strategies. However, long COVID is incompletely understood and characterised by a wide range of manifestations that are difficult to analyse computationally. Additionally, the generalisability of machine learning classification of COVID-19 clinical outcomes has rarely been tested.

Methods
We present a method for computationally modelling PASC phenotype data based on electronic healthcare records (EHRs) and for assessing pairwise phenotypic similarity between patients using semantic similarity. Our approach defines a nonlinear similarity function that maps from a feature space of phenotypic abnormalities to a matrix of pairwise patient similarity that can be clustered using unsupervised machine learning.

Findings
We found six clusters of PASC patients, each with distinct profiles of phenotypic abnormalities, including clusters with distinct pulmonary, neuropsychiatric, and cardiovascular abnormalities, and a cluster associated with broad, severe manifestations and increased mortality. There was significant association of cluster membership with a range of pre-existing conditions and measures of severity during acute COVID-19. We assigned new patients from other healthcare centres to clusters by maximum semantic similarity to the original patients, and showed that the clusters were generalisable across different hospital systems. The increased mortality rate originally identified in one cluster was consistently observed in patients assigned to that cluster in other hospital systems.

Interpretation
Semantic phenotypic clustering provides a foundation for assigning patients to stratified subgroups for natural history or therapy studies on PASC.

Open access, https://www.thelancet.com/journals/ebiom/article/PIIS2352-3964(22)00595-3/fulltext

 

This reminds me of an old folk joke around here. Goes something like this. It's not really funny but whatever, let's call it a joke.

A farmer wants to know how cold winter will be, and goes to see the local native American wise guy for guidance, since they must know a lot about winter.

The wise guy tells the farmer that winter will be cold, so the farmer goes and chop wood.

The farmer then goes back again to make sure he chopped enough wood, and the wise guy says it will be even colder, so the farmer goes and chop more wood.

Back and forth a few times and a lot more chopped wood, the farmer decides to ask how the wise guy knows winter is going to be cold, and says he looks at how much wood got chopped for the winter, and noticed that a lot more wood just got chopped, that's how he knows winter will be colder than normal.​

This study tells us nothing about the patients, it only tells us about what healthcare services are recording about some of those patients, and we already know that for this entire category of illness they are completely incapable of doing this. It's circular logic as they use their past behavior to inform their future behavior, which is all guided by... analyzing their past behavior.

How can such basic lapse in logic never get noticed at all? There goes the exact same hubris behind the first NIH study: we already know that standard medical care is useless at not only treating this but even at merely noticing it. And they base an entire study based on what gets noticed. Good grief. We need adults in the room over there, this is insane.