Shocker: People are Really Bad at Evaluating their own Symptoms

Cognitive Bisases are everywhere

You know the Dunning-Kruger effect? It's a cognitive bias that says that the less knowledgeable someone is on a topic, the more likely they are to be over-confident on their knowledge. Spend any time on twitter or linkedin and you see this effect in action.

The Problem with Subjective Symptoms Evaluation

Turns out there is a similar phenomena in the realm of clinical practice. This fascinating recent paper in Lung proves that there is a big difference between people's perception of how much they cough (subjective cough rate) and how much they actually cough (objective cough rate).

We are well into the era of precision medicine - where clinical decisions are based on objectively measured data (sensor readings, lab work), yet there are still corners of medical sciences where clinical decisions rely almost exclusively on patients' own evaluation of their symptoms. Cough management is one of them .

The fact that people are REALLY bad at evaluating their own symptoms, even while they are REALY confident that their evaluation is on point, is not just a fascinating quirk to be studied by sociologists.?It actually has massive (negative) impact for both patient care and the broader landscape of medical research and science.

The last time the FDA approved a cough medication was in 1958 . Is this because coughing has been a solved problem in modern times? Not at all - people cough more than they ever have and conditions like chronic coughing, asthma or COPD continue to affect signiifcant segments of population world-wide.??

Or could it be that we reached peak respiratory health science in the '50s and no additional innpvation is needed or possible? Definitely not, In fact, the drug approved in 1958 is closer to morphine (or snake oil) than it is to modern medicine .

So what is going on here?

Turns out that our scientific standards have evolved significantly from the 50s, while our ability to generate rigorous data to satisfy those standards - for things like cough counting - has not.?

That is because, in respiratory care we still rely on patient reported data when it comes to quantifying things like coughing. And because these reports are unreliable, the foundational data for research can be inherently flawed. This not only affects the validity of studies but can also lead to misguided treatment recommendations or therapeutic strategies.

For instance, in the development of a new drug for a condition with subjective symptoms, if the baseline symptom severity is inaccurately reported by patients, the perceived efficacy of the drug might be over or underestimated. This could lead to inappropriate dosing, side effect misattribution, or even the premature discontinuation of a potentially beneficial treatment.

Well, modern regulators do not like all of that,?at all.

it gets worse still. This inability to report/ recall one's own symptoms becomes even more pronounced over extended periods of time.

Consider the case of chronic coughers. A patient might experience a persistent cough for months or even years, gradually adapting to it as a 'new normal'. Over time, they may underestimate the severity or frequency of their cough, or even dismiss associated symptoms like fatigue or shortness of breath. This misjudgment can lead to delayed diagnoses, inappropriate treatments, or missed opportunities for early intervention.

The Wrong Technology Makes Things Worse

Incumbent technology solutions - such as they are - are not helping. Turns out cough is highly stochastic, so cough rates in chronic sufferers can vary wildly from a day to the next. The proverbial good days and bad days that chronic sufferers are familiar with.

Incumbent tech solutions have been limited to max 24h monitoring . This is because these solutions are basically tape recorders that record everything that is going on in a patient's life, and then humans listen to the tape looking for symptoms.?

Here is an example of 30 days in the life of a cougher whose hourly cough rate varies between 1 cough/ hour to more than 400 coughs/ hour (!!).

Imagine a doctor would limit clinical observation on any random 24h in the life of this person. They would 100% miss what is going on here.

Or imagine a health system screening this person. If they'd be monitored during a random low-cough window, they might be dismissed as perfectly healthy, although they'd be coughing their brains out (400 coughs/ h) next day.

We can do better. We need objective measures in medical assessments and clinical trials.

AI To the Rescue

Hyfe 's technology can objectively detect coughing in real time, all the time.

This is changing our understanding of coughing - for things like predicting COPD exacerbations and is also unlocking unprecedented opportunities to innovate.?The tech has already been validated in more than 45 studies world wide and more than a dozen peer reviewed papers .?