People are more willing to let software route their money than explain their blood test, and they are not being irrational. The consequences of a wrong answer are different, and the ability to detect a wrong answer is far lower.
What actually erodes trust
- Unsourced confidence. A figure stated with no indication of where it came from cannot be checked, so it has to be taken on faith — which is exactly what a sceptical reader will not extend.
- Never admitting uncertainty. A system that is equally confident when it is right and when it is guessing has trained the reader to discount everything it says.
- Overreach. One answer that strays into diagnosis undermines the hundred that were careful.
- Opacity about data. If someone cannot tell what happens to their results, they will assume the worst, and quite often they will be right.
What rebuilds it
Showing the working. Naming the source and its date. Pointing at the range printed on the document rather than at an authority the reader cannot see. Saying plainly when the answer is not available. And stating the limits before being asked, because in clinical software the refusals are more informative than the features.
The uncomfortable part
All of this makes a product sound less impressive. A system that says “this is what the number is, here is the range your laboratory printed, here is what is worth asking” is a smaller claim than one that says it will analyse your health.
It is also the only version a reasonable person should believe, and the smaller claim is the one that survives contact with a real patient and a real doctor.
This article explains how our technology works. It is health information, not medical advice.