A Year of SamaClip: What We Got Right and What We Learned

SamaClip went into routine use at our centre in June 2025. A year on, it is worth saying plainly what has held up, what surprised us, and the things we had to change our minds about. This is not a victory lap. It is the kind of stock-take we would want from anyone asking us to trust their work.

What held up

The core bet was that a non-invasive, finger-clip screen would get people through a health check who would otherwise avoid one. That held. The single biggest barrier we kept hearing about was the finger prick, and removing it changed who walked in. People who had refused blood tests for years sat down for a screen that took about two minutes and drew no blood. That was the whole thesis, and it survived contact with reality.

The second bet was that we could validate accuracy across the skin tones of the people we actually serve, not just on average. That held too. When the validation results came in early this year, the oxygen-saturation accuracy stayed within a tight band across every skin-tone group we tested, including the darkest. We had been saying we would prove it; the data let us stop saying "we expect" and start saying "we measured."

What surprised us

We underestimated how much the operator matters compared to the device. A finger clip is easy. A person who can put a frightened patient at ease, explain a result without alarming them, and know when to insist on a referral is not easy, and is far more decisive for whether screening actually helps. The technology removed the needle. The operator did almost everything else. We have shifted a lot of our attention from the device to how we train and support the people using it.

We also underestimated the gap between screening someone and that person actually reaching follow-up care. It is one thing to flag a likely anaemia or a heart-rhythm irregularity. It is another for that patient, who may have travelled far and lost a day's wages, to come back, get the confirmatory test, and start treatment. A flag with no follow-through is close to useless. Much of our second year has been about closing that gap rather than screening more people faster.

What we had to change

We had to get stricter, not looser, about what we claim. Early on it was tempting to let the screen sound like a diagnosis and let the follow-up monitoring band sound like it had the screen's evidence behind it. We pulled hard in the other direction. The screen flags; clinicians diagnose. The monitoring band tracks trends and carries no diagnostic claim. We would rather under-promise and keep trust than oversell and lose it the first time a claim does not hold.

We also got more honest about counting. It is easy to quote a footfall number or a catchment population and let it blur into an impression of how many people were actually screened with the validated tool. We separated those numbers deliberately and wrote about why. The validated-screen count is smaller and less impressive than the footfall figure, and it is the only one we will stand behind as a screening claim.

What the year actually taught us

If there is one lesson under all of this, it is that the hard part of community health was never the gadget. The gadget is the easy, fundable, photogenic part. The hard part is trust, follow-through, honest measurement, and showing up in the same place long enough that people believe you will be there next month. The screen got us in the door. Everything that matters happened after that, and most of it had nothing to do with technology.

We are going into year two with a tool we can defend, a team we trust, and a much clearer view of how far a screen alone can take you. Which is to say: through the door, and not one step further on its own.