Reducing user error - A big win flying under the radar?

Several years ago, I bought a new cycling helmet. I asked the shop assistant about the practical difference between a £30 helmet and those priced in excess of £100. I only had three concerns, safety, comfort and how stupid it would make me look. It turns out that the big difference was weight – and we’re not talking bags of sugar difference, we’re talking about a few grams. I tapped my belly, smiled – and picked up the cheaper one.

I completely understand that in the world of cycling components, weight is a big factor. Furthermore, if cycling is your passion or you compete at a reasonable level, you should want the best kit that you can afford. However, for a tight-fisted bloke who wants a Sunday ride out with the kids on a mid-90s steel frame mountain bike and packing several kilos of excess body fat, a 25g lighter helmet was not the big win.

When it comes to lateral flow technology, we’re obsessed with all manner of marginal gains – and rightly so. We should be pushing for greater sensitivity, better limits of detection, more sustainable materials, longer shelf life, more affordable products, new applications and reducing test variability.

One area of improvement is (I believe) something of an afterthought. When developing a new test, we tend to consider usability in terms of the number of user steps. Fewer steps equal fewer chances for someone to mess things up. While UX (user experience) is well-considered, asking experts in lateral flow technology to think of every way that a complete novice might innocently misuse the test is folly. It’s an area where familiarity is a disadvantage.

Is 'reducing user error' going under the radar?

For example, imagine that someone buys a test that they need to conduct in private. They remove the components, read the instructions for use, lay everything out in the bathroom, take a shower, dry-off and prepare their sample. They mix the sample in the buffer as instructed and run the test.

Little do they know that the chances of an accurate result evaporated while they were in the shower as the humidity slowly destroyed the properties of the materials required for the device to work. We know that the sealed foil pack and desiccant protects the test – the user doesn’t – and why would they?

There are more extreme examples of user-error that no one could legislate for. I am told that several years ago, someone complained about a negative pregnancy test reading when they were in fact pregnant. It caused some consternation but upon investigation, it turned out that the woman concerned wasn’t feeling great, so asked her mum to do the test for her. Neither the mother or ‘mother-to-be’ were aware of their error.

Reducing user-error isn’t easy but if an assay had a chance of catastrophic failure because of a weakness in the chemistry, we would doubtless spend several weeks of expensive lab time resolving the issue. The instruction for use (IFU) however, is sometimes reduced to an illustration on a package insert. Even the very best IFUs, while accurate, are probably inadequate.

I guess my solution would be to throw belts and braces at the issue. Cover as many bases as possible to improve the chances of everyone being able to run the test as it was intended. Employ a decent design agency to produce the IFU and stress test them with the aid of an uninitiated user group. Don’t bury the IFU in the small print. Record an instructional video to explain everything step-by-step and explain why the steps are important. Don’t ignore questions and complaints online. The last thing you need is a disgruntled subreddit community undermining your product simply because the UX was sub-par.

This isn’t a potential marginal gain; it’s a potential big win. Whatever helps, seriously consider doing it.

As ever. Thoughts welcome.

Richard