One Card, Many Targets: The Case for Multiplexed Diagnostics
One Card, Many Targets: The Case for Multiplexed Diagnostics
Every point-of-care test that screens for exactly one thing is, in a sense, a missed opportunity. Patients rarely present with a single question. A fertility workup isn't really about one hormone. A wellness panel isn't really about one marker. And yet the default architecture for most rapid diagnostics is still single-analyte: one card, one target, one answer. Multiplexing — getting multiple results from a single sample application — isn't a nice-to-have feature. It's closer to what point-of-care testing should have looked like from the start.
Why Single-Analyte Testing Falls Short
The cost of single-analyte testing isn't really about the price of the test itself. It's the cost of fragmentation. A patient or user who needs three pieces of information has to run three tests, manage three samples, and reconcile three separate readouts — often taken at different times, under different conditions, which introduces its own source of error when those values actually need to be interpreted together. A hormone panel where the ratios between markers matter more than any single value in isolation is a perfect example of where single-analyte testing structurally can't deliver the right answer, no matter how accurate each individual test is.
Two Real Approaches to Multiplexing
There isn't one way to build a multiplexed diagnostic, and it's worth being specific about what's actually involved, because the engineering trade-offs are real.
The first approach is a multi-spot vertical flow format — a chemical spot test with several discrete reaction zones on a single card. Each spot is its own independent test for a different analyte, which means sample has to be applied to each spot individually rather than added once and distributed automatically across the card. That's a meaningfully different user experience than a single-sample-application test, and it's a design trade-off worth being upfront about: you gain multiple results on one card, but you don't get the simplicity of a single application step. What you do get is a clean separation between assays, since each spot's chemistry doesn't have to coexist with or interfere with the others.
The second approach builds on traditional lateral flow with AI-assisted calibration layered on top. This is the harder engineering problem, because lateral flow tests weren't originally designed with multiplexing in mind, and packing multiple target lines onto one strip raises real questions about cross-reactivity between analytes. The honest answer is that AI calibration helps with this, but it doesn't solve it outright. In well-behaved assay configurations, the calibration model can often read past low-level cross-reactivity and still return an accurate result. In more complex configurations — where multiple analytes have overlapping chemistries or similar optical signatures — that cross-reactivity can be hard to fully correct for, and some combinations remain genuinely difficult to multiplex reliably with this method. That's not a failure of the approach; it's a constraint worth naming so that assay developers go in with the right expectations rather than assuming AI calibration is a universal fix.
Why This Is Worth the Engineering Difficulty
Given those trade-offs, it would be easier to just keep building single-analyte tests and call it a day. The reason that's the wrong call is that the value of a diagnostic test often isn't in any single data point — it's in the relationship between data points. A single hormone value tells you a fact. A panel of related hormones tells you a story, and that story is usually what the clinician, the consumer, or the field technician actually needed in the first place.
The diagnostics industry has spent a long time treating multiplexing as a premium feature reserved for lab-grade equipment. It shouldn't be. The point-of-care setting is exactly where the ability to get a fuller picture from a single sample matters most, because there often isn't a second chance to draw more sample or run another test. Getting multiplexing right at the point of care — honestly, including its real limitations — is one of the more underrated problems in diagnostics right now, and one of the more solvable ones.
