Smart toilets and passive vitals: promising hardware, thin independent evidence

The pitch for smart-home health hardware — toilets that analyse what you leave behind, seats that read your heart from underneath, mirrors that estimate your pulse from your face — is that measurement should happen passively, without anyone remembering to do anything. The engineering is genuine and, in places, elegant. What is largely missing is the second half of the sentence: independent validation at the scale that would let a clinician trust a number, and any evidence that watching these numbers changes what happens to a patient.

The research prototype

The most cited proof-of-concept is a "smart toilet" developed at Stanford and published in Nature Biomedical Engineering [s1]. The mounted system combined uroanalysis test strips, video-based uroflowmetry, and stool classification, and identified individual users by a fingerprint scanner and an "analprint" — the distinctive features of a person's anoderm — so that a household's readings could be attributed to the right person [s1]. It is a serious piece of work, and it demonstrates the core idea: fixtures people already use every day could, in principle, collect longitudinal health data with zero added effort [s1].

But a proof-of-concept is exactly that. The study demonstrates feasibility on a small scale, not that the readings predict disease, that they hold up across diverse bodies and plumbing, or that acting on them improves health [s1]. The distance between "we built a device that measures X" and "measuring X in the home is clinically useful" is the entire game, and it is a distance most smart-home health products have not travelled.

The commercial version

The pattern repeats in the products reaching market. Casana's Smart Seat, a toilet seat that passively estimates blood pressure trends, heart rate, blood-oxygen saturation and respiratory rate from sensors reading an electrocardiogram, a photoplethysmogram and a ballistocardiogram, is the clearest commercial example [s2]. The company says it received regulatory clearance and reports, in its own validation, systolic blood-pressure estimates biased by roughly −2.7 mmHg and diastolic by about 1.2 mmHg against a reference monitor, with a blood-oxygen error of around 2.3% [s2].

Those are the company's figures, and that is the point worth flagging. Company-run validation against a reference monitor in a controlled setting is a necessary first step, but it is not the same as independent, peer-reviewed testing across a representative population, and it says nothing about whether a home blood-pressure trend read off a toilet seat improves outcomes over the validated arm cuff that guidelines already endorse. The house rule in reading any of this: a manufacturer's accuracy claim is a claim, not an endorsement.

Where contactless vitals actually stand

The broader field of contactless vital-sign sensing — radar and camera systems that estimate heart rate, breathing and sleep without touching the body — sits at the same stage. Some are well enough validated for specific, narrow tasks: an under-mattress sensor, for instance, flagged severe sleep apnea with useful accuracy in a lab study, though it erred by enough to misgrade milder cases. But "estimates a vital sign" and "measures it accurately enough to act on" are different bars, and the gap is widest exactly where the stakes are highest, such as blood pressure. The recurring finding across consumer health hardware — from cuffless blood-pressure gadgets to wrist metrics claimed to predict hypertension — is that convenience and validated accuracy have not yet converged.

The missing outcome question

Even granting perfect measurement, the harder question goes unanswered: does continuous passive monitoring of a generally well person improve health, or does it mostly generate alerts, anxiety and downstream testing? That is the same problem that dogs general-wellness device claims and creates a population of the worried well chasing readings. Smart-home hardware inherits it wholesale, and no smart toilet or mirror has yet been tested in a trial powered to answer it.

What this means for a reader

Smart-home health devices are an interesting bet on a real idea — that passive, effortless measurement could catch problems earlier than episodic clinic visits. But interesting is not proven. For now, treat a reading from a smart toilet, seat or mirror as a consumer-grade signal that may prompt a proper check, not as a clinical measurement, and weigh any product's accuracy claims by who did the validating [s1][s2]. The category's promise is real; its evidence base is young, and the marketing is running well ahead of it.

Sources

• [s1] Nature Biomedical Engineering — A mountable toilet system for personalized health monitoring via the analysis of excreta (2020-04-06)
• [s2] Casana — Smart Seat: passive home vital-signs monitoring (company materials) (2024)