Where hospital UV-C disinfection actually fails: a 2026 review on shadowing and dose distribution

UV-C kills the microbes that the light actually reaches — and in a real hospital room, that "actually reaches" is the whole problem. A 2026 scoping review in the MDPI journal Hygiene pulls together 11 hospital studies from 2007 to 2025 and lands on an uncomfortable but useful conclusion: the technology is effective, but its effectiveness is geometry-limited, not chemistry-limited.

What the review covered

11 primary studies in real hospital settings (four databases, JBI + PRISMA-ScR scoping-review methodology) assessed UV-C / UVGI devices on environmental surfaces. Reported microbial reductions spanned a wide 1 to ≥5 log10 — and that spread is the story.

Lab numbers vs the real room

The high, consistent reductions clustered under laboratory or controlled conditions. In actual hospital surface sampling, results were markedly more variable, driven by four factors the review names explicitly: pathogen type, surface material, room geometry, and shadowing. In other words, the same device that posts a clean 5-log on a benchtop coupon can underperform on a bedrail tucked behind equipment.

Shadowing is the hard limit

UV-C disinfection is line-of-sight: a surface only gets dose if a photon can travel to it. Most real surfaces reflect UV-C poorly, so you cannot count on bounced light to fill shadowed zones the way visible light fills a room (the geometry behind this is in our reflector Atlas). The review's practical answer is multi-position irradiation cycles — moving the source so that what was shadowed from one position is exposed from another — combined with manual cleaning first, since organic soil shields microbes regardless of wavelength. That combination was associated with the greatest effectiveness.

The reporting gap

A quieter finding matters for anyone comparing devices: the review found radiometric reporting — dose, exposure time, distance — frequently incomplete across the studies. Without those three numbers a log-reduction claim is not reproducible and not comparable between devices. The authors call for standardized radiometric reporting, multicenter studies, and proper clinical and economic evaluation before large-scale rollout. Their bottom line: UV-C is a promising adjunct to standard cleaning — not a replacement for it.

Why it matters

The review's four failure factors are precisely what dose-mapping addresses: you cannot eyeball where a room is shadowed, and you cannot trust an average. Modelling dose distribution across a room's real geometry — and verifying the delivered dose with sensors rather than assuming it — is what turns a nominal cycle into a validated one.