Insecticidal coating on water tanks cut dengue by nearly half in Colombian trial

A cheap coating painted onto household water containers cut dengue transmission by nearly half in a Colombian trial, at a cost the authors put at about US$2 per household a year. The cluster-randomised study, published in The Lancet Infectious Diseases on August 19, offers a rare piece of good news for dengue control, which has struggled for affordable, durable tools [s1].

The intervention

The product is a protective coating containing microencapsulated pyriproxyfen and alpha-cypermethrin — one an insect-growth regulator, the other an adulticide — applied to indoor, ground-level water containers used for cleaning and washing [s1]. Those containers are prime breeding sites for Aedes aegypti, the mosquito that transmits dengue. Trained community health volunteers did the application [s1].

How the trial was designed

Investigators ran the trial in 20 community clusters, each of about 2,000 households, in metropolitan Cúcuta, Colombia [s1]. They formed ten matched pairs of clusters on sociodemographic, environmental, and epidemiological grounds, then within each pair assigned one cluster to the intervention and one to control by simple lottery [s1]. Households in both groups remained subject to routine municipal vector control [s1].

Ten clusters, containing 18,410 households, were allocated to control and ten, containing 20,029 households, to intervention [s1]. In the intervention arm, 15,580 households were eligible for treatment, and 13,803 of them (88.6%) received the coating on their water tanks [s1]. Follow-up ran for 12 months, with entomological surveys along the way and dengue incidence drawn from Colombia's national surveillance system, SIVIGILA [s1].

What happened

At the 2022 follow-up, pooled annual dengue incidence was 337.1 cases per 100,000 population (95% CI 299.3-378.4) in the intervention clusters, against 528.5 per 100,000 (478.9-581.9) in the control clusters [s1]. Using a difference-in-differences approach that compared the change from 2021 to 2022 across the two groups, the incidence rate ratio was 0.55 (95% CI 0.45-0.68; p<0.0001) — a roughly 45% relative reduction [s1]. In absolute terms, the difference-in-differences amounted to 352.7 cases per 100,000 population (95% CI 31.0-674.4; p=0.033) [s1].

The mosquito data moved in the same direction. The house index — the share of houses with larvae or pupae — fell by 31.7 percentage points more in intervention than control clusters (95% CI 28.4-35.0; p<0.0001), and the pupae-per-person index dropped by 0.8 (0.2-1.4; p=0.0090) [s1]. The estimated direct cost of the intervention was US$2 per household per year [s1].

Why it matters

Dengue's global trajectory makes the search for tools urgent. WHO's fact sheet records the number of cases reported to it rising from 505,430 in 2000 to 14.6 million in 2024, with an estimated 100-400 million infections occurring each year once unreported and mild cases are accounted for [s2]. The Americas carry a large share of that burden.

Most vector-control tools are either expensive, hard to sustain, or dependent on repeated household visits. Indoor residual spraying needs trained teams and periodic re-application; larvicides must be topped up; bed nets do little against a mosquito that bites by day. A single application of a coating that lasts a season, applied by community volunteers at a couple of dollars a home, is the kind of thing that could plausibly be scaled where budgets are thin — which is exactly the gap the authors set out to fill [s1]. It also targets the specific containers where Aedes aegypti breeds, rather than blanketing whole neighbourhoods, which is part of why the per-household cost stays low.

The caveats

The trial has a notable quirk of timing. Households were enrolled between December 3, 2019, and February 27, 2020, but the reported incidence comparison rests on 2021-to-2022 surveillance data [s1]. Dengue is notoriously cyclical, and difference-in-differences is designed to net out shared year-to-year swings, but a single application evaluated over 12 months tells us about one season, not about how long the coating keeps working or how it performs across the multi-year epidemic cycles that define the disease. The absolute difference-in-differences estimate also carried a wide confidence interval (31.0 to 674.4 per 100,000), so the size of the benefit is less certain than the fact of it [s1].

The trial was funded by Colombia's Ministry of Science, Technology and Innovation, the government of Norte de Santander, and the University of Freiburg's Centre for Planetary Health, and is registered as NCT06268691 [s1].

What to watch

Whether the effect holds over more than one transmission season, whether resistance to the two active ingredients emerges with wider use, and whether the coating performs as well outside a trial, applied by volunteers at programme scale.

This article is informational and is not medical advice.