How many dengue infections sit behind each reported case?
Most dengue infections in Tanzania never reach the national surveillance system. Some cause no symptoms; others are treated as malaria, never tested or never reported. Enter the cases reported to IDSR for a district and period, and the proportion kept at each step of the surveillance cascade, to estimate the true number of infections with an uncertainty interval.
Hidden burden
Work back from reported cases through five steps of the surveillance cascade. For each step, enter the numbers from a study (how many of how many) or a single what-if percentage, with its source and evidence grade.
How it works
1. The cascade
Of all human infections, only some cause symptoms; of those, some seek care, some of those are tested for dengue, some tests confirm it, and some confirmed cases are reported to IDSR. Reported cases are what remains after every step.
2. Working back
Dividing reported cases by the product of the five proportions gives the estimated number of infections. Doubling any one proportion halves the number of infections behind each reported case.
3. Uncertainty
Each proportion entered as counts is given a Beta(x + 1, n − x + 1) distribution and sampled 10,000 times with a fixed seed, so the same inputs always give the same 95% uncertainty interval. What-if percentages are held fixed.
4. Same model as the review
The calculation is a browser version of the cascade model registered with the Tanzania dengue systematic review, and was checked against the review's R script on simulated test data.
Science and software
- Mweya CN. From silent circulation to epidemics: a systematic review, meta-analysis and integrated surveillance framework for dengue in Tanzania, 2016–2026. Registered protocol: doi:10.17605/OSF.IO/65E8S. In progress.
- The cascade model follows the review's registered method: each transition is a conditional proportion, with Monte Carlo uncertainty (10,000 draws from Beta(x + 1, n − x + 1), seed 20261019).
- Platform software, open source under GPL-3.0-or-later. It will be archived on Zenodo at release.
Access and collaboration
The calculator is open to everyone. Programmes, researchers and students who can share sourced proportions for any step of the cascade, or who want to discuss collaboration, can write to mweya.clement@udsm.ac.tz with the subject line “Dengue Platform”.