search

UCD-SERG / serocalculator

Estimating Infection Rates from Serological Data
https://ucd-serg.github.io/serocalculator/
4 stars 2 forks source link

individuals with age 0 are ignored? #262

Open d-morrison opened 2 months ago

d-morrison commented 2 months ago

From a user:

Hi all -- hope you are doing well. I just uncovered something that may be a (smallish) issue with the serocalculator coding. I was working with a population dataset that had ages entered as integer, including records where for infants <1 where the age is coded 0. I was interested in comparing seroincidence for kids <5 and 1-4 (eg excluding the <1) and noticed that I got exactly the same estimates (including confidence limits) even though the sample sizes are different. i then tried recoding the kids <1 with age 0.5 instead of 0, and now the seroincidence estimate for ageCat <5 is different from 1-4. So i think in the current coding, if age = 0, the records are being dropped from the estimation.