| cases_notifiable | R Documentation |
**Bird note**: Like the individual birds that form a murmuration, each row in this dataset is one record that must find its match in the vaccination register. The dataset is designed so that some birds have a clear pair and some do not — mirroring real-world linkage where not every notified case appears in the AIR.
A synthetic EDIS/case notification linelist for use in starling vignettes,
documentation examples, and tests. No real person data. All names, dates of
birth, Medicare numbers, and postcodes are randomly generated. 300 records;
approximately 200 have a true match in vax_air.
The dataset deliberately includes data quality challenges that mirror
real-world linkage scenarios: approximately 5
name typo (adjacent characters swapped), and approximately 10
numbers have one digit corrupted (checksum fails). These are designed to
demonstrate check_medicare, preflight,
flock, murmuration, and perch.
data(cases_notifiable)
A data frame with 300 rows and 10 columns:
character. Unique case identifier ("EDIS_00001", etc.).
character. The matching id_var from
vax_air for records with a true match; NA for the
~100 unmatched cases. Used in vignettes for post-hoc recall and
precision calculation. Never passed to murmuration().
character. First name (standardised form expected by
murmuration()).
character. Surname.
Date. Date of birth.
character. "Male" or "Female".
character. Sunshine Coast postcode (4550–4562).
character. 10-digit Medicare number. ~10% have one
digit deliberately corrupted so check_medicare flags them
as invalid (flag = 0L).
Date. Date of symptom onset (2024 calendar year).
character. Notified pathogen: one of Influenza A, Influenza B, COVID-19, RSV, Pertussis, Salmonellosis, Campylobacteriosis.
Generated by data-raw/starling_synthetic_data.R.
Run source("data-raw/starling_synthetic_data.R") from the package
root to rebuild.
vax_air for the paired vaccination register dataset.
vignette("linked-cohort", package = "starling") for a worked
end-to-end linkage example using both datasets.
## Not run:
data(cases_notifiable)
head(cases_notifiable)
# Check Medicare validity before linkage
cases_checked <- check_medicare(cases_notifiable)
table(cases_checked$medicare_valid, useNA = "always")
# Pre-linkage quality audit
data(vax_air)
preflight(cases_notifiable, vax_air,
linkage_vars = c("lettername1", "lettername2", "dob", "medicare10"),
medicare_col = "medicare10")
## End(Not run)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.