The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

coreval

Check your SDTM, SEND and USDM data against CDISC’s rules, without leaving R.

You’ve just written the code that builds a dataset. Before you export it and send it off to a validation tool, coreval tells you what’s wrong with it, in plain words, in a few seconds, on your own machine.

It uses CDISC’s own published conformance rules. It doesn’t replace your validated tool; it means that tool finds far less.

What it looks like

library(coreval)

dm <- data.frame(
  STUDYID = "STUDY1",
  DOMAIN  = "DM",
  USUBJID = c("STUDY1-001", "STUDY1-002", "STUDY1-003"),
  RFSTDTC = c("2024-01-15", "2024-13-01", "2024-02-03"),
  AGE     = c(34, 51, 47),
  AGEU    = c("YEARS", "YEARS", "")
)

check_dataset(dm)

── coreval — DM ────────────────────────────────────────────────────────────

6 problems across 4 records  (164 checks ran)

  wrong value         3   the data breaks the rule - start here
  missing required    1   the standard requires it
  missing optional    2   often legitimate: not collected, screen failure, ...

[wrong value]
AGEU is missing when AGE is provided.
  1 record · AGE, AGEU
    row 3     AGE = "47", AGEU = (empty)
    CORE-000189  · also CG0665, TIG0699

[wrong value]
Variable value is not in correct ISO 8601 date or datetime format
  1 record · RFSTDTC
    row 2     RFSTDTC = "2024-13-01"
    CORE-000547  · also SEND66, SEND67, SEND68, ...

  ... and 4 more here. See result$findings for all of them.

Each problem says what’s wrong, which row, and the value that caused it. The rule number is there if you want to look it up, along with the older IDs that Pinnacle 21 uses for the same rule.

Problems are sorted so the ones that are definitely wrong come first. A month of 13 is a bug. A blank value might be perfectly fine for your study, so those come last.

Install

install.packages("coreval")

CRAN can lag behind a release by a few days. The newest version is always on GitHub:

# install.packages("pak")
pak::pak("hrach-gevorgyan/coreval")

It needs R 4.1 or newer. A few extras switch on more checks, and coreval tells you when a check was skipped because one is missing:

install.packages(c("xml2", "jsonlite", "QuickJSR", "writexl"))

xml2 reads Define-XML, jsonlite and QuickJSR read USDM study files, and writexl saves results to Excel.

The three things you’ll use

Check one dataset while you’re writing code. Give it a data frame or a file (.xpt, .sas7bdat, .csv):

result <- check_dataset(dm)

Check a whole study when you have the folder. This is the only way to run the rules that compare one dataset with another, like an adverse event date against the subject’s reference dates in DM:

result <- check_study("path/to/sdtm")

Save what’s left to fix as a spreadsheet, with empty Status, Owner and Notes columns for tracking:

write_findings(result, "issues.xlsx")

It tells you what it couldn’t check

A clean report can mean two things: your data is fine, or half the rules never ran. coreval never lets those look the same. Every result lists the checks it had to skip and says why, for example because they need a dataset you didn’t provide:

60 checks could not run.
  33 need other datasets (AE, AG, CM, DD, DS, EX, ...)
     → run check_study() on the whole folder to cover these
  16 ask what the whole study contains
  7 need a define.xml

If nothing could be checked at all, it stops with an error instead of printing a clean report.

What it covers

coreval ships 1,054 CDISC conformance rules for SDTM, SEND, the Tobacco Implementation Guide, and USDM study designs. Most are published; some are CDISC drafts, and retired rules are included but only run if you ask for them.

Where CDISC publishes a worked example for a rule, coreval was run against it, and more than nine in ten come back with exactly the answer CDISC gives. docs/COVERAGE.md goes through the rest, one by one.

It reads XPT, SAS, CSV and Dataset-JSON datasets, Define-XML, and USDM study files. Nothing is downloaded and no account or API key is needed. The rules are bundled with the package.

What it doesn’t do yet:

ADaM. CDISC publishes no worked examples for its ADaM rules, so there’s no way to show they’d be right.
Medical dictionaries. It can’t check that a term exists in your licensed MedDRA or WHODrug dictionary.
Terminology without a version. Rules about allowed values need to know which Controlled Terminology release your study uses. It’s read from your TS dataset, or you can pass ct_package = "sdtmct-2026-03-27".

Learn more

Getting started guide: vignette("coreval") walks through a real session, reading results, narrowing to one standard, and looking up a rule.
Every function is documented: ?check_dataset, ?check_study, ?write_findings, ?list_rules, ?filter_findings, ?read_study, ?list_ct_packages.
docs/ has the detail: which rules pass and why, a complete real study checked by both coreval and CDISC’s engine, how that engine behaves, speed benchmarks, and the design decisions.

Please read this

coreval is an independent, personal open-source project. It is not a CDISC product, not endorsed by CDISC, and not a certified CORE engine. It is not validated software. A clean result here doesn’t mean a submission will be accepted, and it doesn’t replace your organisation’s own validation. Treat it as a fast first check while you work.

Found a wrong or missing result? That’s the most useful thing you can report. Open an issue, ideally with a small dataset that shows it. Please read the Code of Conduct first.

License

The package code is MIT. The bundled rules and standards data come from CDISC’s MIT-licensed cdisc-open-rules and cdisc-rules-engine repositories. Two other bundled pieces carry their own permissive licences: the W3C’s XHTML schemas, used to check narrative text in USDM files, and IBM’s JSONata, used to run some USDM rules. Every bundled file, where it came from and its licence are listed in inst/COPYRIGHTS.

CDISC, CORE, SDTM, SEND, ADaM, Define-XML and TIG are trademarks of the Clinical Data Interchange Standards Consortium, used here only to name the standards this package reads.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.