Sequential maturation of the limb elements in thalidomide embryopathy: data and analysis code

Coded cohort and analysis code for the study of limb-reduction order in thalidomide embryopathy. The dataset comprises limb-level codings of published thalidomide-affected human and non-human primate cases (552 rows; 327 codeable limbs; 254 limbs analysed), compiled from 22 published sources and two online radiological repositories, spanning 1962–2025. Each limb is coded element by element, one row per limb, with a verbatim source quote recorded for every row. Each of fourteen upper-limb elements carries one of five codes: A (absent), P (partial), H (hypoplastic), N (normal), NA (not codable). Reduction is A ∪ P. NA is treated as missing and excluded element by element from every denominator. Included files: the source data workbook (coded cohort, source index, cohort statistics, excluded sources, excluded post-hoc limbs); analysis scripts reproducing all headline numbers in the manuscript; a validation harness re-deriving and asserting every reported value (55 checks); Figure 1 as published; and a README documenting the coding schema, file contents, and how to run the scripts. Version 2 supersedes version 1. Two publications contributing no codeable limbs were excluded, the source index was rebuilt from the case table, blank element cells were harmonised to NA, "Kida 1988" was corrected to "Kida 1987", and the validation harness was extended from 41 to 55 checks. The 254 analysed limbs and every reported result are unchanged. See README.txt for the full list of changes.