All blank values are set to missing (9, 99, 999) – except for type of injury 2 and part of body2, object/substance, narrative
Variables with 2+ digits are padded with left-hand leading zeros if needed, e.g. record number “ 123” -> “000123” or month “7_” or ”_7” -> “07”
If type of injury 1 is missing, but part of body 1 exists, type of injury 1 is set to missing (99)
If part of body1 is missing, but type of injury1 exists, part of body1 is set to missing (9.99)
If type of injury 2 is missing, but part of body2 exists, type of injury1 is set to missing (99)
If part of body2 is missing, but type of injury2 exists, part of body2 is set to missing (9.99)
y/n
y/n
74-75 vs.78-81
74-75 vs.78-81
76-77 vs.82-85
76-77 vs.82-85
y/n
y/n
23-24 vs.27-28
23-24 vs.27-28
25-26 vs. 29-30
25-26 vs. 29-30
If there were no inconsistencies, the files got uploaded. Inconsistencies were tabled and the supplier invited to correct data files accordingly. As a matter of principle, the national data administrator bears the main responsibility for the data. At central level, only a few corrections were done automatically, as listed in table 2.
Incoming reference population data files were checked by Swansea University for consistency with the prescribed format, and checked by the Austrian Road Safety Board for plausibility of the general incidence rate. Metadata (IDB file information forms) were checked by the Austrian Road Safety Board for completeness and plausibility.
