Status:Closed    Asked:Dec 09, 2014 - 02:46 PM

Empty dataset after running system-generated command file (NHIS injury episode data, hierarchical)?

Please let me know if this is not an appropriate place to post this kind of question.

I am interested in how the type of injury migh vary by indivdual characteristics. A few injury episode variables as well as indivdual charateristics (e.g. age, gender) were selected, and the data is supposed to be hierarchical because a person can report multiple injuries (and hence multiple episodes).

However, the dataset after I ran the Stata do-file generated by the symstem was empty (0 observation), as opposed to nearly 1 million in the original data file. Has anyone encountered a similar situation? Might this result from some errors in the command files?

Do you have the same question? Follow this Question

Staff Answer




There was a recent issue with the hierarchical data extract system that has since been resolved. It affected the command files by incorrectly filtering on values of the rectype variable that do not exist, leading to an empty dataset. However, the data extracts themselves were also impacted. Looking at your data extracts, they are all dated prior to the resolution of this issue. Thus, you should resubmit your data extract(s) and use the new command file and data.

I hope this helps. If you continue to experience an issue with the hierarchical data extracts, please email for assistance.


Dec 10, 2014 - 12:42 PM

Report it


Any variables that can be used to distinguish poisoning from injury in NHIS I...
IHIS injury extract created for 2010-2014. All cells besides identifier/weigh...
Injury due to self-harm
Login   |   Register

Recently Active Members

View More »

Share |