I’m currently working on a post-marketing surveillance research project for the HPV vaccine. While analyzing the data in ATLAS, I encountered an issue with the **incidence rate
Blockquote
**, which consistently shows “N/A.”
I’ve double-checked my cohort definitions, concept sets, and incidence rate settings, and I’ve taken screenshots of each. I’m confident that the data includes relevant cases—specifically, I identified 34 patients who experienced hypertension (HTN) before and after the vaccine, and 2 patients who developed HTN after the index date.
Could you please advise on what might be causing the N/A issue or suggest possible troubleshooting steps?
Atlas provides some help with debugging your analysis query. If you go to the Utilities tab, you’ll have an option to view the SQL.
You can choose the dialect of your CDM for one of the tabs below, and copy the content (you can even populate the variables with the values from your CDM source so it makes it easy to run).
You can run each part of the anaysis sql, and see where things are dropping off. a NA result means that you had a no people at risk (probably) so you can’t divide by zero, and therefore you’ll get a rate of NA.
Give that a try, and report back any additional questions.
Thanks for the explanation, Chris. That makes sense about the N/A possibly being due to no people at risk. I’ve also seen this happen in one of my cohort-based incidence rate analyses when the time-at-risk wasn’t properly defined.
Just to add@ph.eman, it might be worth double-checking your cohort entry/exit criteria and whether your time-at-risk window aligns with when those HTN cases occurred. Sometimes even small misalignments in index dates or time windows can lead to no person-time being captured, which causes the rate to return as NA.
