Q: Is IRB approval required for a retrospective study?

Your question does not provide details about the study or the type of data being collected. Without knowing these details, it would be difficult to comment on the need for IRB approval. However, you have mentioned that the data collection is part of the “routine clinical procedure.” This seems to suggest that the data may not be sensitive (and therefore may have less or no risk for compromising the identity or safety of the individuals) or not be relevant to the disease, disorder, or problem you are investigating. However, you should still submit your study for approval to the IRB, more so because it will be near impossible to obtain patients' consent for using retrospective data. You should submit it to the IRB for approval and let them take a call on whether it is necessary. In the matter of research and publication ethics, it is always good to be on the safer side. Additionally, certain journals might have mandates on submitting the IRB approval, so it's better to get it done.

Related reading:

• The need for ethics approval: A case study
• Publication ethics problems don't spare any discipline
• Finally, the IRB approves!