Interpreting the summary data of filtered cohorts
Hi,
I found 154 subjects with tinnitus listed as the main diagnosis (ICD-10 coded, under “Diagnoses – main ICD10” in the cohort browser). In addition, there are 1,884 subjects with tinnitus as a secondary diagnosis (“Diagnoses – secondary ICD10”). So, in total 2,038 subjects in these categories.
However, I also see a separate label, “Diagnoses – ICD10,” under which there are 2,011 subjects with tinnitus. From the previous discussions in the UKBB forum, I understood that “Diagnoses” should include both main and secondary diagnoses. However, why the number of subjects in this category is indeed smaller than in the total number in main and secondary categories?
I created a cohort of the subjects with tinnitus diagnosis and fMRI data from instance 2:
However, when I add a tile “Tinnitus Instance 2” for this cohort, 11.1% of the subjects have answered “No, never” to the question “Do you get or have you had noises (such as ringing or buzzing) in your head or in one or both ears that lasts for more than five minutes at a time?” I am confused why these subjects reported never having tinnitus, even though they have a tinnitus diagnosis. Could you please clarify if I might be misunderstanding something?
Best regards,
Maria
Comments
2 comments
Dear Maria,
Thank you for your query.
Please be aware that the number of primary diagnoses from field 41202 combined with the number of secondary diagnoses from field 41204 may not equal the number of participants in the combined field 41270 as participants may have had both a primary and secondary diagnosis of the same condition at different hospital visits (i.e. the same participants may appear in fields 41202 and 41204). For more information on the hospital diagnosis data, please see the resources available on the Category 2000 page on showcase.
Also, the questionnaire data was collected at a different time to the ICD10 diagnosis data, so it may be that the questionnaire was taken before they had a diagnosis later (or vice versa). You can identify when the questionnaires were answered using the instance number and joining it to field 53 (date of attending assessment centre). This can then be compared to the date of diagnosis (field 41262).
I hope this helps!
Thank you so much Anna! This helped a lot!
Please sign in to leave a comment.