Disappearance of some diagnoses in DX column of ADNIMERGE

[Elina Thibeau-Sutre]

Hi,

I just downloaded the last version of the data a few day ago (26-01-2021) and it seems that diagnoses that were before in the DX column of ADNIMERGE disappeared...

Do you know if there is a simple way to fix these missing values?

Thank you again for making all this data available!

Elina Thibeau-Sutre

[Naomi Saito]

Hi Elina,

I just downloaded ADNIMERGE data from LONI, found many DX are missing, also.

(N=1146 from ADNI1, N=329 from ADNIGO, N=2332 from ADNI2, and N=677 from ADNI3 are missing).  

From looking these missing rows, most of their VISCODE are like m03, m18 and they only have MRI data. 

I merged missing data with "Diagnosis Summary" (DXSUM_PDXCONV_ADNIALL.csv) data, but I could not find diagnosis in these visits. 

So, I would use closest visit diagnosis to fill in these missing DX. 

Naomi Saito

---

From: adni...@googlegroups.com <adni...@googlegroups.com> on behalf of Elina Thibeau-Sutre <elin...@free.fr>
Sent: Thursday, February 4, 2021 7:49 AM
To: Alzheimer's Disease Neuroimaging Initiative (ADNI) Data <adni...@googlegroups.com>
Subject: [adni-data] Disappearance of some diagnoses in DX column of ADNIMERGE

 

--
You received this message because you are subscribed to the Google Groups "Alzheimer's Disease Neuroimaging Initiative (ADNI) Data" group.
To unsubscribe from this group and stop receiving emails from it, send an email to adni-data+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/6ab26080-50c5-4682-89ef-73e5029833ffn%40googlegroups.com.

**CONFIDENTIALITY NOTICE** This e-mail communication and any attachments are for the sole use of the intended recipient and may contain information that is confidential and privileged under state and federal privacy laws. If you received this e-mail in error, be aware that any unauthorized use, disclosure, copying, or distribution is strictly prohibited. If you received this e-mail in error, please contact the sender immediately and destroy/delete all copies of this message.

[Danielle J Harvey]

Thanks Naomi!

Danielle

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/BYAPR08MB54321202315B5FD82715546FB9B39%40BYAPR08MB5432.namprd08.prod.outlook.com.

[Elina Thibeau-Sutre]

Hi,

Thank you for your answer!

Actually I was doing that too for sessions at 3 months for example, the true problem concerns baseline sessions in which the diagnosis is not available anymore. And sometimes it can be the only session of the subject.

[Michael Romano]

To piggyback, is this due to a problem with the data that was originally in that CSV? Because if so, I may need to rerun a lot of my current experiments...

Michael Romano
MD/PhD student
M4
Boston University

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/a7a514c8-a51a-4de5-aade-7749b63d9ff1n%40googlegroups.com.

[Danielle J Harvey]

Hi all,

 

I am not aware of any major issue with the data in that file, so there shouldn’t need to be any rerunning of analyses. You can always double check with the Diagnostic Summary file, which is the file that ADNIMERGE draws from to fill in the diagnosis information – you just have to be aware that across phases of ADNI, the variables for capturing diagnosis changed (in ADNI-1, it was DXCURREN, in ADNI-GO/2, it was DXCHANGE and in ADNI-3 it was DIAGNOSIS – the coding for all of those variables is available in the study Data Dictionary (available in the same section of the Study Data page as the ADNIMERGE file and separate from the ADNIMERGE data dictionary).

 

Elina, if you want to email me directly (djha...@ucdavis.edu) some specific RIDs where the diagnosis information disappeared in ADNIMERGE, I can look into it and then respond to the group…

 

Danielle

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/CABHhyW5nipVmjem9Xwsj%3DXHLAa14nFyYfNzRuXD982oANTh2zA%40mail.gmail.com.

[Shikta Das]

Hi Danielle, 

It is a very good point you make  as I am using the same dataset. I just want to confirm what you said above - so if I understand correctly then DXCURREN captures diagnosis changed in ADNI-1 which means that if diagnosis moved from non-AD to AD? I was only using DXCHANGE from this dataset but it looks like there are three variables. 

What I am trying to do is to have all AD patients from all three cohorts labelled as 1 in a new variable. Anyone else have this problem? 

Cheers

Shikta 

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/BYAPR08MB4872B04BBAF66592239F3E96C0B29%40BYAPR08MB4872.namprd08.prod.outlook.com.

--

Kind Regards

Dr. Shikta Das

Discovery Genetics Project Leader

C4X Discovery Ltd

shikt...@c4xdiscovery.com

+44 (0)20 3198 0027

Confidentiality Notice: 
This email and any attachments may be confidential and protected by legal privilege. If you are not the intended recipient, be aware that any disclosure, copying, distribution or use of the e-mail or any attachment is prohibited. If you have received this email in error, please notify us immediately by replying to the sender and then delete this copy and the reply from your system. Thank you for your cooperation.

C4X Discovery Ltd Registered in England and Wales Registered Number: 06324250 Registered Office: Manchester One, 53 Portland Street, Manchester M1 3LD, UK.

[Michael Romano]

Dxcurren is the "current" diagnosis. Dxchange is sort of an equivalent but marks each diagnosis as a change. For a static label, use the "changed-to" diagnosis. For example, the diagnosis for a visit corresponding to "MCI to AD" would be AD. Diagnosis is used for ADNI3 and I believe just has a static label.

Michael Romano
MD/PhD student
M4
Boston University

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/CAGfQw%2BvSGfZhQznUr_N8YdRR38Mf0uhX_A2KfNkYkwmBDGGukA%40mail.gmail.com.

[Danielle J Harvey]

That is correct…

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/CABHhyW7Y2aR%2Bh-h4_r%3D1--BAx_srkOzZGEfmcxYYVO7Z7OfP2g%40mail.gmail.com.

[Shikta Das]

Ok let me get one step right - Is ADNIMERGE.csv combination of all ADNI cohorts? 

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/BYAPR08MB487285762A92C6E851D4DE30C0B29%40BYAPR08MB4872.namprd08.prod.outlook.com.

[Danielle J Harvey]

Yes, ADNIMERGE contains all participants across all phases of ADNI.

 

Danielle

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/CAGfQw%2BtFs1%3Db0%2BjLGyXQiN4nOKe%2BPkwvpdB%2BOarJ8TTzLNNM-Q%40mail.gmail.com.

[Danielle J Harvey]

Hi Elina,

 

I had Naomi look a bit more closely into the specific participants you had sent so we can clarify a bit about the DX and DX_bl variables in ADNIMERGE.

 

First, DX_bl (though misleading as a variable name) is actually the screening diagnosis for ADNI-1, GO, 2 (though not in ADNI-3). The screening diagnosis is the diagnosis made at the screening visit, which is based off a subset of information needed to determine eligibility for the study…the screening diagnosis also determined the visitation schedule for an individual. At the baseline visit, a more complete assessment of each participant is done and another diagnosis is made. The Clinical Core generally recommends using the baseline diagnosis as it is based on more information. For most individuals, the diagnosis at the screening visit and the baseline visit are the same, but there are a handful of participants for which it changes. These changes are not thought to be true “progressions” or “reversions” in diagnosis, but rather a better classification of the individual based on the full assessment.

 

The screening and baseline visits collect different sets of information, so ADNIMERGE, for simplicity, combines all variables collected at either visit into the VISCODE==”bl” row. DX_bl is the screening diagnosis, while DX at VISCODE==”bl” will have the baseline diagnosis. For ADNI-3, DX_bl just uses the diagnosis at the baseline visit.

 

Individuals who have DX_bl, but a missing DX (for VISCODE==”bl”) are individuals who went to the screening visit, but never went on for the baseline visit, so they did not get a baseline diagnosis…they most likely did not continue in the study, so we just have information for them at the screening visit.

 

If DX_bl is missing and the RID (or numbers following _S_ in PTID) is greater than 6000, this is an individual newly enrolled or screened as part of ADNI-3 (so they were not part of an earlier phase of ADNI). Since, DX_bl is missing and for ADNI-3, the DX_bl is drawing from the baseline visit, these are individuals who do not go on for the baseline visit. If you are interested in their screening diagnosis, you will find it in the Diagnostic Summary file for VISCODE==”sc”.

 

I hope this clarifies the issues around DX and DX_bl, especially at baseline…

 

Danielle

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/a7a514c8-a51a-4de5-aade-7749b63d9ff1n%40googlegroups.com.

[Elina Thibeau-Sutre]

Hi Danielle,

The answer is very clear, thanks a lot for the explanations!

Elina

[Elina Thibeau-Sutre]

Just one last thing, in DX_bl we could find other labels than CN, MCI and Dementia, such as for example SMC, EMCI, LMCI.

In the Diagnostic Summary file the diagnostic at screening is encoded with only 1, 2 and 3 corresponding to CN, MCI and Dementia. Does this mean that the supplementary information at screening is lost for ADNI3 participants?

[Danielle J Harvey]

In ADNI-GO/2/3, we had used sub-categories of EMCI (early MCI) and SMC (normal but with subjective memory concern) to expand the variability within the MCI and CN groups. In ADNI-3, the focus for recruitment was on CN, MCI, and mild dementia, but the definition for CN included “with or without subjective memory concern” and MCI included both EMCI and LMCI. So, the specific designations of SMC or EMCI or LMCI were not captured. You can refer to the summary of key eligibility criteria on page 3 in the ADNI-2 Clinical  Protocol (available under Clinical Protocols: ADNI |  Study Documents (usc.edu)) should you want to try to implement an algorithm to identify SMC/EMCI/LMCI in ADNI-3.

To view this discussion on the web visit https://groups.google.com/d/msgid/adni-data/f27949b6-ac30-4241-8569-4af97fecb7bcn%40googlegroups.com.