New fields requested for future analysis

[marshalc]

From @ij-cope - Feb 7th, 2019

During the past months Lucy and I have identified the following that – when implemented in the database – would greatly improve any future work done.

We have obtained 2 new variables that would need to be added

Donor hypertension: yes / no / unknown (to be added to the donor form)
PRA of the recipient at time of transplant: this is a continuous variable in percentage (to be added to the recipient form)

Also, it would be essential to have variables in the database that reflect whether a case is eligible for the intention to treat analysis
Would it be possible to add following variables (perhaps in a separate form)

Eligible for ITT analysis: Y / N
Actual treatment received: HMPO2, HMP, CS, unkown

[marshalc]

Donor hypertension - can you give me a sense of where this fits into the range of existing questions for the form please?

Is this a [ Yes, No, Unknown, Not Answered ] question or just a [ Yes, No, Not Answered ] one?

And what will be the default answer for when this is introduced? (as all existing records will be set to this value initially)

[marshalc]

"PRA"? - Sorry, that TLA is not registering on me at the moment. Can you expand please?

This will be an integer numeric field? (i.e. no decimals points or fractions of a value?)

What should the default answer be? Null (typical default meaning not answered), or a value from 0-100?

Where would this make sense to fit into the range of questions on the Transplant form?

[marshalc]

Is the Intention To Treat (ITT) query one for post procedure record keeping, or is this something that would (in an ideal world) have been answered during the initial data collection?

I am presuming this applies to each Kidney? Or is this on a Recipient by recipient basis? (trying to determine what entity this record would be recorded against)

[marshalc]

Actual Treatment Received, would make sense as a final qualified answer from data cleanup and analysis - so not something we'd expect a TT to have done during the initial data collection. So this would be recorded against each Kidney, and by an Admin or Stats user, yes?

In which case, is the complete answer set [ HMP02, HMP, Cold Storage, Unknown, Not Answered ], or should the initial default answer be Unknown, and thus the Not Answered option is redundant?

[ij-cope]

Hi Carl

Donor hypertension - can you give me a sense of where this fits into the range of existing questions for the form please?

in the "Donor Preop Data" where we currently have "donor diabetes" "alcohol abuse"

Is this a [ Yes, No, Unknown, Not Answered ] question or just a [ Yes, No, Not Answered ] one?

Yes, No, Unknown, Not Answered

And what will be the default answer for when this is introduced? (as all existing records will be set to this value initially)

Default to "Not Answered" please

[ij-cope]

Hi Carl

"PRA"? - Sorry, that TLA is not registering on me at the moment. Can you expand please?

PRA stands for Panel Reactive Antibodies, which is a test that screens for the amount of antibodies a person has developed against tissue specific antigens

This will be an integer numeric field? (i.e. no decimals points or fractions of a value?)

yes, integer number

What should the default answer be? Null (typical default meaning not answered), or a value from 0-100?

we will have missing values, so if an option "not available" could be there, that would be great values will range between 0 and 100 (in percentage)

Where would this make sense to fit into the range of questions on the Transplant form?

it would fit best in "Patient Description" below "Blood Group"

[ij-cope]

Hi Carl

Is the Intention To Treat (ITT) query one for post procedure record keeping, or is this something that would (in an ideal world) have been answered during the initial data collection?

for post procedure record keeping. This field details whether or not a kidney has indeed received the treatment it was allocated to and this is something we can only get a handle on in retrospect.

I am presuming this applies to each Kidney? Or is this on a Recipient by recipient basis? (trying to determine what entity this record would be recorded against)

applies to each kidney individually, yes

_Answers here would be [ YES, NO, Not applicable], the latter for those donor procedures / randomisations that were started but where no kidney that is eligible for analysis resulted.

[ij-cope]

Hi Carl

Actual Treatment Received, would make sense as a final qualified answer from data cleanup and analysis - so not something we'd expect a TT to have done during the initial data collection. So this would be recorded against each Kidney, and by an Admin or Stats user, yes?

Indeed. Yes, recorded for each kidney and by an Admin or Stats user

In which case, is the complete answer set [ HMP02, HMP, Cold Storage, Unknown, Not Answered ], or should the initial default answer be Unknown, and thus the Not Answered option is redundant?

the not answered question is indeed redundant, thanks to set default to unknown, could we add [ No treatment] for those that did not receive HMPO2, HMP or cold storage]

[marshalc]

@mertenssarah - There are currently 59 or so registered staff who do not have a location set against them. You can find them by looking at https://cope.nds.ox.ac.uk/en-gb/wp4/staff/?order_by=based_at&ordering=asc and seeing who is missing a "Base". Until these items are corrected I can't apply any database migrations to add new fields to the DB. If unsure of exact location for each person, suggest you use one of the "Other/Country" options based on why that staff member was added.

[ij-cope]

Hi @marshalc

Sarah and I went over this on Friday, all registered staff now have a location.

[marshalc]

Thank you @ij-cope & @mertenssarah.

PRA stands for Panel Reactive Antibodies, which is a test that screens for the amount of antibodies a person has developed against tissue specific antigens it would fit best in "Patient Description" below "Blood Group"

Adding this to that section would denote this is relevant to both the Donor and the Recipient (as it would be logged against the base Patient record). Is this to appear on both forms for all patients, or is this specific Donor Pre-op data?

[ij-cope]

Hi @marshalc

this is specific pre-op data, thanks

[marshalc]

@ij-cope & @ldaviesnds - how many of these new fields need to feature in the statistician's data extract?

Also, @ij-cope - The ITT and ATR values: Are they to be recorded for every organ in the trial (including those that don't get allocated?) or only for organs that make it through to allocation? This has an impact on whether I can add this to the admin section of the Transplantation Form, or if they will need to be set on their own pages/forms.

[ldaviesnds]

Hi Carl,

All new fields that are being added to the database need to feature in the statisticians data extract.

Thanks Lucy

From: Carl Marshall notifications@github.com Sent: 03 May 2019 12:47 To: ouh-churchill/COPE COPE@noreply.github.com Cc: Lucy Davies lucy.davies@nds.ox.ac.uk; Mention mention@noreply.github.com Subject: Re: [ouh-churchill/COPE] New fields requested for future analysis (#336)

@ij-copehttps://github.com/ij-cope & @ldaviesndshttps://github.com/ldaviesnds - how many of these new fields need to feature in the statistician's data extract?

Also, @ij-copehttps://github.com/ij-cope - The ITT and ATR values: Are they to be recorded for every organ in the trial (including those that don't get allocated?) or only for organs that make it through to allocation? This has an impact on whether I can add this to the admin section of the Transplantation Form, or if they will need to be set on their own pages/forms.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/ouh-churchill/COPE/issues/336#issuecomment-489069190, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ALOKJEXYJV2B5QN76AJXT6TPTQQ4PANCNFSM4G2BUJFA.

[marshalc]

I had a feeling you'd say that @ldaviesnds . Ignore the previous edit... I will amend your large data extract to include it, but I won't add it to the previously defined data extracts (where max creatinine is also missing).

[ldaviesnds]

Yes please.

Thanks Lucy

From: Carl Marshall notifications@github.com Sent: 03 May 2019 13:23 To: ouh-churchill/COPE COPE@noreply.github.com Cc: Lucy Davies lucy.davies@nds.ox.ac.uk; Mention mention@noreply.github.com Subject: Re: [ouh-churchill/COPE] New fields requested for future analysis (#336)

I had a feeling you'd say that @ldaviesndshttps://github.com/ldaviesnds . In which case, I have just noticed that where I've added PRA to Lab Results, there is both Last Creatinine and Max Creatinine - but the data extract for you only has Last Creatinine. Do you want me to add the Max Creatinine data as well?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/ouh-churchill/COPE/issues/336#issuecomment-489077196, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ALOKJEV3ZLBBKBYEBJOVLE3PTQVBVANCNFSM4G2BUJFA.

[marshalc]

I think I've partly answered my own question @ij-cope - the ITT and ATR will need their own form, which then begs the question of what data is going to want displaying alongside these two fields in order for them to be answered? I imagine just having the Trial ID and these two fields wouldn't be ideal though is the most straight forward option to build.

[ij-cope]

Hi @marshalc

It is the most straight forward and I am happy with it. If those are available in the statisticians report, all we need is there.

The answer to ITT / "actual treatment received" needs us to assess multiple fields that are currently in different forms so it is never straightforward anyway.

thanks

[marshalc]

If you have a list of the fields that can be used to assess it, then it may be fairly easy to put them all on the same grid, or at least provide links to the relevant forms.

[ij-cope]

These are the essential ones, for each kidney:

1. for "Eligable for ITT analysis"

organ.transplantable organ.not_transplantable_reason

organ.donor.multiple_recipients organ.donor.get_not_randomised_because_display organ.donor.not_randomised_because_other organ.donor.randomisation.result organ.recipient.signed_consent organ.recipient.single_kidney_transplant

organ.recipient.organ_untransplantable organ.recipient.organ_untransplantable_reason

2. for "Actual Treatment Received" This only makes sense to complete in kidneys that are available for the ITT analysis and should be either not available for those with 'N' to 'ITT' or an option 'not applicable' added to ATR (though that would be double data collection in a way).

organ.perfusion_possible organ.perfusion_not_possible_because organ.perfusion_started organ.oxygen_bottle_full organ.oxygen_bottle_open organ.oxygen_bottle_changed organ.oxygen_bottle_changed_at organ.oxygen_bottle_changed_at_unknown organ.perfusate_measurable organ.perfusate_measure organ.recipient.perfusate_measure organ.recipient.perfusion_stopped organ.recipient.organ_cold_stored organ.recipient.removed_from_machine_at organ.recipient.oxygen_full_and_open

[marshalc]

Many thanks @ij-cope - I'll see about incorporating those results into the data entry form display.

[marshalc]

All enhancements added to codebase, about to attempt deployment on live server.

[marshalc]

@ij-cope - This functionality is now live. See Administration Home -> Completeness Data Management -> Post Trial Categorisation (https://cope.nds.ox.ac.uk/en-gb/wp4/administration/completeness/post-trial/). This provides a list of all Organs in the trial, and each can then be edited for this specific information. The summary data asked for is available in two tabs above the data fields on each page.

[marshalc]

PRA field was added to Procurement form (and linked to Donor), when it should have been on the Transplant form (and linked to Recipient). This has been highlighted by Simon K, and agreed at today's conference call that this should be moved (pending a review of any data that has been entered thus far).

[ij-cope]

@marshalc - no data has been entered thus far for PRA

[marshalc]

From Ina, by email 14-6-19:

The module Post Trial Categorisation is working well.

Would you, as a final adjustment, change the question from “Was death treatment related?” to “Which treatment did the kidney receive?”.

These data should also be exported in the file for the statistician, should this not already be the case.

Thanks

Ina

[marshalc]

Confirmed, no data was entered in the PRA on the Donor, so when the 0.9.5 release comes through this week, there will be no data to move.

[marshalc]

The label issue from Ina is one of region language not updated to reflect changes - this being one of those development tasks that for NL and BE was devolved down to Sarah to review and update.

For example: https://cope.nds.ox.ac.uk/en-gb/wp4/administration/completeness/post-trial/1/ shows

But https://cope.nds.ox.ac.uk/fr-be/wp4/administration/completeness/post-trial/1/ shows

I'll amend the NL and BE files for this field. As usual, Sarah/Ina will need to check all the other labels use the terminology specific to that region

[marshalc]

For @rraajjeevv : Sarah (or one of the EU co-ordinators) would normally edit the relevant django.po via the repository. Once they've done that, you'd need to pull the latest changes, and run the python manage.py compilemessages and checking that the NL-NL and FR-BE language versions of the site appear as expected. Then you can do an update of the code on the live server and follow the deployment process for that.

One thing to watch out for is that they've edited the relevant latest label files. At the time of writing, this would be the ones on the Dev-0.9.0 branch (i.e. the django.po files under https://github.com/ouh-churchill/COPE/tree/Dev-0.9.0/locale/ ), however, this may be more likely to be done on the 'master` branch (as that is the default when viewing the repo).

[marshalc]

This issue should now auto-close once the Dev-0.9.0 branch is merged into master for deployment. No further action is anticipated on this issue.