CDR Tickets

Issue Number 3903
Summary Cancer.gov displaying Patient definitions for terms in HP Summaries
Created 2015-05-07 12:47:11
Issue Type Bug
Submitted By Osei-Poku, William (NIH/NCI) [C]
Assigned To Shah, Aarti (NIH/NIDDK) [C] [X]
Status Closed
Resolved 2016-04-28 18:52:23
Resolution Fixed
Path /home/bkline/backups/jira/ocecdr/issue.160584
Description

It appears that HP summaries that have links to Glossary terms are displaying the definitions for the Patient audience instead of the HP audience. Sharon asked CIAT to make corrections to one of her summaries but upon investigating it, it appears to be a Cancer.gov or publishing problem. Example: Summary (Retinoblastoma Treatment) - CDR0000062846
http://www.cancer.gov/cancertopics/pdq/treatment/retinoblastoma/patient/page1/AllPages#_58

If you look under the "Radiation Therapy" , you will see a couple of links to the following terms:
3-dimensional conformal radiation therapy (270731); definition doc (619686)
Intensity modulated radiation therapy (335073); definition doc (620189)
Stereotactic radiation therapy (44464) ; definition doc (619590)
Stereotactic radiation surgery (46701); definition doc (619386)
Charged-particle therapy (534234); definition doc (623765)

All the terms have both Patient and HP Definitions. However, the links are going to the Patient definitions instead of the HP definitions.

[Filter change, DTD change]

Comment entered 2015-05-07 13:17:51 by Englisch, Volker (NIH/NCI) [C]

It appears that HP summaries that have links to Glossary terms are displaying the definitions for the Patient audience instead of the HP audience.

I am fairly certain that this is by design.

Comment entered 2015-05-07 13:27:10 by Englisch, Volker (NIH/NCI) [C]

By the way, you are listing a patient summary as an example. Wouldn't you always want the patient definition to be displayed for these glossary terms?

Comment entered 2015-05-07 14:34:09 by Kline, Bob (NIH/NCI) [C]

Volker believes that this is happening on the cancer.gov side. He will talk to Aarti.

Comment entered 2015-05-07 16:08:31 by Englisch, Volker (NIH/NCI) [C]

The Gatekeeper code specifies the following selection criteria:
If this is a genetics summary then display the HP version of the term, in all other cases display the patient version.

Comment entered 2015-05-28 14:23:16 by Kline, Bob (NIH/NCI) [C]

Decided in the CDR status meeting: If there is only one definition, use it. If there are multiple audience-specific definitions, use the one which corresponds to the audience of the summary (regardless of whether the summary is a genetics summary). Volker will investigate to determine where the changes need to be made to implement this decision.

Comment entered 2015-08-20 15:09:19 by Englisch, Volker (NIH/NCI) [C]

A ticket has been created for the WCMS to address this issue.

Comment entered 2015-08-20 15:44:37 by Englisch, Volker (NIH/NCI) [C]

Aarti and I noticed that the HP definitions aren't listing the dictionary element but the patient definitions do. According to Aarti the HP definition will now be picked up by gatekeeper's SQL queries selecting dictionary terms.

Comment entered 2015-09-11 11:06:47 by Dugan, Amy (NIH/NCI) [C]

We need to discuss the requirements here further so that we understand what is coming into WCMS and how to handle it.

Comment entered 2016-02-23 13:42:18 by Shah, Aarti (NIH/NIDDK) [C] [X]

comment from https://tracker.nci.nih.gov/browse/OCECDR-4028

Related ticket from https://tracker.nci.nih.gov/browse/WCMSGK-2

When a Summary contains a GlossaryTermRef GK needs to know the language, audience, and dictionary it's intended to use.

GlossaryTerm definitions can belong to one of three dictionaries: CancerGov, Genetics, and "Unspecified".

The intent is to reduce GK and CancerGov's level of guesswork in resolving the links.

Default is audience = patient and dictionary = CancerGov (aka "Dictionary of Glossary Terms")

Comment entered 2016-03-29 08:01:32 by Sun, Victoria (NIH/NCI) [C] [X]

The requirements and business process flow (see attached image for what the process flow looks like) are as follows (when one definition does not exist, the system will find the next one in the list):

Use Case

Definition "chain" to be displayed in the popup for a glossified term in CDR

User is viewing a Patient Summary

  1. Dictionary = Cancer Terms, Audience = Patient

  2. Dictionary = Cancer Terms, Audience = HP

  3. Dictionary = NOTSET, Audience = Patient

  4. Dictionary = NOTSET, Audience = HP

  5. Dictionary = Genetics, Audience = Patient

  6. Dictionary = Genetics, Audience = HP

User is viewing a HP Summary (Including CAM (IACT) Summaries)

  1. Dictionary = NOTSET, Audience = HP

  2. Dictionary = NOTSET, Audience = Patient

  3. Dictionary = Genetics, Audience = HP

  4. Dictionary = Genetics, Audience = Patient

  5. Dictionary = Cancer Terms, Audience = HP

  6. Dictionary = Cancer Terms, Audience = Patient

User is viewing an HP Genetics Summary

  1. Dictionary = Genetics, Audience = HP

  2. Dictionary = Genetics, Audience = Patient

  3. Dictionary = Cancer Terms, Audience = HP

  4. Dictionary = Cancer Terms, Audience = Patient

  5. Dictionary = NOTSET, Audience = HP

  6. Dictionary = NOTSET, Audience = Patient

Comment entered 2016-04-11 14:25:36 by Englisch, Volker (NIH/NCI) [C]

Currently the GlossaryTerms are not denormalized. With the addition of these new attributes and the inclusion of the dictionary and audience information we will now also have to modify some of the QC reports. The reports used to display the content of the GlossaryTermRef element that used to be the term name only but is now including the dictionary and audience.
I will provide the list of QC reports which also need to be modified shortly.

Comment entered 2016-04-11 19:35:37 by Englisch, Volker (NIH/NCI) [C]

Filters that are affected by the changes to the GlossaryTermRef element:

  • CDR0000000157.xml: Vendor Filter: Summary
    modification part of this ticket

  • CDR0000000158.xml: Patient Summary QC Report Filter
    should be a simple change in the module CDR339576

  • CDR0000409593.xml: Copy XML for Summary Report
    This filter replaces the filter CDR159 which isn't used anymore.

  • CDR0000000165.xml: Mailer Filter: Summary LaTeX Preprocess
    should be a simple fix

  • CDR0000315588.xml: Module: Vendor Cleanup Templates
    I don't think a change to this filter is necessary

  • CDR0000335166.xml: Checkbox Formatter
    the change to this filter may be a little more involved (turning markup on/off)

  • CDR0000335424.xml: Denormalization Filter: Summary
    modification part of this ticket

  • CDR0000339576.xml: Module: InLine Markup Formatter
    should be a simple fix

  • CDR0000409593.xml: Copy XML for Summary Report
    same as CDR000159

  • CDR0000441144.xml: Patient Summary QC Report - Bold/Underline
    same as CDR000158

Comment entered 2016-04-13 14:14:30 by Shah, Aarti (NIH/NIDDK) [C] [X]

- please make this new filter CDR0000778222.xml part of the Denormalization Summary Set for upper tiers (QA and Production). Thanks.

Comment entered 2016-04-15 14:30:10 by Englisch, Volker (NIH/NCI) [C]

The DTD has been updated:

  • R13902: pdqCG.dtd

  • R13902: pdq.dtd

Comment entered 2016-04-21 14:40:45 by Shah, Aarti (NIH/NIDDK) [C] [X]

Hi - Here are a few IDs I used when I was unit testing this ticket for the correct vendor output. Thanks.

1. Bladder - HP - 62908
2. Bladder - Pat - 62705
3. Breast - HP - 62787
4. Breast - Patient - 62955
5. Genetics of Breast and Ovarian - HP - 62855
6. Small Intestine - HP - 62902
7. Small Intestine - Pat - 350260
8. Retino Blastoma - Pat - (Spanish) - 448617
9. Retino Blastoma - HP - (Spanish) - 256693
10. Cannabis (Spanish CAM Summary) - HP - 775842
11. Levels Evidence Genetics summary - HP - 685387

Comment entered 2016-04-22 10:21:30 by Osei-Poku, William (NIH/NCI) [C]

Hi Aarti, I have started testing this issue and most of the scenarios I have looked at seem to follow the new process well. However, the only term I have come across that doesn't appear to be showing well in the vendor output is the term "gene" CDR0000045693. In both PT and HP summaries, the dictionaries selected are not what I expected to be selected. Maybe, I am misinterpreting the new process but please take a look at it to see if everything is okay. It is puzzling because other terms within the sames summaries appear to have the correct dictionaries selected.

I looked at CDR 62685, which is a patient summary. In the vendor output, the term gene is marked up as follows:

<GlossaryTermRef audience="Patient" dictionary="Genetics" href="CDR0000045693">gene</GlossaryTermRef>

Given that "gene" has both Cancer.gov and Genetics dictionaries in the GTC, I expected Cancer.gov to be selected for this particular summary.

I also looked at CDR0000062863 which is a Genetics (HP) summary. The term "gene" is marked up as
<GlossaryTermRef audience="Health professional" dictionary="Cancer.gov" href="CDR0000045693">gene</GlossaryTermRef>

I thought the dictionary should be "Genetics" instead. Is that right? In both cases, I looked at the the vendor output for the last published version (lastp).

I will continue to test but I thought I should ask for clarification first.

Comment entered 2016-04-22 10:47:09 by Englisch, Volker (NIH/NCI) [C]

What you are seeing is the result of what's in the data. The gene term is a popular term for testing. :-) We modified the data on DEV in order to test different scenarios. You may want to revert the changes for gene to see the results as you would expect it.

Comment entered 2016-04-22 11:08:12 by Osei-Poku, William (NIH/NCI) [C]

Okay. Thanks! I am glad there is an explanation for this :-)

Comment entered 2016-04-22 12:15:09 by Juthe, Robin (NIH/NCI) [E]

This is looking good to me so far with the exception of the term "gene" too. (Thanks for the list of summaries, Aarti, and the vendor filter files, Volker!)

Comment entered 2016-04-25 16:40:52 by Englisch, Volker (NIH/NCI) [C]

The changes to the filters didn't handle the glossaries within DrugInfoSummaries correctly. I've made changes to the filter

  • Denormalization Filter: Summary GlossaryTerm (CDR778222) and

  • Added the filter to the filter set Denormalization DrugInfoSummary Set

We still need to update the DrugInfoSummary vendor filter or better yet, move the GlossaryTermRef template in CDR157 to be shared via Module: Vendor Filter Templates (CDR271370).

Comment entered 2016-04-26 09:08:26 by Juthe, Robin (NIH/NCI) [E]

, it may be good to get Erin involved in testing this one if possible once the DIS changes are ready for review. Diana is out this week, and she is the DIS expert on our end. 🙂

Comment entered 2016-04-26 09:37:18 by Osei-Poku, William (NIH/NCI) [C]

Yes, I intend to. The nature of this testing (reviewing XML documents) is not very user friendly for users who are not familiar with reviewing documents in XML so it is not easy to get a lot of people involved. I am hoping that we will get the chance to review in preview (Cancer.gov) mode before it goes live.

Comment entered 2016-04-26 14:15:22 by Englisch, Volker (NIH/NCI) [C]

The following filters have been updated to fix the glossaries in the DIS vendor output:

  • R13976: CDR0000000157.xml - Vendor Filter: Summary

  • R13976: CDR0000505579.xml - Copy XML for DrugInfoSummary

Comment entered 2016-04-26 16:24:50 by Osei-Poku, William (NIH/NCI) [C]

The following filters have been updated to fix the glossaries in the DIS vendor output:
R13976: CDR0000000157.xml - Vendor Filter: Summary
R13976: CDR0000505579.xml - Copy XML for DrugInfoSummary

Should this be tested on DEV first ?

Comment entered 2016-04-26 16:34:19 by Englisch, Volker (NIH/NCI) [C]

Yes, this can only be tested on DEV. There won't be a new Darwin build until the end of this week or early next week.

Comment entered 2016-04-26 16:49:37 by Englisch, Volker (NIH/NCI) [C]

I found another bug in the filters: The filter changes are ignoring the text entered by the user for the glossary and replaces it with the TermNameString. For instance, in the summary CDR256491, the word dye that is linking to the glossary CDR409764 (contrast material) is replaced by the term name.

I'm re-opening this ticket.

Comment entered 2016-04-26 18:19:07 by Englisch, Volker (NIH/NCI) [C]

I'm currently working on a fix for the above problem. I'm in the process of testing proper handling of Insertion/Deletion markup for GlossaryRefs and LOERefs.

Comment entered 2016-04-28 18:52:10 by Englisch, Volker (NIH/NCI) [C]

The following filters have been updated on DEV and QA:

  • R13899 : CDR0000335424.xml - Denormalization Filter: Summary

  • R13980 : CDR0000000165.xml - Mailer Filter: Summary LaTeX Preprocess

  • R13980 : CDR0000505579.xml - Copy XML for DrugInfoSummary

  • R13988 : CDR0000780696.xml - Denormalization Filter: Summary GlossaryTerm

  • R13988 : CDR0000000157.xml - Vendor Filter: Summary

  • R13988 : CDR0000335166.xml - Module: Checkbox Formatter

  • R13988 : CDR0000339576.xml - Module: InLine Markup Formatter

  • R13988 : CDR0000409593.xml - Copy XML for Summary Report

I also fixed a bug in the QC report that had not been reported. The LOE terms in the table displayed at the end of HP summaries did not mark-up the inserted/deleted terms but does so now.

I resumed my original task - running diff reports on QA.

Comment entered 2016-04-29 04:48:27 by Shah, Aarti (NIH/NIDDK) [C] [X]

The DTD was also updated for the CDR. The same changes have been made on the GateKeeper side.

Comment entered 2016-04-29 10:20:27 by Juthe, Robin (NIH/NCI) [E]

Is this ready for testing on QA? Just want to be sure I understand. Thanks.

Comment entered 2016-04-29 10:22:52 by Shah, Aarti (NIH/NIDDK) [C] [X]

Hi Robin - yes this ticket is ready for QA. Thanks.

Comment entered 2016-04-29 11:40:48 by Osei-Poku, William (NIH/NCI) [C]

We've started looking at these again also. So far, they're looking good. The terms in the DISs (and other summary types) are selecting correctly and the LOERefs appear be selecting correctly as well. We'll continue to test and report any issues.

Comment entered 2016-04-29 12:28:58 by Juthe, Robin (NIH/NCI) [E]

This looks good to me so far, too. Even the gene term seems to be going to the right place 🙂. I've asked Sharon to take a look at a few of the Peds summaries as well and I'll keep you posted.

Comment entered 2016-04-29 12:39:10 by Englisch, Volker (NIH/NCI) [C]

We have a German saying:

Was lange währt wird endlich gut

Translated it's something like:

All good things are worth waiting for

:-)

Comment entered 2016-04-29 16:56:59 by Osei-Poku, William (NIH/NCI) [C]

Moving this ticket to QA verified. We've reviewed several summaries and all of them looked good.

Comment entered 2016-06-17 15:27:15 by Englisch, Volker (NIH/NCI) [C]

I confirmed that the CDR is providing the necessary attributes needed for Cancer.gov to properly display the HP/patient definitions.
Closing ticket.

Attachments
File Name Posted User
GTPopUp_Future_Process.png 2016-03-29 08:02:25 Sun, Victoria (NIH/NCI) [C] [X]

Elapsed: 0:00:00.001464