OHDSI Home | Forums | Wiki | Github

Do you want to phenotype? Phenotype Development and Evaluation Workgroup

General
#1

Please join the OHDSI Phenotype Development and Evaluation workgroup here.

We had a great conversation today and one of our OKRs for Q1 2022 is

Enable 10 new OHDSI collaborators to develop and evaluate phenotype using OHDSI phenotype best practices.

As part of Phenotype Phebruary - we invite you to Phenotype and @AzzaShoaibi and @Gowtham_Rao are here to help you in your journey. We will have office hours and in those office hours we can help you with

  1. What are the steps needed to phenotype
  2. We will guide you thru clinical description and literature review
  3. We will guide you thru building concept sets
  4. We will guide you thru building cohort definitions
  5. We will guide you running cohort diagnostics and reviewing results

If this is of interest to you, please put your name, time frame , phenotype of interest here. The phenotypes voted by the OHDSI community will be preferred. List here Welcome to Phenotype Phebruary!

As of Feb 11th 2022 - here are the phenotypes that have been taken

image
image1852Ă—905 97.9 KB

Prevalence of Autoimmune Disease - OMOP Script available (execution time 3 to 10 min)
Logical problems with relationships between some “sepsis” concepts in OMOP vocabularies
Device_concept_ids invalidated with no 'Maps To' or 'Concept Replaced By'
#2

@david_vizcaya post

Hi! Marcela Rivera and I would like to take our chance to create and evaluate a phenotype on Acute Kidney Injury. We are based in Europe, so we would rather take earlier hours for US-based colleagues (any time before your midday should work…more or less)

#3

@Marcela and @david_vizcaya - that would be great. Acute Kidney injury is on the list - so lets pick a date and start working together towards developing and evaluating this phenotype

2 Likes
#4

Hi @Gowtham_Rao and @AzzaShoaibi,

On behalf of my department, here’s a few we’d like to participate in. Now, I still have some work to do to ensure CohortDiagnostics works with Databricks/Spark, and we are very novice at PHOEBE and PheValuator. But we’re set with Atlas!

  • @CarolynB and I are poking at SLE a bit in our Atlas instance, and raised a few initial questions in the thread
  • @jshaw03 (I hope I tagged the right name) and I are interested in the acute MI phenotype.
  • I have a few oncology focused colleagues I will be assisting in participating in the prostate cancer and NSCLC phenotypes

Thanks,
Ajit

#5

Great - come join the conversation here for MI

Its on the OHDSI Phenotype Development and Evaluation workgroup - right now I am doing a read of medical text book (Harrisons - just refresh my memory) and put my notes (its on the teams channel).

image
image1110Ă—460 30.5 KB

#6

@agolozar is signed up for posting these. @agolozar please let me know if i or folks in the Phenotype workgroup may be of assistance

#7

@Gowtham_Rao Hello, I’d like to work on COPD phenotyping

#8

@Gowtham_Rao I am interested in joining the conversation in the next Prostate Cancer and HIV phenotypes targets and we will try to execute the cohort definitions in our CDM :smiley:

#9

Hi @Gowtham_Rao, we’ve started with the clinical definition, would be great to get feedback from you and @AzzaShoaibi before implementing in Atlas/Phoebe(?)

1 Like
#10

Hello @AzzaShoaibi @Gowtham_Rao ! I am interested in created a phenotype for Acute Myeloid Leukemia (AML). I am also interested in having a conversation about how to create a phenotype to identify patients receiving chemotherapy (any chemo). I am unfortunately out of town later this week into next week but would love to collaborate after 2/21/22.

Thanks!
Ryan

1 Like
#11

That’s awesome! Let’s start with a clinical description. Please let me know if you want to meet with us and we can set up office hours

#12

@Marcela @david_vizcaya @AzzaShoaibi and I are working on this phenotype on ohdsi MS teams channel. If anyone else is interested please message us

#13

Ok great! I will work on the clinical description and reach out next week to set up a time to meet. Thanks again!

1 Like
#14

Hi @mmayer, I will send a meeting invite soon to start the work on prostate cancer.

Thanks @Gowtham_Rao! Will be in touch soon. :slight_smile:

#15

Please, I am interested in participating in the discussion and execute the cohort definition in our CDM in the depression phenotype.

1 Like
#16

That’s awesome @mmayer . have you or anyone you know run CohortDiagnostics?

#17

Hello!

I’ve really been enjoying the activity around Phenotype Phebruary and greatly admire this community’s effort in collaborating to develop standards around phenotype definitions.

Even ahead of this post, I’ve been working with a group of students to develop a library of phenotypes - could there be a way for us to share those that we’ve developed and have them be apart of this effort?

Related - I wonder if there’s interest in collaborating to conduct large-scale studies of drug effects on these phenotypes? I’m in the protein-protein interaction network space, and using these methods you can predict more drug-disease effects than evidence supports. I’m interested in doing a few proof of principle experiments and then would be interested in advice from this community on how to realistically scale this project).

I could envision a few proof of principle experiments:

  • measuring one drug’s effects on several phenotypes (I know that CohortMethod already facilitates this)
  • measuring several drug’s effects on a single phenotype (we could easily test the phenotypes characterized from Phenotype Phebruary)

If anyone is interested in the latter, let me know and perhaps I can start a separate thread about this topic.

1 Like
#18

Yes ofcourse - the avenue would be to thru the OHDSI Phenotype Development and Evaluation workgroup. The workgroup meets on 2nd and 4th friday of every month from 10 to 12 est.

there are workgroups in OHDSI that is focused on evidence generation and in methods development that would be very interested in such efforts. The workgroup page is here OHDSI Workgroups – OHDSI

t