How to handle when won't know a participant's arm at the beginning of a study

[Isaac Petersen]

I'm creating a REDCap project for a study in which the participant would be assigned to one of various arms based on their responses in the screening. Thus, I will not know which arm the participant will be in at the beginning of the study. According to this link (https://www.utsouthwestern.edu/edumedia/edufiles/research/center_translational_medicine/redcap-faq.pdf), you can set up an arm as a "screening and enrollment" arm. Then, once a subject becomes enrolled, the participant can be added to an "active" arm. But it's not clear to me how to add a new arm to a participant (or how to change their arm).

One challenge is that the first instrument of my project (i.e., the screening instrument) is a public survey link. Thus, it requires using auto-numbering of record_id. This means that, in the "Add / Edit Records" section, I don't have the option to "Enter a new or existing Record ID". How can I add/change a participant's arm to one of the active arms after they were in the screening arm?

Your help would be greatly appreciated.
-Isaac

[Rick Watts]

'Arms' in REDCap are not as sophisticated as they sound, and they are not quite synonymous with how I would use the term. Really they are just a way of representing different event/instrument schedules. In actual fact, the same participant can exist in more than one study arm at the same time. So if you were manually assigning numbers you could start a participant in one arm (let's say screening) and then later create a participant with the same record_id in a different arm (Treatment A or Treatment B for example). The new arm/record would represent the continuation of the same participant. But then there's the question of the study ID. When you are auto-assigning study IDs REDCap assigns them for all the arms and you do not have the opportunity to assign your own ID or re-use IDs in more than one arm. However, some of the 'magic' is in the URL. When you view the URL for a Record Home Page it will look something like:

• example.com/redcap_v10.6.2/DataEntry/record_home.php?pid=391&id=3&auto=1&arm=1
  

The above is the URL for record 3 in arm 1 and the auto=1 parameter tells REDCap that we are autonumbering. However, if you enter that URL manually, you can enter a URL for a record that doesn't already exist and REDCap will create a new one!

•   example.com/redcap_v10.6.2/DataEntry/record_home.php?pid=391&id=3&auto=1&arm=2

Entering this URL (above) will create a new record 3 in arm 2.

This is not very helpful, since the URL has to be typed but, since this is a regular URL it can be used in a bookmark, or in a link in descriptive text (embed HTML and use piping) or in a survey redirect. This could allow you to create a mechanism to create the new record from within the study without having to type the URL by hand.

So, I've not tested ALL this fully, but it seems to me that what you need is a study with three arms. The participant starts in arm 1 and, at some point, a participant with the same id gets created in either arm 2 or arm 3 depending on your 'randomization' or other factors. Also, I can't guarantee that Vanderbilt may "fix" this feature (the URL tweaking) in the future and so break your project. (Then there's the question of whether you really need multiple arms. Could you create the effect you need with branching? Could you get by simply displaying guidance for what instruments a coordinator should complete based on the participant's context?)

Rick Watts

Team Lead, Research Informatics

rick....@ualberta.ca

T: (780) 248-1170

Women and Children’s Health Research Institute

University of Alberta

5-083 Edmonton Clinic Health Academy (ECHA)

11405 87 Avenue NW Edmonton, AB T6G 1C9

F:  (780) 248-5616

WCHRI is a partnership between the University of Alberta and Alberta Health Services, funded by the generosity of the Stollery Children's Hospital Foundation and supporters of the Lois Hole Hospital for Women.

--
You received this message because you are subscribed to the Google Groups "redcap open" group.
To unsubscribe from this group and stop receiving emails from it, send an email to redcap_open...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/redcap_open/CAPBn5Xsg4F38V7Y-46J3HJDNEX9w2u6Vd7xD3xbSqV6ojjj%3DBA%40mail.gmail.com.

[Peter Macisaac (POP)]

Isaac and Rick I agree that the concept of “arm” in study design thinking is not confused with “arms” in redcap implementation..

For example  common model is to use a redcap form or survey to screen individuals who if eligible can then be randomised (if required) or entered into a study. That way only some of the records progress to the followup data collection stages.

If there is a real practical need to screen participants then move them to different studies I have used the methods you describe of creating a record in the new study by generating the intended study URL -  REDCap will create a record if you feed it the right URL

I have been providing assistance to a team at an external Uni who have had the following  scenarios:

I have used this where the subject of the screening (e.g. family) is different from the subject of the study (e.g. child)

We use one REDCap project to collect data about the family and then create in a second redcap project records relating to the indvidual - this helps as the redcap iD signifies the subject of interest and here we have to different types of subjects of interest _ families and children

In another project I have used one project to screen the participants then created a another project to collect study data on eligible particpants - normally this can be handled in one project however we are using the mobile app and there are large numbers in the study so removing extraneous data during data collect was a recommended strategy to make the mobile app work better

A third scenario is one where a set of participants are part of multiple studies however you want to keep an eye on which studies they are involved in by having a master list - the master list controls the documentation of the studies and particpants and then generates the study ID with consistent participant identifier in many sub-studies.

Here is the secret sauce:

<h2><a href="https://capricorn.bc.edu/redcap/redcap_v10.3.4/DataEntry/index.php?pid=802&id=[record_id_new]&page=household_enrollment_form&event_id=2965"  target="_blank"">Click on this link to complete record creation for [household_headname] family</a></h2></p>

Create the new project for one of the arms

Create a single record - record ID 1

Then copy the URL

Add in the piping from the setup up project to give you the ID you want to create own the target project. -  above this is 

[record_id_new]

Add the target = blank attribute to open the new record in another tablet -  it helps usually to do this

Add in the text to describe what the URL does.  eg.  "Click on this link to create new…"

Create the URL and add it to a descriptive text type of field 

So now you have the ability to invoke it in a form for eligible particpants 

PRESTO you open a new form in the right study

And with cross project data sharing external module you can even copy the data collected for screening into the new project.

HOWEVER before going down the above route which is complex and is using undocumented features of REDCap, knowledge of HTML and external modules,   I would encourage further analysis  of Isaac's requirements to be sure that his needs cant be done in a more standard way.  Simple is better.

Isaac if you can more fully document the story of your project - like a synopsis of the study design - that will help structure the data needed to  answer the question - how to get REDCap to work for you in your project.

Kind regards

Peter MacIsaac

PS.Isaac,   If you need more detailed help on whether you need to do this or how to achieve it if you get stuck you can contact me offline

To view this discussion on the web visit https://groups.google.com/d/msgid/redcap_open/CAAh1wLtcU-VCTJ9TJNeTyNx1qJgKLoz2p3a%3D3owkjsMc2CC1pQ%40mail.gmail.com.