Promis & CATs (Computer Adaptive Tests)

CATs & PROMIS Instruments from the REDCap Shared Library

PROMIS^® (Patient-Reported Outcomes Measurement Information System) is a set of person-centered measures that evaluates and monitors physical, mental, and social health in adults and children. It can be used with the general population and with individuals living with chronic conditions. PROMIS questionnaires were developed by the NIH and incorporated into the REDCap Shared Library in 2015.

Initial Considerations

Note: PROMIS forms cannot be edited because they (and CATs-the adaptive PROMIS forms) are constructed a very specific way so as to work with the PROMIS API server, which performs calculation of the score.

Users may select from dozens of CATs (computer adaptive tests) from the REDCap Shared Library, including PROMIS instruments provided by the Assessment Center API software (https://www.assessmentcenter.net/).

There are four types of PROMIS instruments:

Adaptive
Auto-Scoring
Normal (require user to calculate scores). This service is offered by the Clinical Translational Science Institute’s (CTSI) PHR Statistics Qualitative Research foundation
Battery: is a battery of multiple instruments that are bundled together

[Adaptive Instrument] indicates that the instrument is adaptive with scoring, and can only be implemented in survey format as one question at a time. The adaptive PROMIS forms are computer adaptive test (CAT), which means that questions are generated dynamically using previous answers in that survey

[Auto-Scoring Instrument] indicates that the instrument contains scoring that is automatically performed and saved after being completed. The instrument is static (not adaptive), and can only be implemented in survey format as one question at a time. The Auto-scoring PROMIS scores each question behind the scenes. When viewed as a data entry form the questions and scored calculations are read only.

[Normal] instruments can be modified.

[Battery]: is a battery of multiple instruments that are bundled together.

Most instruments of all types offer short forms with few questions. For more information on PROMIS instruments, see http://www.nihpromis.org/.

The Process

From your REDCap Online Designer, click Import a new instrument from the REDCap library
Enter topic for your PROMIS instrument in the Keyword search (e.g. anxiety).
View results options for the search. Each is labeled as Normal instrument or Auto-scoring or Adaptive instruments.
1. The number appended to the end of the instrument title indicates the number of questions in the survey (circled in green).
Click on the arrow next to the desired instrument to expand and see Acknowledgement /Terms of Use.

The instrument can also be viewed as a PDF or webpage by clicking in the upper right corner. This can be send to the IRB as supporting documentation of explanation of your study
Curated instruments highlighted with a star have been approved for inclusion by the REDCap Library Oversight Committee (REDLOC) after review for research relevance, accuracy in function and coding (see guidelines), and copyright issues. Other instruments and forms are shared by individuals or groups from consortium institutions on "as-is" basis.

Click on Import into my REDCap Project
Agree to the terms

Additional Considerations & Questios

Q: Can an Autoscoring instrument be modified?

No - because the instrument is an auto scoring survey , it cannot be modified as noted below when you go into the record status dashboard

Q: Can an Adaptive instrument be modified?

A: No – in an adaptive instrument, questions are generated dynamically base on previous answers. Because it is dynamic, it can only be taken in survey form. All fields are then permanently locked and are not editable from the record status dashboard

Q: When I look at the data dictionary for PROMIS and other adaptive, battery and auto-scored instruments, the coding looks incorrect because some answer choices have the same code value. Why are the coding rules different here?

PROMIS, Neuro-QOL, TBI and other instruments developed by the HealthMeasures Team are unique in nature in that they work with an API service to generate questions and scores. They are created, tested and verified by their team, not the REDCap Shared Library Committee, for accuracy and proper functionality. If you notice in the data dictionary the same code is used for multiple similar responses, it is normal.

Q: Do the "auto-scoring" instruments have any special features that a typical instrument would not have access to – e.g., scoring algorithms, perhaps, that would require external calls to the PROMIS API? Can PROMIS instruments be edited in the Online Designer after download?

A: Auto-scoring, adaptive PROMIS forms cannot be edited because they (and CATs-the adaptive PROMIS forms) are constructed a very specific way so as to work with the PROMIS API server, which performs calculation of the score.

Theoretically, if you know the algorithm, you could probably create a calc field to do the scoring at the end. However, Northwestern, who owns PROMIS, is the group that owns that algorithm, and so you might not be able to obtain it unless you pay for the license to that assessment, which may cost hundreds or thousands of dollars. But if you're curious, it might be worth asking them if you can get it for free.

Q: My survey instructions entered in survey settings are no longer appearing for the PROMIS questionnaires. What happened?

A: Keep in mind that if the surveys have already begun (i.e., participant has pressed Submit or Next Page and is returning), then it will no longer display the instructions. The instructions are only ever displayed before the survey has begun.

Q: Is there a way to import data from an existing project into PROMIS auto-scoring instruments?

A: Unfortunately, there's no real way to import data for PROMIS auto-scoring instruments. Since the scoring is often the critical piece of those, the score can only be obtained by taking the survey in real time, in which the PROMIS API server does all the number-crunching to determine the score before it is saved. So for that reason manual importing of the data is not permissible. But we might think about being more flexible with that in the future, as we have recently with BioOntology fields.

Q: When I only type ‘anxiety’ in the search button a list of non-Promis instruments appear (some of which are adaptive or autoscored). How do these differ from the Promis adaptive / autoscored instruments?

A .In these types of instruments (SCI-FI, ASCQ-Me, SCI-QOL, and TBI-QOL), none of these are literal PROMIS assessments (i.e., have "PROMIS" in the name) but are all CATs, Auto-scoring instruments, or traditional instruments and are distributed via the Assessment Center API.

Results when search for anxiety

Results when search for Promis anxiety

Q: What is the PROMIS API?

A: Both the adaptive and auto-score PROMIS forms use of Assessment Center's Application Programming Interface (API), that allows data collection systems to administer self- and proxy-reported HealthMeasures including PROMIS®, ASCQ-Me® and Neuro-QoL.

Q: The red star icon next to the instrument indicates a REDLOC curated instrument. What does this mean?

A: After instruments have been suggested by researchers, reviewed by REDLOC librarians and recommended for inclusion by REDLOC members, they must then be translated into REDCap format using standardized nomenclature for things like instrument names, field variable name & coded values

Q: The Promis instruments ask one question at a time with the same choices as shown below. Can I modify the instrument such that the questions are displayed as a matrix rather than one question at a time?

A: The PROMIS Adaptive instruments cannot be modified in any fashion including changing to a matrix format