1. What is the decision or change (e.g., clinical topic, practice guideline, system design, delivery of care) you are facing or struggling with where a summary of the evidence would be helpful?

Understanding the uncertainty of prediction of both classical and machine learning decision support tools. Not all predictions have the same validity, and the error bound are highly dependent of the individual prediction.

2. Why are you struggling with this issue?

I am not struggling with this, but my clinical colleagues always ask the question. Ok the predictive value, area under the curve for the population in the training and test sets are great, but what about my next patient. How about unrepresented and underserved populations that are not included in the training sets.

3. What do you want to see changed? How will you know that your issue is improving or has been addressed?

A well develop literature that provide best practices and implementation guidelines of assessing uncertainty quantification at the point of care.

4. When do you need the evidence report?

Friday, 06/30/2023

5. What will you do with the evidence report?

Extend our previous work on the topic and engage clinicians in identifying the most critical issues.

Optional Information About You

What is your role or perspective?  Scientist