We are updating our Privacy Policy and Terms and Conditions to help you clearly understand how your personal information is collected, stored and managed. Learn More
Menu Back to Poster-Presentations-Details

W 52: Visualizing Patients' ADaM Data via SAS and R





Poster Presenter

      Bella Feng

      • Statistical Programming Manager
      • Amgen Inc
        United States

Objectives

This poster intends to share our experience on creating an ADaM dataset and some graphs for an FDA RTQ question. The challenge was, how to show the patients' concurrent medications in parallel with dosing information and adverse event while those information are scattered in separate ADaM datasets.

Method

We used SAS to create the datasets and graphs during the time of the RTQ. As hindsight, we realized R might do a better job and create the graphs easily. Therefore, this poster is going to show using SAS and R separately and the pros and cons of each method.

Results

The quantitative data was in separate CDISC ADaM datasets: ADCM, ADAE, ADLB, ADEX and ADSL. The intent is to see if there is any trend between patients calcium supplement intake and their dosing reduction and IPTH lab results. The request is particularly challenging for programming when the patients take several kinds of calcium supplements, and when there are overlaps between the end date of one supplement with the next one. We did our programming in SAS during the time of RTQ. Right afterwards, afterwards, I attended an AMGEN internal conference and got inspired in two aspects: 1. We could use an alternative data step in SAS to elegantly combine different ADaM datasets to show the patients' various data at one point of time. 2. I also attended an introductory R talk and found out that R offers a dplyr package to manipulate data. It could also be the right tool to visualize the data and show the trend much more easily than SAS. Therefore, the different strategies will be explored. In terms of data processing, I will compare SAS proc transpose with dplyr in R studio. My hypothesis is R will have a lot of advantages over SAS in terms of programming time and efficiency.

Conclusion

After the different alternatives are tried out, it’s concluded that R (dplyr) works more efficiently in transposing the data and putting them together for visualizing. It offers an excellent alternative tool for dealing with RTQ questions. If it can’t be used for submission, at least it could be helpful for statisticians in validating SAS results.