Upcoming Planned Work:
- Start to test available data and document the results for quality control and data validation.
The 2BLAST Pipeline Development: (Biostatistical and Bioinformatic Longitudinal Analysis of Symptoms and Toxicities in cancer patients)
Integration and data quality assurance of cluster symptoms and toxicity data in combination with other clinical data is challenging due to a lack of processes and pipelines. 2BLAST will generate automated processes for integrating and analyzing institution-wide data routinely, to help clinical patient management, and support a wide range of research endeavors from genomic to health services research.
AIM 1: Integrate chemotherapy and radiation treatment data with laboratory, outcomes, and patient-reported symptom and toxicity data at Princess Margaret (PM) into a research data repository, with automated quality assurance evaluation of key-variables through a machine-learning approach.
AIM 2: Demonstrate proof-of-principle clinical utility of these pipelines.
- Reviewed each of the following Princess Margaret data sources and identified data elements that would be important to include in the 2BLAST pipeline: OPIS, MOSAIQ, EPR, BDM, CRR, Cancer Registry.
- Documentation for each data source created to properly define the data to be pulled out and to map the lineage of each data element.
- Requests to access data from OPIS, MOSAIQ, and BDM have been approved.
- AIM 2 will demonstrate the proof of concept and will start once AIM 1 is complete.
Last modified: July 17, 2020