MSACL 2017 US Abstract

Strategies for Human-Proofing High-Throughput Data Analysis

Julia Drees (Presenter)
Kaiser Permanente Regional Laboratories

Bio: Julia Drees is a Scientific Director in Chemistry for The Permanente Medical Group at the Kaiser Permanente Regional Laboratories in Northern California. Her work there includes endocrine testing, the applications of LC-MS/MS and liquid handling robotics for small molecule quantitation, and database mining for reference range studies and quality assurance. Julia earned her Ph.D. in biochemistry from the University of Wisconsin, Madison and completed a fellowship in Clinical Chemistry at the University of California, San Francisco.

Authorship: Julia C. Drees
Kaiser Permanente Regional Laboratories

Short Abstract

Data analysis and reporting can be an error-prone and time consuming part of LC-MS/MS testing even for low volume clinical laboratories. This session will describe how data analysis software and middleware rules can streamline workflows and reduce errors by automatically flagging samples with quality failures and preventing the uploading of those results. A case study will be presented demonstrating how one high volume clinical laboratory further increased throughput by implementing data review by exception. At the conclusion of this session, participants will be able to assess their own workflows and implement solutions that reduce time and errors spent analyzing data.

Long Abstract

Peak review and identifying quality failures in LC-MS/MS is particularly challenging in high throughput laboratories, but data analysis can be an error-prone part of LC-MS/MS testing even in low volume clinical laboratories. This session will describe the data analysis workflows that have made LC-MS/MS analysis in one high-throughput clinical laboratory bearable. The solutions used include data analysis software from LC-MS/MS vendors and off-the-shelf middleware—no custom coding required. Automated flagging of quality failures coupled with suppression of these results guards against human error and greatly streamlines data review. This session will also present a case study of how data review by exception was implemented for one high volume LC-MS/MS assay to increase throughput. The case study will highlight changes made, how they were validated, and the impact they had on reducing data analysis time. This session is intended for participants who are working in high volume laboratories or for those who are interested in learning ways they can optimize and automate data analysis. At the conclusion of this session, participants will be able to assess their own workflows and implement solutions that reduce time spent analyzing data while also reducing the potential for human errors.


References & Acknowledgements:

I would like to acknowledge Judy Stone for original development of middleware rules.


Financial Disclosure

DescriptionY/NSource
Grantsno
Salaryno
Board Memberno
Stockno
Expensesno

IP Royalty: no

Planning to mention or discuss specific products or technology of the company(ies) listed above:

no