2005 Scientific Session Abstracts
Simple Programming for the Molecular Diagnostics Laboratory: Automated Data Analysis for Increased Accuracy and Efficiency
Alexis B. Carter, MD 1 ; (carterab@upmc.edu) Ronald P. Busch, BS(MT) 1; Jeffrey A. Kant, MD, PhD 1, 2; James H. Harrison, MD, PhD 1. 1 Departments of Pathology, University of Pittsburgh Medical Center , Pittsburgh, PA; 2 Human Genetics, University of Pittsburgh Medical Center, Pittsburgh, PA.
Context: Automation in specialized molecular laboratories lags the main clinical laboratory because of the relative newness of molecular techniques and instrumentation as well as lack of impetus due to lower sample volumes. With increasing workloads, manual entry of results into calculation programs and laboratory information systems is becoming more time-consuming. Manual entry is also prone to transcriptional error and requires verification of information at each step.
Technology: Until mid 2004, data from viral load assays performed on an Applied Biosystems (ABI) PRISM ® 7000 Sequence Detection System in our laboratory were manually transferred from the ABI printout to a Microsoft Excel 2000 worksheet pre-configured with formulas to convert copies per reaction into IU/mL. Internal positive controls and quality controls were manually verified for performance. Data analysis required 30-45 minutes of technologist time per run. Rechecking data occupied an additional 30 minutes each of fellow and faculty time.
Design: Data for each viral load run was exported from the ABI 7000 as a delimited text file. A program was constructed with Visual Basic for Applications in Microsoft Excel 2000 which automatically imported this file into the correct fields on the spreadsheet, performed calculations and checked performance of controls. Samples or controls not meeting predetermined criteria were automatically highlighted and rules-driven comments applied.
Results: Programming took approximately 10 hours. The program was parallel-tested against manual methods over a period of two months. The program, which is now in routine clinical use, runs in three minutes. Approximately 2 hours of technologist and 1.4 hours each of fellow/faculty time is saved per week.
Conclusions: This program has provided significant benefits including increased satisfaction of technologists and professional staff, reduced transcriptional errors, and time savings which have enabled the laboratory to adjust to increasing workload without adding personnel.