Skip to Content
SBMI Horizontal Logo

Evaluation of Real-Time PCR Assays and Liquid Bedd Arrays for the Molecular Detention of Gastrointestinal Pathogens in a Pediatric Population

Author: Laura Millecker (2013)

Primary Advisor:  Sriram Iyengar, PhD

Committee Members: Trevor Cohen, MBChB, PhD

Masters thesis, The University of Texas School of Biomedical Informatics at Houston.

Abstract:

Introduction: Physicians rely on molecular techniques for infectious disease detection. As demand for molecular technologies has increased so too has that of software capable of data analysis. Limited resources can hinder the automatic analysis and electronic transfer of patient results leaving molecular technologists to manually perform data analyses. Therefore, data analysis software needs to not only provide results quickly, but also needs to be easy to learn and user friendly. A task analysis and user satisfaction survey was utilized to evaluate the usability and likability of molecular data analysis software applications at Texas Children's Hospital.

Methods: A total of 202 previously positive and negative stool samples were obtained and analyzed with a Roche LightCycler 2.0 and a Luminex 200 instrument. Sensitivity, specificity, positive predictive values (PPV) and negative predictive values (NPV) were calculated for target agreement within each panel and overall for each panel. CogTool and a System Usability Scale were used to perform a task analysis and a usability survey for each platform's software, respectively.

Results: Ninety seven samples (970 targets) were tested with the Acute Bacterial panel (Table 2), 77 samples (154 targets) with the C. difficile panel (Table 3) and 99 samples (495 targets) with the Viral Panel (Table 4) yielding target agreements of 963/970 (99.3%), 1301154 (84.4%), and 454/495 (91.7%), respectively. Combined sensitivity and specificity for all panels was 155/159 (97.5%) and 102/114 (89.5%), respectively. CogTool revealed the LightCycler software required 6 steps and 9.7 seconds for data analysis, while the Luminex software required 5 steps and 7.5 seconds. The SUS scores for the LightCycler software range from 67.5-80, with an average of77.2.  For the Luminex TDAS LSM software, SUS scores range from 72.5-85 and have an average of 79.7.Discussion: The Luminex assays enhance the current molecular test menu at Texas Children's Hospital  by enabling the detection  of multiple GI pathogen targets and include data analysis software that is highly sought-after  and user-friendly.