Innovative Statistical Approaches in Clinical Microbiology Research

• Activity Number: 291
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Statistics Without Borders
• Author(s): Jayawant Mandrekar*+
• Companies: Mayo Clinic
• Keywords: Inverse prediction ; Deming's regression ; Bayesian latent class models

Abstract

Academic medical centers offer opportunities to consult on clinical projects that involve novel applications of statistical methods. The focus of this presentation is to provide brief illustrative examples of projects from infectious diseases and clinical microbiology. Specifically, statistical approaches including; 1) use of inverse prediction to estimate the true known concentration on the basis of the value obtained from real-time PCR, 2) use of Deming's regression to compare the performance of different assays in clinical microbiology studies, and 3) use of Bayesian latent class models to estimate the performance of diagnostic tests (sensitivity, specificity, negative predictive value and positive predictive value) when there is no gold standard. Findings from these studies directly impact patient care and/or offer cost savings with improved efficiency.