Master's degree in Statistics and Operations Research

Statistical analysis of diagnostic tests.

Date:

July 6 to 10. AFTERNOON: 15 to 18h.

Instructor

Anna Felip-Badia, Natàlia Pallarès Fontanet, Sara Perez-Jaume

Anna Felip-Badia: Anna obtained her Bachelor’s Degree in Mathematics in 2023 from the Universitat Politècnica de Catalunya (UPC). She later specialised in Biostatistics through the Master’s Degree in Statistics and Operations Research jointly offered by the UPC and the Universitat de Barcelona (UB), which she finished in 2025. Since July 2024, Anna has been working as a biostatistician at the Developmental Cancer Research Laboratory of the Fundació Sant Joan de Déu – Pediatric Cancer Center Barcelona. She is also a substitute professor in the Statistics and Operations Research Department of the UPC, teaching statistics to undergraduate engineering students.

Natàlia Pallarès Fontanet: Natàlia Pallarès has a Bachelor’s Degree in Mathematics from the UPC, a Master's Degree in Statistics and Operations Research from the UPC-UB and a PhD in Medicine and Translational Research (UB). From 2015 to 2023, she worked as a biostatistician at the Hospital del Mar Research Institute (IMIM) and at the Bellvitge Biomedical Research Institute (IDIBELL). Since 2023, she has worked as a senior biostatistician at the Germans Trias i Pujol Research Institute (IGTP) Biostatistics Support and Research Unit, providing statistical advice to researchers. Since 2015, she has also worked as an associate lecturer at several Catalan universities, including the Universitat Autònoma de Barcelona (UAB) and the UB.

Sara Perez-Jaume: Sara holds a Bachelor's Degree in Mathematics from the UPC, a Master's Degree in Statistics in Operations Research from the UPC-UB and a PhD in Medicine and Translational Research from UB. Sara has a ten-year experience working as a biostatistician at the Developmental Cancer Research Laboratory of the Fundació Sant Joan de Déu – Pediatric Cancer Center Barcelona. She currently works as a professor in the Biostatistics Unit of the Department of Basic Clinical Practice at UB.

Language

English

Description

Studies of diagnostic accuracy are of the utmost importance in clinical and biomedical research. The development of new biomarkers is an active area of research nowadays, and a proper analysis and evaluation of these biomarkers is essential to reach valid conclusions. In this course, we review basic concepts on the analysis of diagnostic tests, such as sensitivity, specificity, predictive values and other accuracy measures. More advanced topics such as Receiver Operating Characteristic (ROC) curves and meta-analysis are also covered. We also explain strategies for designing diagnostic accuracy studies and for analysing the resulting data. The course covers both binary and continuous diagnostic tests and includes many real case examples to illustrate the methodology.

Course goals

This course aims to provide participants with the tools to compute, understand and interpret measures of diagnostic accuracy from both statistical and practical perspectives. Participants will also learn how to analyse data from diagnostic studies using a wide variety of R packages. These methods, while widely used in biostatistics, are also useful in economics and related fields, and the R tools covered in the course can also be applied to such settings.

Course contents

  1. Introduction to diagnostic tests
  2. Assessment of binary diagnostic tests: measures of accuracy, comparison of binary tests
  3. Assessment of continuous diagnostic tests: ROC curve, ROC surface, optimum threshold estimation in the two- and three-state settings
  4. Diagnostic tests with time-to-event outcome: time-dependent ROC curve, optimum threshold estimation with right-censored outcomes
  5. Introduction to meta-analysis of diagnostic accuracy studies: meta-analysis of binary tests, summary ROC curve

Prerequisites

Knowledge of R, basic knowledge of statistics.

Targeted at

Master's degree students, PhD students and biostatisticians giving methodological support to physicians or biologists.

Evaluation

An exercise to do at home in which participants will be required to apply the techniques explained in the course.

Teaching Methodology and Activities

Theoretical and practical sessions will be combined.

Software requirements

R version 4.5.0 or higher (free). The use of a personal laptop is strongly recommended.