Login / Signup

The objective function controversy for group testing: Much ado about nothing?

Brianna D HittChristopher R BilderJoshua M TebbsChristopher S McMahan
Published in: Statistics in medicine (2019)
Group testing is an indispensable tool for laboratories when testing high volumes of clinical specimens for infectious diseases. An important decision that needs to be made prior to implementation is determining what group sizes to use. In best practice, an objective function is chosen and then minimized to determine an optimal set of these group sizes, known as the optimal testing configuration (OTC). There are a few options for objective functions, and they differ based on how the expected number of tests, assay characteristics, and testing constraints are taken into account. These varied options have led to a recent controversy in the literature regarding which of two different objective functions is better. In our paper, we examine these objective functions over a number of realistic situations for infectious disease testing. We show that this controversy may be much ado about nothing because the OTCs and corresponding results (eg, number of tests and accuracy) are largely the same for standard testing algorithms in a wide variety of situations.
Keyphrases
  • infectious diseases
  • healthcare
  • machine learning
  • systematic review
  • deep learning
  • high throughput
  • decision making
  • single cell