Project overview

This website is the result of a collaboration between researchers from Achira Labs (Bangalore, India) and McGill University Health Centre (MUHC, Montreal, Canada)—a project funded by Grand Challenges Canada.

The larger goals of this project are to review and clarify methods for the statistical characterisation (e.g. calibration, analytical validation) of new quantitative assays (pre-market) with a special focus on POCT.

Specifically:

• Develop the statistical model used to estimate sample concentration from an observed response—the calibration curve.
  
• Characterise a test’s accuracy in terms of:
  • Reliability: the variation of the test within and across stipulated conditions;
    
  • Validity (or agreement): The extent of bias and random variation in the new test compared to a reference method.

Collectively, we will refer to these as analytical validity.

The application of these methods is illustrated in a set of tutorials using R software. The literature on calibration curve-fitting and pre-clinical diagnostic test validation is synthesised in brief, narrative reviews (in development).

A summary of the terminology can be found in the glossary.