The test-negative design: What bias is it intended to address?
Addressing what type of bias the test-negative design (TND) addresses, and showing the key bias that arises due to misclassification of the outcome.
Test-negative studies recruit ‘cases’ who test positive for a particular disease; ‘controls’ are patients undergoing the same tests for the same medical reasons and who test negative. The design is often used for vaccine effectiveness studies, but it has also been used to assess risk factors for disease in the context of the Covid-19 pandemic, and previously for other diseases. Although the test-negative design (TND) has been widely used, and also explored theoretically, there has been little agreement on what type of bias the TND is intended to address, or what the corresponding Directed Acyclic Graph (DAG) is.
In this seminar, we consider the situation where the only available disease outcome is from the natural process of people being tested in a particular population (i.e. no additional testing is done for the study), and those who test positive (the cases) are compared to the rest of the “general population” (or a sample of general population controls), not all of whom have been tested. In this situation, bias arises due to misclassification of the outcome. This bias occurs irrespective of whether the entire source population is studied, or a “standard” case-control study (with “general population” controls) is conducted.
The TND intends to address this bias by restricting the source population to the subpopulation who get tested. Once again, this approach works whether or not all of this new subpopulation is studied, or a TND case-control study is conducted. Either approach controls (fully or in part) for determinants of who gets tested, including health-seeking behaviour (HSB).
We then consider more general and complex situations in which some biases (including residual confounding) may occur even after restriction to those who are tested. However, even in this situation, any remaining bias is likely to be small, since restriction to those who are tested, indirectly adjusts for health seeking behaviour (HSB). We conclude that the TND is a design that is excellently suited to give valid estimations of a wide range of risk factor-outcome associations, and can be applied to large populations since its restriction to tested persons largely eliminates important sources of bias.
Speakers
- Professor Neil Pearce - Professor of Biostatistics and Epidemiology, LSHTM
- Professor Elizabeth Williamson - Professor of Biostats and Health Data Science, LSHTM
- Dr Thiago Cerqueira Silva – Assistant Professor, LSHTM
Event notices
- Please note that you can join this event in person or you can join the session remotely
- Please note that the recording link will be listed on this page when available
Admission
Contact
