Evaluating the Strength of Instrumental Variables in in the Evaluation of Time‐Varying Treatments

There is a growing interest in the use of instrumental variable analysis to evaluate time-varying treatments in comparative effectiveness research. In this setting, a major challenge is to identify an instrumental variable that is strongly associated with treatment assignment at each time point. The strength of association of an instrument is often assessed using the F-statistic. However, the standard F-statistic assumes that the instrument and the treatment are homoscedastic and serially uncorrelated, which is unrealistic in the evaluation of time-varying treatments. This paper proposes the use of a robust F-statistic that allows for the autocorrelated and heteroskedastic nature of time-varying instruments. A simulation study assesses the relative performance of the proposed F-statistic in comparison to standard, commonly used F-statistics. We illustrate the practical advantages of the different approaches in an evaluation of second-line oral anti-diabetic treatments among people with type 2 diabetes mellitus in the UK.

Authors

Institutions

Publication Details

Journal
Biometrical Journal
Published
2026-11-20
DOI
https://doi.org/10.1002/bimj.70200
Primary Topic
Advanced Causal Inference Techniques
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Evaluating the Strength of Instrumental Variables in in the Evaluation of Time‐Varying Treatments

Daniel Tompsett, Manuel Gomes
2 links
Biometrical Journal
Advanced Causal Inference Techniques
article

Evaluating the Strength of Instrumental Variables in in the Evaluation of Time‐Varying Treatments

Daniel Tompsett, Manuel Gomes
article en

Abstract

There is a growing interest in the use of instrumental variable analysis to evaluate time-varying treatments in comparative effectiveness research. In this setting, a major challenge is to identify an instrumental variable that is strongly associated with treatment assignment at each time point. The strength of association of an instrument is often assessed using the F-statistic. However, the standard F-statistic assumes that the instrument and the treatment are homoscedastic and serially uncorrelated, which is unrealistic in the evaluation of time-varying treatments. This paper proposes the use of a robust F-statistic that allows for the autocorrelated and heteroskedastic nature of time-varying instruments. A simulation study assesses the relative performance of the proposed F-statistic in comparison to standard, commonly used F-statistics. We illustrate the practical advantages of the different approaches in an evaluation of second-line oral anti-diabetic treatments among people with type 2 diabetes mellitus in the UK.

Biometrical JournalVol. 68(6)
University College London (GB)
Queen Mary University of London, Barts Charity, Medical Research Council
Good health and well-being
Advanced Causal Inference Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon