2019/07/24 by Benjamin Ackerman, Juned Siddique, Ackerman, Benjamin +3
Mathematics · Medicine · #Advanced Causal Inference Techniques #Applications (stat.AP) #FOS: Computer and information sciences #Methodology (stat.ME) #Nutritional Studies and Diet #Statistical Methods and Inference
paper · pdf · doi:10.48550/arxiv.1907.10722
openalex publication_date 2019/07/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Many lifestyle intervention trials depend on collecting self-reported\noutcomes, like dietary intake, to assess the intervention's effectiveness.\nSelf-reported outcome measures are subject to measurement error, which could\nimpact treatment effect estimation. External validation studies measure both\nself-reported outcomes and an accompanying biomarker, and can therefore be used\nfor measurement error correction. Most validation data, though, are only\nrelevant for outcomes under control conditions. Statistical methods have been\ndeveloped to use external validation data to correct for outcome measurement\nerror under control, and then conduct sensitivity analyses around the error\nunder treatment to obtain estimates of the corrected average treatment effect.\nHowever, an assumption underlying this approach is that the measurement error\nstructure of the outcome is the same in both the validation sample and the\nintervention trial, so that the error correction is transportable to the trial.\nThis may not always be a valid assumption to make. In this paper, we propose an\napproach that adjusts the validation sample to better resemble the trial sample\nand thus leads to more transportable measurement error corrections. We also\nformally investigate when bias due to poor transportability may arise. Lastly,\nwe examine the method performance using simulation, and illustrate them using\nPREMIER, a multi-arm lifestyle intervention trial measuring self-reported\nsodium intake as an outcome, and OPEN, a validation study that measures both\nself-reported diet and urinary biomarkers.\n