Objective: Carotid intima-media thickness (CIMT) is a marker of atherosclerosis that is commonly used to assess the effect of therapeutic interventions. It is currently unclear to what extent biologically implausible values affect treatment effects. We evaluated the impact of biologically implausible CIMT values on the estimated rate of change in CIMT.
Methods: Data were used from the METEOR (Measuring Effects on Intima-media Thickness: an Evaluation of Rosuvastatin) trial. METEOR was a randomized, placebo-controlled trial showing that rosuvastatin reduced the 2-year change in CIMT among low-risk individuals with subclinical atherosclerosis. In the main METEOR analysis, the data were analyzed without exclusion of biologically implausible data. In this post-hoc analysis, we constructed twelve definitions to define mildly or extremely biologically implausible values using distance from the interquartile range, median or mean. We evaluated the effect of removing implausible values on the estimated rate of change in CIMT.
Results: The percentage of biologically implausible CIMT values ranged from 0.6% to 9.7%, depending on the definition used. Across all definitions, removal of biologically implausible CIMT values marginally reduced standard errors and did not change the primary outcome (i.e., a nonsignificant change in the rosuvastatin group, significant progression in the placebo group, and a statistically significant difference between treatment groups).
Limitation: This study was focussed on the impact of implausible values in the analytical part of a CIMT study. Ultrasound images were not re-examined to determine whether an implausible measurement was due to measurement error or temporal morphological thickening,
Conclusion: Removal of biologically implausible CIMT values marginally decreased the variability of the estimated rate of change in CIMT without having a large impact on the estimated rate of change.
Trial registration: ClinicalTrials.gov NCT00225589.