The impact of ignoring measurement error when estimating sample size for epidemiologic studies

Area of Science:

• Biostatistics
• Epidemiology
• Health Research Methods

Background:

• Sample-size estimation is critical for study power and validity.
• Measurement error in study variables can bias these estimates.
• Ignoring such errors may lead to underpowered or over-resourced studies.

Purpose of the Study:

• To illustrate the impact of measurement error on sample-size calculations.
• To provide methods for adjusting sample-size estimates for misclassification and measurement error.
• To guide researchers in planning studies with accurate sample-size requirements.

Main Methods:

• Demonstration through two illustrative examples.
• Analysis of outcome misclassification bias in sample-size estimates.
• Evaluation of classical measurement error in predictor variables using linear regression.

Main Results:

• Ignoring outcome misclassification can significantly affect sample-size accuracy.
• Classical measurement error in independent variables can lead to severe underestimation of required sample size.
• Methods for adjusting sample-size estimates are presented.

Conclusions:

• Accurate sample-size estimation requires accounting for measurement error in both outcome and predictor variables.
• Researchers should consider measurement error adjustments during study planning.
• Literature review provides further guidance for study planners.