Assessing Linkage Eligibility Bias in the National Health Interview Survey

Area of Science:

• Health Services Research
• Biostatistics
• Survey Methodology

Background:

• Linking health survey data with administrative records enhances analytical capabilities.
• However, incomplete participant eligibility for linkage can introduce new sources of bias.
• This bias can arise from differences between the full survey sample and the linked subset.

Purpose of the Study:

• To evaluate linkage eligibility bias in health survey data.
• To demonstrate methods for mitigating this bias through questionnaire design and weight adjustments.

Main Methods:

• Utilized data from the National Health Interview Surveys (2000-2013).
• Estimated linkage eligibility bias across diverse sociodemographic groups.
• Assessed bias related to various health-related variables.

Main Results:

• Identified specific sociodemographic and health variables affected by linkage eligibility bias.
• Quantified the extent of bias in estimates derived from linked data.
• Demonstrated the potential impact of bias on research findings.

Conclusions:

• Researchers must acknowledge and assess linkage eligibility bias in linked health data.
• Employing strategies like survey weight adjustments is crucial for reducing bias.
• Careful consideration of these factors ensures the validity of analyses using linked survey and administrative data.