Conflicting race/ethnicity reports: lessons for improvement in data quality

Area of Science:

• Health Informatics
• Cancer Registry Data Quality
• Patient Demographics

Background:

• Accurate race and ethnicity data are crucial for understanding health disparities.
• Conflicting reports between hospital discharge data and cancer registries can impact data integrity.
• The Data Improvement Project on Patient Ethnicity and Race (DIPPER) aimed to address these data quality issues.

Purpose of the Study:

• To determine the frequency of conflicting race/ethnicity reports in cancer cases.
• To analyze patterns associated with these reporting discrepancies.
• To identify strategies for improving the quality of race and ethnicity data.

Main Methods:

• Analysis of matched hospital discharge data and central cancer registry data from 2009.
• Comparison of race and ethnicity information for cancer patients across the two datasets.
• Examination of a sample of 2,356 patients, identifying 187 with conflicting race reports (7.9%) and 357 with conflicting ethnicity reports (15%).

Main Results:

• Unknown ethnicity responses were nearly twice as common as unknown race responses in hospital discharge data.
• Approximately 85% of mismatched race cases were reported as non-white in hospital data and white in registry data.
• The most frequent ethnicity mismatch involved an 'unknown' code in hospital data versus 'non-Hispanic' in registry data.

Conclusions:

• Hospital cancer registrars often face challenges accessing complete race and ethnicity data.
• Addressing discrepancies through flagging, verification, and improved staff training is essential.
• Integrating race and ethnicity data within electronic medical records and considering ethnicity for registry standards could enhance data accuracy.