Methods for Retrospectively Improving Race/Ethnicity Data Quality: A Scoping Review.

Improving race/ethnicity data quality is imperative to ensuring underserved populations are represented in datasets used to identify health disparities and inform healthcare policy. We performed a scoping review of methods that retrospectively improve race/ethnicity classification in secondary datasets. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, searches were conducted in MEDLINE, Embase and Web of Science Core Collection in July 2022. A total of 2,441 abstracts were dually screened, 453 full-text articles were reviewed, and 120 articles were included. Study characteristics were extracted and described in a narrative analysis, including: method type used for race/ethnicity classification; races/ethnicities targeted for classification; publication year; method inputs; reference population (if applicable); target population; and whether the article included a validation process. Six main method types for improving race/ethnicity were identified: Expert Review (n=9; 8%), Name Lists (n = 27; 23%), Name Algorithms (n=55; 46%), Machine Learning (n=14; 12%), Data Linkage (n=9; 8%), and Other (n=6; 5%). The main racial/ethnic groups targeted for classification were Asian (n = 56; 47%) and White (n = 51; 43%). Eighty-six articles (72%) included some form of validation evaluation. We discuss the strengths and limitations of different method types and potential harms of identified methods. We recommend the need for innovative methods to better identify racial/ethnic subgroups and further validation studies. Accurately collecting and reporting disaggregated data by race/ethnicity is critical to address the systematic missingness of relevant demographic data that can erroneously guide policymaking and hinder the effectiveness of healthcare practices and intervention.