Search

Background: Structural racism lies at the root of inequities; however its impact on rheumatology care is understudied. Redlining was a US federal government-sponsored practice that mapped areas with high concentrations of Black and immigrant residents as hazardous for investment. We aimed to investigate the association of historical redlining and present-day racialised economic segregation, on health-care utilisation among individuals with rheumatic conditions in the US state of Massachusetts and surrounding areas., Methods: This retrospective observational cohort study used multihospital data from the Mass General Brigham Research Patient Data Registry to identify individuals aged ≥ 18 years living in Massachusetts and surrounding areas, with two or more International Classification of Diseases codes for a rheumatic condition. Individuals were included if they received care between Jan 1, 2000, and May 1, 2023, at rheumatology practices affiliated with Mass General Brigham (Boston, MA, USA). Addresses were geocoded and overlaid with 1930s Home Owners' Loan Corporation (HOLC) redlining files. The Index of Concentration at the Extremes (ICE) for combined racial and income polarisation was constructed from US Census data. We used multilevel, multinomial logistic regression models to examine the odds of health-care utilisation separately by historical HOLC grade (A [best] to D [hazardous]) and ICE quintile (most deprived [1] to most privileged [5] race and income), adjusting for demographics, insurance, and comorbidities. People with lived experience of a rheumatic condition were not involved in the design or implementation of this study., Findings: The cohort comprised 5597 individuals; 3944 (70·5%) of 5597 patients were female, 1653 (29·5%) were male, 657 (11·7%) were Black, 224 (4·0%) were Hispanic, and the median age was 63 (50-73) years. 1295 (23·1%) of 5597 individuals lived in the most historically redlined areas (HOLC D) and 1780 (31·8%) lived in areas with the most concentrated present-day racialised economic deprivation (ICE quintile 1). Individuals in historically redlined areas (HOLC D) had greater odds of having four or more missed appointments (odds ratio [OR] 1·78 [95% CI 1·21-2·61]; p=0·0033) and of three or more emergency department visits (2·69 [1·48-4·89]; p=0·0011) compared with those in the most desirable neighbourhoods (HOLC A). Individuals in areas with highly concentrated racial and economic deprivation (ICE quintile 1) had greater odds of four or more missed appointments (OR 2·11 [95% CI 1·65-2·71]; p<0·0001) and of three or more emergency department visits (2·97 [2·02-4·35]; p<0·0001) versus those in areas with highly concentrated privilege (ICE quintile 5)., Interpretation: Historical redlining could be a structural determinant of inequities in present-day health-care utilisation patterns. Policy interventions that dismantle structural racism could reduce inequities in access to care for individuals with rheumatic conditions., Funding: Bristol Myers Squibb Foundation., Competing Interests: Declaration of Interests CHF and LS received grant support to their institution from the Bristol Myers Squibb Foundation for this work. CHF has also received grants to her institution from the National Institutes of Health and the Arthritis Foundation unrelated to this work. CHF has consulted for OM1, Harvard Pilgrim, Bain Capital LP, the University of Alabama, the Lupus Foundation of America, and the American College of Rheumatology (ACR); received honoraria from the University of Alabama, Birmingham and Northwestern University for lectures; serves on the ACR Diversity, Equity and Inclusion (DEI) committee, the Arthritis Foundation DEI Task Force, the Lupus Foundation of America Medical-Scientific Advisory Board, the Editorial Board of Arthritis Care & Research and as an Associate Editor of Lupus Science & Medicine and past Associate Editor of ACR Open Rheumatology. The other authors declare no competing interests., (Copyright © 2024 Elsevier Ltd. All rights reserved, including those for text and data mining, AI training, and similar technologies.)