Racial and ethnic disparities in aortic stenosis within a universal healthcare system characterized by natural language processing for targeted intervention

Abstract

Aims: Aortic stenosis (AS) is a condition marked by high morbidity and mortality in severe, symptomatic cases without intervention via transcatheter aortic valve implantation (TAVI) or surgical aortic valve replacement (SAVR). Racial and ethnic disparities in access to these treatments have been documented, particularly in North America, where socioeconomic factors such as health insurance confound analyses. This study evaluates disparities in AS management across racial and ethnic groups, accounting for socioeconomic deprivation, using an artificial intelligence (AI) framework.

Methods and results: We conducted a retrospective cohort study using a natural language processing pipeline to analyse both structured and unstructured data from > 1 million patients at a London hospital. Key variables included age, sex, self-reported race and ethnicity, AS severity, and socioeconomic status. The primary outcomes were rates of valvular intervention and all-cause mortality. Among 6967 patients with AS, Black patients were younger, more symptomatic, and more comorbid than White patients. Black patients with objective evidence of AS on echocardiography were less likely to receive a clinical diagnosis than White patients. In severe AS, TAVI and SAVR procedures were performed at lower rates among Black patients than among White patients, with a longer time to SAVR. In multivariate analysis of severe AS, controlling for socioeconomic status, Black patients experienced higher mortality (hazard ratio = 1.42, 95% confidence interval = 1.05-1.92, P = 0.02).

Conclusion: An AI framework characterizes racial and ethnic disparities in AS management, which persist in a universal healthcare system, highlighting targets for future healthcare interventions.

Keywords: Aortic stenosis; Artificial intelligence; Ethnicity; Health equity; Transcatheter aortic valve replacement.

Graphical Abstract

Figure 1

Cohort overview. (A) Analysis framework to examine race and ethnicity differences in aortic stenosis pathology and clinical pathways. (B) Electronic health record data-types. (C) CONSORT diagram of patient inclusion and exclusion criteria.

Figure 2

Aortic stenosis presentation. (A) Radar plots showing the prevalence of cardiac symptoms at aortic stenosis diagnosis, plotted separately by ethnicity. The symptoms categories are chest pain, breathlessness, palpitations, dizziness, presyncope, and syncope. The percentages for each symptom are displayed within the corresponding radar slice. (B) Forest plot showing the adjusted odds ratios for presenting with any cardiac symptoms at the point of diagnosis for each ethnicity group. A logistic regression model is adjusted for age, sex, aortic stenosis disease severity, and socioeconomic deprivation. Error bars represent the limits of the 95% confidence interval for the odds ratio.

Figure 3

Aortic stenosis diagnosis. (A) Line plot showing cumulative percentage of aortic stenosis clinical diagnosis in the first year after positive echocardiogram findings, stratified by race and ethnicity. (B) One-year rate (percentage) of aortic stenosis clinical diagnosis stratified by race and ethnicity. (C) Forest plot showing the adjusted odds ratios for aortic stenosis clinical diagnosis in the first year after positive echocardiogram findings for each ethnicity group. A logistic regression model is adjusted for age, sex, and aortic stenosis disease severity. Error bars represent the limits of the 95% confidence interval for the odds ratio.

Figure 4

Aortic stenosis intervention. (A) Bar plots showing the percentage of patients diagnosed with severe aortic stenosis receiving an intervention (transcatheter aortic valve implantation or surgical aortic valve replacement), stratified by ethnicity (left) or by deprivation quintile (right). (B) Time between severe aortic stenosis diagnosis and transcatheter aortic valve implantation, stratified by ethnicity. Bar plot (left) shows mean time difference. Forest plot (right) shows coefficients for a linear regression model adjusted for age, sex, and socioeconomic deprivation. (C) Time between severe aortic stenosis diagnosis and surgical aortic valve replacement, stratified by ethnicity. Bar plot (left) shows mean time difference. Forest plot (right) shows coefficients for a linear regression model adjusted for age, sex, and socioeconomic deprivation. Error bars for bar plots represent the SEM. Error bars for forest plots represent the limits of the 95% confidence interval for the model coefficient.

Figure 5

Aortic stenosis mortality. (A) Kaplan–Meier plot showing overall survival outcomes from severe aortic stenosis diagnosis stratified by intervention. (B) Kaplan–Meier plot showing overall survival outcomes from diagnosis of severe aortic stenosis stratified by ethnicity and intervention status. (C) Forest plot showing hazard ratios for multivariate Cox analysis of overall survival outcomes from diagnosis of severe aortic stenosis stratified by ethnicity, age, sex, and socioeconomic deprivation.

Figure 6

Aortic stenosis prevalence. (A) Pie charts showing aortic stenosis prevalence by race and ethnicity in >100 000 echocardiograms for unselected indication. Top row shows data for people over the age of 65 years; bottom row shows data for people below the age of 65 years.