Abstract A021: Cancer disparities among Muslim women: Population identification by name

Background: A total of 3.5 million American Muslims live in the United States (US), representing 1% of the total population. Over 50% of American Muslims are foreign-born [1]. Racially, American Muslims are very diverse religious groups in the US with no majority race, split as 25% Black, 24% White, 18% Arab, 18% Asian, 7% mixed race, and 5% Hispanic [2]. In terms of origin, Middle Eastern Countries represent 41% of the entire Muslim community in the US [3]. In the last two decades, the population growth trajectory of the American Muslim community has been doubling; the projection was 2.6 million (0.8 %) in 2010 and is expected to be 6.2 million (1.7%) in 2030 [4]. Virginia has the second-largest Muslim population in the US by the percentage of the population, with an estimated 2.7% [2]. Much of that population is concentrated in Northern Virginia and the Richmond Area [5]. Our previous work showed that Muslim women seek cancer care at an advanced stage. However, we had a limitation in identifying Muslim women in the VCU health system as most Muslims do not share their religion at registration [6]. We created an algorithm to identify Muslim patients by name to address this issue. Methods: Identifying patient populations by name has been done by the North American Association of Central Cancer Registries (NAACCR) to better identify Hispanic populations. Their algorithm uses family names and county of residence information [7]. Our algorithm uses Names only at this time. We used a collection of Middle-Eastern names from a publicly available database [8]. We excluded names frequently used in the United States; examples include Maria, Sarah, and Daniel. Then we used the Names in the database to heuristically identify patients with Middle-Eastern origin. Results: We identified 165 female Muslim cancer patients between 2012 and 2021 at VCU Health System, a 260% improvement compared to our previous work without the new identification algorithm. Similar to our previous work, the extended cohort shows that Muslim women are diagnosed with late-stage breast cancer more frequently than non-Muslim women (20% vs. 14%, p-value = 0.04). Conclusion: Our algorithm to identify Muslim patients made a difference in the patient cohort size, making it possible to do a more robust analysis and draw statistically significant conclusions. Future work: Our algorithm needs further improvement. Currently, we are focusing on Middle-Eastern names, we plan to be more inclusive with additional Muslim majority areas.

Citation Format: Tamas S. Gal, NourEldin Abosamak, Fariha Tariq, Nevena Skoro, Rana Ramadan, Asmaa Namoos. Cancer disparities among Muslim women: Population identification by name [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr A021.