Abstract
Background: Several cancer organizations have emphasized the need for etiologic and prognostic research targeting underserved populations. Most studies have operationalized the term underserved as uninsured or impoverished, but unknown is whether these characteristics sufficiently represent the underserved. Therefore, we aimed to assess whether conventional definitions of underserved adequately represent an underserved cancer population.
Methods: We used data from the JPS Center for Cancer Care institutional registry (accredited by the Commission on Cancer). This center is part of a public hospital network and is the primary source of care for underserved cancer patients in Tarrant County, Texas. We also obtained data for all cancer diagnoses in Tarrant County from the Texas Cancer Registry for comparison. Our eligible population included individuals aged ≥18 years diagnosed with a first primary cancer between 2008 and 2015. We applied conventional definitions of underserved to cancer patients in Tarrant County, where underserved was defined as uninsured or residence in a census tract where >20% of residents were below the poverty level (i.e., impoverished). We estimated ratios of relative frequency (RRF) and bootstrapped 95% confidence limits (CL) comparing sociodemographic and cancer characteristics between the conventionally defined population (i.e., uninsured or impoverished) and the JPS reference population. RRFs unequal to 1 suggested that the distribution of the characteristic differed between the conventionally defined population and the reference population.
Results: Our study population comprised 7,751 underserved cancer patients at JPS, of whom 82% were aged <65 years, 46% male, and 54% racial/ethnic minorities. The most common cancer diagnoses were female breast (14%), lung (12%), and colorectal (10%), and 34% were diagnosed with advanced-stage cancers. RRFs for sociodemographic characteristics ranged between 0.47 and 0.91 when comparing uninsured to the reference population. RRFs for sociodemographic characteristics ranged between 1.1 and 2.5 when comparing impoverished to the reference population. Similar departures from RRF=1 were observed for cancer characteristics.
Discussion: Our results suggest that conventional definitions of underserved may not adequately represent underserved cancer populations. The complex characteristics of underserved populations may be underappreciated. Public hospitals that form the health care safety net warrant greater attention for cancer research targeting the underserved.
Citation Format: Bradford E. Jackson, Yan Lu, Tracey Barnett, Bhavna Tanna, Bassam Ghabach, Rohit P. Ojha. The inadequacy of conventional definitions for representing underserved cancer patients [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr A095.
RSS Feeds