Bridging Health Disparities in the Data-Driven World of Artificial Intelligence: A Narrative Review.

BACKGROUND: Artificial intelligence (AI) holds exciting potential to revolutionize healthcare delivery in the United States. However, there are concerns about its potential to perpetuate disparities among historically marginalized populations. OBJECTIVE: Following the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses, we conducted a narrative review of current literature on AI and health disparities in the United States. We aimed to answer the question, Does AI have the potential to reduce or eliminate health disparities, or will its use further exacerbate these disparities? METHODS: We searched the Ovid MEDLINE electronic database to identify and retrieve publications discussing AI and its impact on racial/ethnic health disparities. Articles were included if they discussed AI as a tool to mitigate racial health disparities with or without bias in developing and using AI. RESULTS: This review included 65 articles. We identified six themes of limitations in AI that impact health equity: (1) biases in AI can perpetuate and exacerbate racial and ethnic inequities; (2) equity in algorithms should be a priority; (3) lack of diversity in the field of AI is concerning; (4) the need for regulation and testing algorithms for accuracy; (5) ethical standards for AI in health care are needed; and (6) the importance of promoting transparency and accountability. CONCLUSIONS: While AI promises to enhance healthcare outcomes and address equity concerns, risks and challenges are associated with its implementation. To maximize the use of AI, it must be approached with an equity lens during all phases of development.

Willingness to participate in various nontherapeutic cancer research activities among urban and rural African American and Latinx healthy volunteers.

PURPOSE: Inclusion of racial/ethnic minorities in cancer research can reduce disparities in health outcomes; however, data regarding barriers and motivators to participation are sparse. This study assessed African American (AA) and Latinx healthy volunteers' perspectives regarding willingness to participate in noninvasive and invasive research activities. METHODS: Using a 38-item questionnaire adapted from the Tuskegee Legacy Project Questionnaire, we assessed willingness to participate in 12 research activities, offering 27 possible barriers and 14 motivators. The sample was segmented into four subgroups by AA/Latinx and rural/urban. RESULTS: Across five states and Puerto Rico, 533 participants completed questionnaires. Overall, participants were more willing to participate in noninvasive versus invasive procedures, although, all subgroups were willing to participate in research if asked. Rural AA were most willing to complete a survey or saliva sample, while rural Latinx were least willing. Urban AA were least willing to provide cheek swab, while rural counterparts were most willing. Self-benefit and benefit to others were among the top three motivators for all subgroups. Curiosity was a primary motivator for urban AA while obtaining health information motivated rural Latinx. Primary barriers included fears of side effects and being experimented on, lack of information, and lack of confidentiality. CONCLUSIONS: Latinx and AAs are willing to participate in the continuum of nontherapeutic research activities suggesting their lack of participation may be related to not being asked. Inclusive enrollment may be achieved by assessing needs of participants during the design phase of a study in order to reduce barriers to participation.

Predictors of Cancer Screening Among Culturally Diverse Men.

Men have higher rates of all cancers and are more likely to die from cancer than women; however, men are less likely to utilize disease prevention services. African American/Black men and Hispanic men have lower cancer survival rates and are less likely to utilize health care services than non-Hispanic White men. The present study examined demographic variables (age, household income, education, marital status, race/ethnicity, health insurance status), motivators to engage in healthy eating, and motivators to engage in physical activity as predictors of culturally diverse, medically underserved men's likelihood of getting a cancer screening (a) at the present time, (b) if no cancer symptoms are present, and (c) if a doctor discovers some cancer symptoms. Analyses were conducted using data from 243 men (47.3% non-Hispanic Black, 29.5% Hispanic, 16.5% non-Hispanic White, and 6.8% "other") recruited at the Men's Health Forum in Tampa, Florida. Age, having a medical or health condition that benefits from eating healthy, and having a commitment to physical activity were significant positive predictors of the likelihood of receiving a cancer screening. Motivation to engage in physical activity because of a personal priority was a significant negative predictor of the likelihood of getting a cancer screening. The findings from this study suggest that interventions to increase cancer screenings among culturally diverse, medically underserved men should be informed at least in part by an assessment of participating men's motivators for engaging in health promoting lifestyle behaviors such as physical activity and healthy eating.

Treatment delays, race, and outcomes in head and neck cancer.

PURPOSE: Patient race has been shown to predict for differences in outcomes and has been attributed to socioeconomic factors such as social support and access to healthcare. In head and neck cancer (HNC), a disease without recommended screening, we sought to investigate the association between race, treatment delays and outcome. METHODS: Records of 1802 patients with non-metastatic squamous cell HNC treated between 1998 and 2013 were retrospectively assessed from an institutional database. Patient demographics, tumor and treatment characteristics, and patient outcomes were abstracted from the chart. Differences between groups were assessed via logistic regression multivariate analysis (MVA). Outcomes including locoregional control (LRC) and overall survival (OS) were then estimated via Kaplan-Meier and Cox-regression MVA. RESULTS: Median follow up was 34 months. Patient races included white (n=1671, 93%), black (n=80, 4%), Asian (n=18, 1%), and other (n=33, 2%). On logistic regression MVA, Black patients were less likely to be married (39% vs. 63%; OR 0.5 95%CI 0.30-0.83, p=0.007) or be currently employed (43% vs. 61%; OR 0.44 95%CI 0.26-0.74, p=0.002) when compared to non-blacks. Black patients were also younger (54 vs. 59 years, p=0.001), more likely to present with advanced tumor stage (T4: 48% vs. 25%), and more often had >45days elapsed from diagnosis to treatment initiation (DTI) (61% vs. 49%, p=0.028). Delays in treatment, such as delayed diagnosis (advanced disease presentation) and delays in DTI>45days were also associated with marital and employment status. Black patients were associated with a lower 3-year LRC rate (65% vs. 81%, p<0.001) and OS rate (43% vs. 69%, p<0.001), compared to non-black patients. Patients with >45days DTI had a detriment in 3-year LRC (77% vs. 83%, p=0.002) and OS (66% vs. 69%, p=0.009). On Cox MVA, black race was independently prognostic for worse LRC (HR 1.62 95%CI 1.04-2.51, p=0.033) and OS (HR 1.55 95%CI 1.15-2.08, p=0.004) vs. non-blacks. CONCLUSION: Black race is independently prognostic for LRC and OS. Delays in HNC treatment, such as more advanced tumor stage presentation and delays in treatment initiation, may be attributed to socioeconomic factors such as employment status and social support. Efforts to accommodate these factors may expedite treatment, in hopes of improving the race related outcome disparity in HNC.

A Community-Level Assessment of Barriers to Preventive Health Behaviors Among Culturally Diverse Men.

There are significant gender disparities in health outcomes and health care utilization in the United States, with men experiencing more of these disparities. It is critical to ascertain the interplay between societal conditions, health behaviors, and access to services and the impact of these factors on health outcomes and utilization of health care. The present study is part of a larger initiative titled, The Men's Health Study: Addressing Healthy Lifestyle Behaviors, which has two purposes-to annually assess the motivators of and barriers to health-promoting behaviors among culturally diverse men attending the Men's Health Forum (MHF) and to use this information to develop an intervention program that facilitates healthy lifestyle behaviors among men. The MHF is a community-driven initiative for medically underserved men in Tampa, Florida that offers free health screenings and wellness exhibitors in order to empower men to lead a healthy lifestyle. The purpose of this article is to identify barriers to engaging in health-smart behaviors (e.g., cancer screenings, physical activity) among culturally diverse men who participated in the MHF and to detect any demographic differences among these barriers. A total of 254 men participated in the study. Findings identify that age was the only demographic variable that had a statistically significant association with any of the cancer-screening barriers. Some cancer-screening barriers appear to exist among all demographic groups since no statistical demographic differences were discovered. Income and education were significantly associated with barriers to engaging in health-smart behaviors. This may give researchers, health educators, and providers information needed to customize interventions to promote health and preventive health care among culturally diverse men.

Cancer and lesbian, gay, bisexual, transgender/transsexual, and queer/questioning (LGBTQ) populations.

This article provides an overview of the current literature on seven cancer sites that may disproportionately affect lesbian, gay, bisexual, transgender/transsexual, and queer/questioning (LGBTQ) populations. For each cancer site, the authors present and discuss the descriptive statistics, primary prevention, secondary prevention and preclinical disease, tertiary prevention and late-stage disease, and clinical implications. Finally, an overview of psychosocial factors related to cancer survivorship is offered as well as strategies for improving access to care.

Disparities in survival by insurance status in patients with Hodgkin lymphoma.

BACKGROUND: The association between insurance status and outcomes has not been well established for patients with Hodgkin lymphoma (HL). The purpose of this study was to examine the disparities in overall survival (OS) by insurance status in a large cohort of patients with HL. METHODS: The National Cancer Data Base (NCDB) was used to evaluate patients with stage I to IV HL from 1998 to 2011. The association between insurance status, covariables, and outcomes was assessed in a multivariate Cox proportional hazards model. Survival was estimated with the Kaplan-Meier method. RESULTS: Among the 76,681 patients within the NCDB, 45,777 patients with stage I to IV HL were eligible for this study (median follow-up, 6.0 years). The median age was 39 years (range, 18-90 years). The insurance status was as follows: 3247 (7.1%) were uninsured, 7962 (17.4%) had Medicaid, 30,334 (66.3%) had private insurance, 3746 (8.2%) had managed care, and 488 (1.1%) had Medicare. Patients with an unfavorable insurance status (Medicaid/uninsured) were at a more advanced stage, had higher comorbidity scores, had B symptoms, and were in a lower income/education quartile (all P < .01). These patients were less likely to receive radiotherapy and start chemotherapy promptly and were less commonly treated at academic/research centers (all P < .01). Patients with unfavorable insurance had a 5-year OS of 54% versus 87% for those favorably insured (P < .01). When adjustments were made for covariates, an unfavorable insurance status was associated with significantly decreased OS (hazard ratio, 1.60; 95% confidence interval, 1.34-1.91; P < .01). The unfavorable insurance status rate increased from 22.8% to 28.8% between 1998 and 2011. CONCLUSIONS: This study reveals that HL patients with Medicaid and uninsured patients have outcomes inferior to those of patients with more favorable insurance. Targeting this subset of patients with limited access to care may help to improve outcomes. Cancer 2015;121:3435-43. © 2015 American Cancer Society.

Lesbian, Gay, Bisexual, Transgender, Queer/Questioning (LGBTQ) Perceptions and Health Care Experiences.

BACKGROUND: The goal study of this was to explore attitudes, health knowledge, and experiences with healthcare setting and providers among gay, lesbian, bisexual, transgender, queer/questioning (GLBTQ) individuals and to identify areas for improvement. METHODS: Members of Equality Florida™ residing in the five counties of the Tampa Bay region were recruited through email invitation to complete a 60-item questionnaire assessing demographics, attitudes, and experiences with healthcare providers (HCPs). Additional open-ended questions focused on experiences with HCPs and suggestions for ways to improve HCPs' cultural competency. RESULTS: 632 respondents completed the survey of which 41% were gay men and 29% were lesbian. The majority of participants was White, non-Hispanic (93%), married/partnered (78%), and had health insurance (88%). The majority (67%) reported they always or often disclosed their sexual orientation/identity to an HCP and few had negative reactions in the healthcare setting (<10%). Healthcare settings with equality signs and gender-neutral language were perceived as safer. Participants' responses suggested need for policy changes and improved cultural competence among HCPs. CONCLUSION: Results show high rates of sexual orientation disclosure, greater acceptance from providers of GLBTQ status, and the need for examination of hospital policies and improved cultural competency.

The Yo me cuido® Program: Addressing Breast Cancer Screening and Prevention Among Hispanic Women.

Breast cancer is less likely to be diagnosed at the earliest stage in Hispanic/Latino (Hispanic) women compared to non-Hispanic White women, even after accounting for differences in age, socioeconomic status, and method of detection. Moffitt Cancer Center created a comprehensive health education program called Yo me cuido (®) (YMC) to address and reduce breast cancer disparities among Spanish- and English-speaking Hispanic women by providing breast cancer and healthy lifestyles awareness and education, and promoting breast cancer screenings, reminders, and referrals for women 40 years and older. The purpose of this paper is to showcase the innovative approaches and methods to cancer prevention and early detection of the YMC program, and to promote it as an effective tool for improving outcomes in community health education, outreach, and engagement activities with Hispanic populations. Key components of the program include educational workshops, mammogram referrals, and a multimedia campaign. The YMC program is unique because of its approaches in reaching the Hispanic population, such as delivering the program with compassionate services to empower participants to live a healthier lifestyle. Additionally, direct follow-up for mammography screenings is provided by program staff. From 2011 to 2013, YMC has educated 2,226 women and 165 men through 93 workshops. About 684 (52 %) women ages 40 and older have had a screening mammogram within their first year of participating in the program. The YMC program is an innovative cancer education and outreach program that has demonstrated a positive impact on the lives of the Hispanic community in the Tampa Bay region.

Affecting African American men's prostate cancer screening decision-making through a mobile tablet-mediated intervention.

African American men experience a 60% higher incidence of prostate cancer and are more than twice as likely to die from it than White men. Evidence is insufficient to conclude that definitively screening for prostate cancer reduces the likelihood of morbidity or death. Patients are encouraged to discuss screening alternatives with health care providers for informed decision-making (IDM). The extent of IDM in clinical or community setting is not known. This study uses data from a community-based, computer-mediated, IDM intervention that targeted 152 African American aged 40 to 70. Pretest-posttest differences in means for prostate cancer knowledge, screening decisional conflict, and screening decisional self-efficacy were examined by two-tailed t-tests. Overall, the intervention significantly improved respondents' prostate cancer knowledge (p<.0001), significantly improved decisional self-efficacy (p<.0001) and significantly reduced decisional conflict (p<.0001). Specifically, the intervention significantly promoted IDM among men who reported more education, being married, having financial resources, and younger age.

Assessing needs and assets for building a regional network infrastructure to reduce cancer related health disparities.

Significant cancer health disparities exist in the United States and Puerto Rico. While numerous initiatives have been implemented to reduce cancer disparities, regional coordination of these efforts between institutions is often limited. To address cancer health disparities nation-wide, a series of regional transdisciplinary networks through the Geographic Management Program (GMaP) and the Minority Biospecimen/Biobanking Geographic Management Program (BMaP) were established in six regions across the country. This paper describes the development of the Region 3 GMaP/BMaP network composed of over 100 investigators from nine institutions in five Southeastern states and Puerto Rico to develop a state-of-the-art network for cancer health disparities research and training. We describe a series of partnership activities that led to the formation of the infrastructure for this network, recount the participatory processes utilized to develop and implement a needs and assets assessment and implementation plan, and describe our approach to data collection. Completion, by all nine institutions, of the needs and assets assessment resulted in several beneficial outcomes for Region 3 GMaP/BMaP. This network entails ongoing commitment from the institutions and institutional leaders, continuous participatory and engagement activities, and effective coordination and communication centered on team science goals.

Establishing the infrastructure to comprehensively address cancer disparities: a model for transdisciplinary approaches.

The Center for Equal Health (CEH), a transdisciplinary Center of Excellence, was established to investigate cancer disparities comprehensively and achieve health equity through research, education, training, and community outreach. This paper discusses challenges faced by CEH, strategies employed to foster collaborations, lessons learned, and future considerations for establishing similar initiatives.

Empowering underserved populations through cancer prevention and early detection.

It is well documented that cancer is disproportionately distributed in racial/ethnic minority groups and medically underserved communities. In addition, cancer prevention and early detection represent the key defenses to combat cancer. The purpose of this article is to showcase the comprehensive health education and community outreach activities at the H. Lee Moffitt Cancer Center and Research Institute (Moffitt) designed to promote and increase access to and utilization of prevention and early detection services among underserved populations. One of Moffitt's most important conduits for cancer prevention and early detection among underserved populations is through its community education and outreach initiatives, in particular, the Moffitt Program for Outreach Wellness Education and Resources (M-POWER). M-POWER works to empower underserved populations to make positive health choices and increase screening behaviors through strengthening collaboration and partnerships, providing community-based health education/promotion, and increasing access to care. Effective, empowering, and culturally and linguistically competent health education and community outreach, is key to opening the often impenetrable doors of cancer prevention and early detection to this society's most vulnerable populations.

Establishing the infrastructure to conduct comparative effectiveness research toward the elimination of disparities: a community-based participatory research framework.

In Tampa, Florida, researchers have partnered with community- and faith-based organizations to create the Comparative Effectiveness Research for Eliminating Disparities (CERED) infrastructure. Grounded in community-based participatory research, CERED acts on multiple levels of society to enhance informed decision making (IDM) of prostate cancer screening among Black men. CERED investigators combined both comparative effectiveness research and community-based participatory research to design a trial examining the effectiveness of community health workers and a digitally enhanced patient decision aid to support IDM in community settings as compared with "usual care" for prostate cancer screening. In addition, CERED researchers synthesized evidence through the development of systematic literature reviews analyzing the effectiveness of community health workers in changing knowledge, attitudes and behaviors of African American adults toward cancer prevention and education. An additional systematic review analyzed chemoprevention agents for prostate cancer as an emerging technique. Both of these reviews, and the comparative effectiveness trial supporting the IDM process, add to CERED's goal of providing evidence to eliminate cancer health disparities.

Racial/ethnic differences in cancer prevention beliefs: applying the health belief model framework.

PURPOSE: An understanding of each racial/ethnic group's beliefs about cancer prevention is important for designing/implementing interventions to reduce cancer-health disparities. The Health Belief Model was used to examine racial/ethnic differences in beliefs about cancer and cancer prevention. DESIGN: The data were from the 2007 Health Information National Trends Survey, a biennial, cross-sectional survey using a random-digit-dial telephone frame and a mailing address frame. SETTING: A weighted, nationally representative sample of American adults. SUBJECTS: The sample consisted of 7452 individuals. MEASURES: Model construct variables (perceived susceptibility; perceived severity; perceived benefits; perceived barriers; cues to action; self-efficacy) and race/ethnicity were assessed. ANALYSIS: The Rao-Scott χ(2) test and multivariate logistic regression assessed racial/ethnic differences. RESULTS: The constructs self-efficacy, perceived benefits, and perceived susceptibility were significantly associated with race/ethnicity. The remaining three constructs were not statistically significant. Multivariate analysis revealed Hispanics were less likely to believe they could lower their chances of getting cancer than did African-Americans and whites. Hispanics, Asians, and African-Americans were more likely to believe they had a lower chance of getting cancer in the future than did whites. CONCLUSION: Culturally relevant health education/promotion interventions need to be developed and tailored to (1) empower Hispanics regarding their ability to prevent cancer and (2) educate racial/ethnic minorities about their susceptibility and risk perception for cancer.

Designing a community-based lay health advisor training curriculum to address cancer health disparities.

INTRODUCTION: Racial and ethnic minorities have disproportionately higher cancer incidence and mortality than their White counterparts. In response to this inequity in cancer prevention and care, community-based lay health advisors (LHAs) may be suited to deliver effective, culturally relevant, quality cancer education, prevention/screening, and early detection services for underserved populations. APPROACH AND STRATEGIES: Consistent with key tenets of community-based participatory research (CBPR), this project engaged community partners to develop and implement a unique LHA training curriculum to address cancer health disparities among medically underserved communities in a tricounty area. Seven phases of curriculum development went into designing a final seven-module LHA curriculum. In keeping with principles of CBPR and community engagement, academic-community partners and LHAs themselves were involved at all phases to ensure the needs of academic and community partners were mutually addressed in development and implementation of the LHA program. DISCUSSION AND CONCLUSIONS: Community-based LHA programs for outreach, education, and promotion of cancer screening and early detection, are ideal for addressing cancer health disparities in access and quality care. When community-based LHAs are appropriately recruited, trained, and located in communities, they provide unique opportunities to link, bridge, and facilitate quality cancer education, services, and research.

High grade prostate intraepithelial neoplasia (PIN) is a PSA-independent risk factor for prostate cancer in African American men: results from a pilot study.

African American men (AAM) demonstrate increased prostate cancer incidence and mortality rates. We investigated known prostate cancer risk factors in AAM. Prostate specific antigen (PSA) and diagnosis of high grade prostatic intraepithelial neoplasia (PIN) were significant prostate cancer predictors. However, even including AAM with low PSA (<4ng/ml), those with PIN had significantly elevated risk, compared to men without PIN (83.3% vs. 6.9%, p<0.0001). In AAM diagnosed with PIN, PSA level was no longer significant (83.3% vs. 92.3%, p=0.593 respectively). Our results suggest that a history of PIN is highly predictive of prostate cancer in AAM, and help provide PSA-independent venues for screening.