Indigenous Data Aggregation Perpetuates Structural Racism

FormalPara Key Points

• Aggregating Indigenous Native Hawaiian and Other Pacific Islander (NHPI) populations with Asian populations masks key NHPI disparities.

• NHPI patients have a higher comorbidity burden and worse survival rates compared to Asian patients.

• Indigenous erasure stifles NHPI research, funding, and public health initiatives.

• Medical researchers should adopt best research practices for NHPI health data collection.

In the United States (USA), the term “Asian Americans and Pacific Islanders” (AAPI) permeates both everyday conversation and mainstream media [1]. Initially established to foster solidarity between two distinct marginalized groups, the term “AAPI” has been widely scrutinized in recent decades [1, 2]. Critics argue that aggregation of the majority Asian American group, which itself is heterogenous, overshadows crucial health outcomes of the frequently excluded Native Hawaiian or Other Pacific Islander (NHPI) population, the Indigenous peoples of the Pacific Islands [2]. While Asian Americans trace their ancestry to over 30 countries across South, East, and Southeast Asia, Pacific Islanders share an Oceanic wayfinding heritage across over 20 island nations and territories in Melanesia, Micronesia, and Polynesia.

In this chapter, we explore the ways that data aggregation negatively impacts Indigenous Pacific Islander communities in the USA and discuss how this form of Indigenous erasure perpetuates structural racism, with a specific emphasis on cancer data. First, we review Indigenous Pacific Islander history in the context of US colonial and imperial ties (Fig. 14.1).

Fig. 14.1

Historical timeline of the NHPI community in the USA, with key time points informing how Indigenous NHPI race terminology came into existence and advocacy efforts for data disaggregation. (Abbreviations: AI/AN American Indian and Alaska Native, API Asian and Pacific Islander, NH Native Hawaiian, NHPI Native Hawaiian or Other Pacific Islander)

In the 1830s, White businessmen arrived in the Kingdom of Hawaiʻi to establish profitable plantations [3]. Economic interests pushed these businessmen to seize power from Indigenous rulers and urged the US government to do the same [3]. By 1893, Queen Liliʻuokalani was forced to abdicate her throne, a direct violation of international humanitarian law [3]. In the coming decades, the Native Hawaiian monarchy was dismantled, the region annexed, and the territory given its statehood [3]. For generations, the Indigenous Native Hawaiian identity was suppressed. US colonialism and imperialism occurred in a similar vein throughout Oceania, in Palau, the Marshall Islands, the Northern Mariåna Islands, Guåhan, the Federated States of Micronesia, and American Samoa [4]. Today, some of these islands, particularly the US Pacific territories, are still under direct colonial rule [4]. These relationships of colonialism, imperialism, and militarization have widely suppressed Indigenous sovereignty across the Pacific.

In the USA, the term “Asian American” likely originated in the midst of the Black Power and American Indian movements of the 1960s [1]. Emma Gee and Yuji Ichioka at the University of California, Berkeley, are credited with creating the term by naming their group the Asian American Political Alliance (AAPA) [1]. The AAPA aimed to create unity by cultivating a pan-Asian identity [1]. Pacific Islanders initially aligned themselves with this movement, but struggled to commit to this racial identity given cultural and historical differences [1]. Nearly a decade later, the US government sought to address the longstanding use of broad, non-standardized race and ethnic categories used in federal data collection [5]. In 1977, the USA created four federally defined US racial categories: American Indian and Alaska Native (AI/AN), Asian and Pacific Islander (API), Black, and White [5]. The aggregated API racial category was first used in the US Census in the 1980s [1].

On July 9, 1997, a Federal Register Notice was issued with racial category recommendations from the Federal Interagency Committee. These included retaining the API category as an aggregated group [6]. The Native Hawaiian congressional delegation, 7000 postcard signees, and the Hawaiʻi legislature all opposed this recommendation. Native Hawaiian advocates wished to combine the Native Hawaiian group with the AI/AN category, expressing solidarity as fellow US-occupied Indigenous populations [6]. However, American Indian Tribal governments disagreed, arguing that aggregation would hinder informative data collection and program administration for AI/AN communities [6]. Native Hawaiian advocates similarly argued that the aggregated API category inhibited data collection and monitoring of NHPI communities [6]. On October 30, 1997, the federal government reported that, while Native Hawaiian advocates initially wished to stand in solidarity with AI/AN, creating a new racial category was a viable option [6]. Thus, the federal government scrapped the API category and created separate Asian and NHPI categories [6]. This established the five federally defined racial categories used in the US today: AI/AN, Asian, Black or African American, NHPI, and White [6]. The disaggregated Asian and NHPI categories were first used in the 2000 Census [1]. Federal programs were expected to comply with these five revised racial categories by 2003 [6].

Unfortunately, widespread awareness has yet to be achieved. Government officials, medical experts, and researchers continue to aggregate NHPI and Asian populations [2, 7]. Advocates argue that NHPI aggregation is structural racism, defined as society’s promotion of racial discrimination through institutions, ideas, and processes [2]. Specifically, data aggregation hides and perpetuates NHPI health issues [2, 7].

Data Disaggregation Hinders Research, Public Health Action, and Funding

The aggregation of NHPI data with those of unrelated non-Indigenous groups perpetuates structural barriers that hinder NHPI health equity. This section explores barriers in research, public health, and funding. Through four case studies, we explore ways in which Indigenous erasure contributes to structurally racist practices.

Data Aggregation and the Impact on Cancer Research

Recent advances in cancer research have come about due to the tireless work of NHPI champions advocating for NHPI data inclusion [7,8,9]. Here, we review two studies that report clinically important differences between Asian and NHPI patients with cancer.

Data Disaggregation in Cancer Research: Case Study 1

In research published in 2023, Taparra et al. investigated predictors of radiation therapy refusal among patients with the top 10 cancers in the USA who were recommended for radiation therapy treatment by an oncologist [8]. Specifically, their analysis stratified the results by race. Earlier studies identified that radiation therapy refusal when recommended by an oncologist was associated with a doubled likelihood of cancer mortality. However, none of these studies had included all five federally defined races including the NHPI population, exemplifying the practice of Indigenous erasure. This Indigenous knowledge gap among NHPI patients with cancer is significant given the high rates of cancer among this population.

The authors compared the rate at which patients refused radiation therapy from 2004 to 2016 by race in a national sample of patients across the country. Compared to all racial groups, NHPI patients with cancer had the highest rates of refusing oncologist-recommended radiation therapy and the largest increase of refusal over the course of the study. Indigenous NHPI and AI/AN patients were significantly more likely to refuse radiation therapy when compared to non-Hispanic White patients. No significant difference in treatment refusal rates was found among Asian patients compared to White patients. In addition, greater comorbidity burden was associated with increased radiation therapy refusal likelihood in NHPI, White, Asian, and Black patients.

While it is important that disparities are reported, they must also be contextualized by Indigenous scholars. While those outside the NHPI community may blame “lack of education” for radiation therapy refusal, the authors of this study underscore the impact of possible multigenerational historical trauma on NHPI radiation refusal. They propose that the history of atomic bomb detonations in the Pacific Islands may contribute to NHPI attitudes. During the Cold War, the US military test-detonated the equivalent of 7200 Hiroshima-sized bombs in the Marshall Islands. These events exposed Indigenous Marshallese communities to dangerous amounts of radiation and forced them to evacuate, all in violation of their human rights. This highlights how historical human rights violations might impact current NHPI cancer treatments. Moreover, it underscores the importance of Indigenous narratives to contextualizing health data.

In a follow-up study disaggregating Asian and NHPI patients, the same authors found that Japanese patients with cancer were not any more or less likely to refuse radiation therapy compared to other East Asian patients, suggesting that this history of atomic bomb detonation may be more specific to the experiences of the NHPI community [9]. Together, these studies provide an example of how cancer treatment decision-making by NHPI patients may differ from that of Asian patients, including Japanese patients whose country experienced the effects of atomic bomb use. Importantly, the authors provide historical context that may contribute to understanding of radiation therapy refusal.

Data Disaggregation in Cancer Research: Case Study 2

In a 2022 study, Taparra et al. reported disparities in survival and comorbidity burden between disaggregated Asian and NHPI patients with cancer [7]. They examined almost 6 million patients with nine of the most common cancers in the USA, with data disaggregated into East Asian, South Asian, Southeast Asian, NHPI, and non-Hispanic White patients.

The authors evaluated comorbidity burden (Charlson-Deyo comorbidity index) and survival outcomes for each of the disaggregated categories and subcategories. All previous studies of this nature had aggregated or omitted the Indigenous NHPI population. However, Taparra et al. found that NHPI patients with cancer have the highest comorbidity burden of all racial groups, with the comorbidity burden statistically higher than that of White and Asian patients. East Asian patients had significantly lower comorbidity burden. For most of the included cancers, Asian patients had improved overall survival outcomes compared to White patients, while NHPI patients had significantly worse overall survival outcomes for the majority of included cancers. Even after controlling for the significantly higher comorbidity burden among the NHPI population, NHPI patients still had significantly worse survival outcomes, suggesting multifactorial genetic and social influences.

These stark findings indicate that current aggregation practices give physicians, policy makers, and patients the false and damaging impression that NHPI patients with cancer have lower comorbidities and more positive outcomes than is the case. Many oncology clinical trials, including those supported by pharmaceutical companies, include strict performance status or comorbidity burden inclusion criteria. The authors suggest that comorbidity burden may play a role in the disproportionate exclusion of NHPI patients from oncology clinical trials.

Data Aggregation and the Impact on Public Health

There is an unmet need for cancer-related public health efforts that focus on the NHPI community. In this section, we review a relevant and contemporary example from the COVID-19 pandemic.

Leveraging Public Health Data: Case Study 3

Kamaka et al. describe how the NHPI data disaggregation movement informed public health efforts during the COVID-19 pandemic [10]. The first case of COVID-19 was reported in the USA in January 2020. By March of that year, the USA was in a state of national emergency. In April, the Centers for Disease Control and Prevention (CDC) excluded the federally recognized NHPI racial category from COVID-19 data reports stratified by race. However, by May 2020, NHPI patients had the highest case rates of COVID-19 compared to other races in 8 out of the 10 states with disaggregated data.

In May 2020, the Hawaiʻi NHPI COVID-19 Response, Recovery, Resiliency (3R) Team was formed. Within a year, the 3R Team successfully collaborated with the Hawaiʻi State Department of Health (HSDH) to address barriers to state-level data disaggregation through a public health lens. With these changes, the HSDH showed that high case rates were largely driven by other Pacific Islander communities, and not necessarily Native Hawaiian communities alone. Overall, COVID-19 remains a key example of a public health effort that has identified Indigenous populations in need.

Data Aggregation and the Impact on Federal Funding

Federal agencies can only fund solutions to identified problems. Since much NHPI health data is aggregated, issues of NHPI health are more likely to persist. Policy makers must decide between funding projects with convincing data or NHPI proposals that lack the same level of quality data. Thus, the realities of Indigenous NHPI people are overlooked and underfunded. Here, we review an example of long-term federal under-investment in the NHPI community.

Examining Federal Funding Trends: Case Study 4

A growing spotlight on the importance of NHPI research prompted scientists to investigate National Institutes of Health (NIH) budget allocation to Asian and NHPI population research [11]. Đoàn et al. examined NIH-funded extramural projects from 1992 to 2018 and found that only 0.17% of the total NIH budget was allocated to studies that focused on Asian American (AA) and NHPI health [11]. At first glance, a significant increase in total budget allocation pre- and post-2000 appeared promising. However, deeper analysis shows that the 0.12% pre-2000 allocation increased to just 0.18% post-2000. Indeed, the increased total NIH budget over this time was not associated with increases in AA and NHPI budget allocation. This demonstrates the NIH’s lack of prioritization of Asian and NHPI communities, despite overall federal health budget increases.

Notably, the authors found that while Asian clinical trial participation increased from 2011 to 2016, NHPI participation decreased. When disaggregated from Asian participant data, almost no NHPI participants were included in the grants the authors examined. In the uphill battle for NHPI data disaggregation, support from the research community is imperative. Đoàn et al. exemplify this allyship with their awareness, inclusivity, and caring tone.

Conclusions and Best Practices

Indigenous erasure via NHPI data omission is harmful and masks health disparities. Researchers should critically evaluate current data reporting methods and adopt best practices to collect, disaggregate, and report NHPI data (Fig. 14.2). Clinical researchers should consider data oversampling to enable valid analysis that reflects the state of NHPI health [12]. Community collaboration is critical to building trust, facilitating participation, and honoring context for the communities the research intends to highlight [8, 10, 12]. Increased language translation would expand accessibility for those more comfortable communicating in Indigenous languages [12]. Demographic survey questions should include the option to specify multiple races given that most NHPI people identify as having multiple racial backgrounds [12].

Fig. 14.2

Recommended NHPI patient data reporting practices

Researchers should use the critical framework of Indigeneity during study design, including taking key steps such as the formation of robust community advisory boards [12]. Results should be socially and culturally contextualized to better capture the nuance of NHPI health outcomes [8]. Importantly, colonialism, imperialism, and globalization throughout Oceania should be factors included in the evaluation of NHPI inequities [2, 12]. It is broadly recommended that researchers reflect on their position relative to the communities they study [12]. Finally, scientists and institutions should empower Indigenous researchers and their sustainability through meaningful allyship [10, 12].

In embracing these inclusive and culturally sensitive data practices, researchers hold the power to illuminate the path toward a brighter future for our Indigenous NHPI communities, where health disparities are recognized and addressed, where research is guided by community-rooted collaboration, and where Indigenous voices are uplifted, fostering hope and resilience for generations to come.

The authors acknowledge that KT was supported by the Stanford Cancer Institute, an NCI-designated Comprehensive Cancer Center. KT was funded by a Stanford Cancer Institute Women’s Cancer Center Innovation Award and the Stanford Cancer Institute Fellowship Award. KT was also funded by the American Society for Clinical Oncology Dr. Judith and Alan Kaur Endowed Young Investigator Award.