Limited Data Collection for LGBTQI+ Health Promotes Bias

Limited Data Collection for LGBTQI+ Health Promotes Bias

SkyInclusive questions about sexual orientation, gender identity, and intersex individuals (SOGII) help identify and address the diverse health needs of LGBTQI+ populations and mitigate health disparities, especially when co-analyzed with race, disability, language, behaviors and other data. SOGII data collection also allows covered entities, such as social services agencies, providers, and insurance plans, to identify the complex health needs of LGBTQI+ people and document compliance with nondiscrimination laws.

Why Information about LGBTQI+ Health Issues is Limited

Despite progressive strides in LGBTQI+ health issues, there is still a lot to learn. The LGBTQI+ community is a vast population full of many communities with diverse lived experiences. Structural oppression of such a multifaceted population has produced heath disparities which are not sufficiently studied and tracked. Existing efforts to collect information is inconsistent, not inclusive, and poorly developed because the health and well-being of LGBTQI+ issues are not prioritized, and often excluded, within the U.S. health care system.

For example, health information about gender diverse populations, like transgender and non-binary people, is mostly an afterthought if a thought at all. Today, the standard question about gender still only asks whether a person is “male” or “female”. This question is found on insurance applications, electronic health records, insurance cards and ID cards, and other health-related documents. CMS and SSA, among other federal agencies, only include the binary gender options despite calls for federal recognition of X gender markers. Not only does it restrict self-identification and “out” transgender, non-binary, and gender diverse people, it is poor data. This informational void is magnified given that over twenty states legally recognize non-binary gender markers, and the majority of states have legal pathways to change gender markers on identity documents. Binary sex-based data buries information about the patients on the other side of the computer screen. LGBTQI+ populations, especially transgender and non-binary individuals, are harmed disproportionately. This is because one’s sex assigned at birth does not convey one’s gender identity, sexual orientation, race, and additional factors that capture complex lived experiences. Even when more expansive LGBTQI+ data is collected, it is often not disaggregated.

The Challenges of Targeting SOGII Health Data

Implementing data collection and addressing related bias is complicated. There is a lot of information about LGBTQI+ health issues, but adequate data is inconsistent and sparse. There is no consensus that SOGII data should be collected. For many of the same reasons, this makes untangling potential intersectional bias in existing health care algorithms difficult. Requirements to collect SOGII vary by agency, county, state and at the federal level. Although some states are collecting COVID-19 related SOGII data, others do not collect it. Some states require SOGII data collection year round and others do not collect it at all. The lack of large-scale, collective buy-in for SOGII data collection has life and death implications in light of past health crises in history, including the deadliest years of the AIDS health crisis.

The conversation on health disparities and SOGII data has gained more visibility as information on COVID-19’s impact on LGBTQI+ populations has surfaced, especially for Black, Indigenous, and People of Color. One thing is agreed upon: there is not enough information. Health systems across the country, including Medicaid, collectively fail to capture accurate and consistent SOGII data. As more states require targeted SOGII data collection, the logistics are complicated. As explained in the first blog to this series, ADS can never be perfect even with every possible best practice, resource, and intension. While ADS are undoubtedly powerful and efficient tools, it matters who is creating them and who is collecting the data. While white, cisgender, heterosexual males still dominate the tech and medical industries, the integrity of ADS remains vulnerable to biases and ignorance.

Existing algorithms and data collection related to LGBTQI+ populations remain deeply flawed. Algorithms like facial recognition technology, SOGII data collection, laboratory or diagnostic testing, or inconsistencies in electronic health records (EHRs) alienate LGBTQI+ individuals and miss key information resulting in underdiagnoses of serious health conditions, such as cancer, substance use, or HIV/AIDS.

How Data Use Currently Impacts LGBTQIA+ Populations

Poorly built ADS is notorious for data-match errors. Currently, people are required to answer if they are “male” or “female” on the single, streamlined application in order to qualify for health coverage under Medicaid, CHIP, or using federal tax credits. Only a few states have more inclusive versions of this question. As a result, transgender, non-binary, and intersex individuals are susceptible to data-match errors when their application is verified through CMS or SSA. Algorithmic bias also arises in the form of “gender conflicts” in the delivery of health services. While algorithm “workarounds” have helped avoid these errors, they are only a band-aid to a structural problem. That is because most workarounds can only capture transgender and gender-diverse populations who are diagnosed with “gender dysphoria” and/or received gender-affirming care services, such as hormone therapy or surgery. Not every transgender or non-binary person is diagnosed with gender dysphoria or seeks gender-affirming services, either by choice or due to safety or health reasons. On top of that, gender diverse patients already experience systemic barriers to gender-affirming care. As a result, transgender and gender-diverse people remain one of the most medically underserved populations in the U.S.

Better SOGII Data Collection Is Critical to Fight LBTQI+ Health Disparities

Successful algorithms and data collection need consistency in instructions and application. Although many resources exist on the best practices for quality and culturally sensitive SOGII data collection, there is no consensus on how to collect the data. Inconsistent instructions lead to confusing and inaccurate information. On top of these factors, disaggregated data is also important to avoid diluting nuances between sexual orientation, gender identity, and behaviors. Without consensus on data collection designs and practices, efforts to tackle LGBTQI+ health disparities are undermined.

Low-income transgender, non-binary, and intersex individuals are caught in the crosshairs of old rules and old systems as legal and medical advancements on LGBTQI+ health progress. It is important to gain consensus on the value of SOGII data and thoughtfully implement collection tools to maximize the quality of the data to avoid bias. SOGII-related data solutions and new uses of such data will be most effective when developed and implemented thoughtfully. Importantly, data and algorithms are likely going to never fully capture and consider a person’s identity and lived experience, and the related impact on their health care needs. Therefore, the use of SOGII data and related ADS must recognize exceptions to protect individual’s rights and their access to equitable health care.


Additional Resources

Race-Based Prediction in Pregnancy Algorithm Is Damaging to Maternal Health

Preventing Harm from Automated Decision-Making Systems in MedicaidCommon

NHeLP AHRQ Comments

Demanding Ascertainable Standards: Medicaid as a Case Study

Q&A: Using Assessment Tools to Decide Medicaid Coverage

Ensuring that Assessment Tools are Available to Enrollees

Medicaid Assessments for Long-Term Supports & Services (LTSS)

Evaluating Functional Assessments for Older Adults

Opportunities for Public Comment on HCBS Assessment Tools – National Health Law Program

A Promise Unfulfilled: Automated Medicaid Eligibility Decisions

Cases

A.M.C v. Smith

Darjee v. Betlach

Hawkins v. Cohen

LS v. Delia

Related Content