How Medical Management & Pharmacy Management Processes Contribute To Health Inequities

Health inequity exists throughout the healthcare delivery system affecting historically marginalized and under-resourced communities more than their White counterparts. The pandemic painfully highlighted health disparities through the disproportionate impact of COVID-19 on American Indian/Alaska Native, Hispanic, and Black Americans, who all experienced higher rates of infections and death.

Persons from historically marginalized groups commonly experience poorer health outcomes like those listed below:

• LGBTQIA+ individuals are less likely to be screened for breast, cervical, and colon cancer
• Hispanic adults are 70% more likely than non-Hispanic White adults to be diagnosed with diabetes
• Asian men are 60% more likely to die from liver cancer
• Black women in America are 3 times more likely to die from preventable pregnancy-related complications than White women in America

Pinpointing the causes of health inequities

As critical stakeholders in the US healthcare system, health payers have an opportunity to examine their systems and structures to investigate for other instances of inequity. When people try to access care by using their health insurance, how do common health plan structures like pharmacy benefit tiering or prior authorization (PA) potentially create differences in outcomes between groups, particularly historically marginalized individuals? This is our first article in a 2-part series exploring how member demographic data can measure and help determine if health insurance systems and processes are inadvertently creating healthcare disparities.

Health equity a focus across all payers

Historical racism and discrimination in the US medical system led to direct harm for BIPOC people, particularly for Black Americans. Lack of respectful care, provider implicit bias, and mistreatment creates a lack of trust in the healthcare system and further contributes to health disparities through delayed preventive care and lower medication adherence.

Health payers are acknowledging health inequity and are taking action. CMS is one of many public payers centering equity by embedding health equity into its Strategic Plan and across all CMS programs. The 2024 CMS Final Rule outlines a new Health Equity Index to be added in the 2027 Star ratings for Medicare Advantage Organizations (MAO). Additionally, the National Committee on Quality Assurance (NCQA) continues to expand health equity-based standards and created Health Equity accreditation pathways to help insurers advance their health equity programming. Leaders in private health plans like Dr. Nwando Olayiwola, Chief Health Equity Officer at Humana, champion the role commercial insurers must play in addressing health disparities:

Data's dual role in uncovering and creating bias in healthcare

Data are crucial to understand the health disparities among and between groups of members. Insurers are increasingly asked to collect more individual-level demographic data such as race, ethnicity, language, sexual orientation, and gender identity. Though essential to disaggregate and identify possible disparities between sub-groups of members, this information can be difficult to obtain for many reasons:

• Legal concerns over state regulations that prohibit the collection of race, ethnicity, and other personal information when individuals apply for coverage but allow collection after enrollment
• Employers that choose to not collect or share all available member demographic data with health plans
• Lack of standardization across data sources pose barriers to collecting, ingesting, and using accurate demographic data
• Reduced data sharing capability by providers with limited or non-adoption of EHRs, such as behavioral health providers or rehabilitation facilities
• Hesitancy to share personal data due to concerns about privacy, fear of misuse, and further discrimination

When health plans respectfully gather and communicate how the data will be used, individuals may be more open to providing it, increasing data accuracy. Individual-level data can be applied to quality and health outcomes measures to surface disparities between groups covered by the insurer, as required with NCQA's Healthcare Effectiveness and Data Information Set (HEDIS). Member-level information enables health plans to take data-informed action to address observed health-related social needs, such as partnering and investing in local community-based services, funding affordable housing, and providing healthy meals for members with chronic conditions.

Inherent biases impact patient health outcomes

Data can also hinder equitable access to care. Health insurance systems and tools, along with the data used to build them, show how discrimination can be baked into the healthcare system. Recent research uncovered built-in bias in structures and processes used by healthcare providers and health plans that may widen existing health disparities including:

• Race-adjusted clinical algorithms used to assess and determine treatment for heart and kidney conditions, score Black patients differently due to race which may inaccurately determine their need for treatment.
• Targeting algorithms used by health plans to identify and offer programs to help manage chronic conditions like diabetes and hypertension. Based on historical care use and claim dollars, these algorithms fail to identify Black Americans as they have lower cost of care due to barriers, such as decreased access.
• Modeling of a medical claim algorithm used by a state Medicaid program to determine if an emergency department (ED) visit was non-emergent and therefore, generated a lower payment to the hospital, showed Black and Hispanic children's ED visits were significantly more likely to be classified as non-emergent. This could further impact revenue for cash-strapped hospitals serving under-resourced communities.

Understanding how bias exists up and down the healthcare system equips all payers to look inward and understand how their structures and processes may create different outcomes between groups of their members. Utilization management (UM) and prior authorization (PA) are 2 such areas of opportunity.

Prior authorization and medical management under scrutiny

Utilization management, including requirements to submit PA for services before delivery of services to patients, are among the most common and hotly contested health insurer practices. All payers use these processes to help ensure members receive the right care for their condition that is effective, safe, and at the appropriate level of care, while also controlling costs. The US spends more on care than other peer countries and has poorer health outcomes due in part to wasteful spending, such as the use of low-value care, estimated to account for over $345 billion each year.

Providers universally see PA processes as frustrating, creating a heavy resource burden, delaying needed care, and negatively affecting the patient-provider relationship when requested care is denied. The net impact on quality of care, health outcomes, and cost-of-care savings from UM programs, including the practice of PA, remain unclear according to examination by KFF:

What little is known about the effects of PA processes comes from a 2022 US OIG report on the use of PA by MAOs, where 13% of MAO PA denials actually met Medicare coverage criteria. This report elicited varied action by different stakeholders:

• CMS issued rulemaking which created greater transparency into prior authorization processes and decision making for Medicare, Medicaid, and qualified health plans on the federal marketplace. 
• A US subcommittee of Homeland Security and Governmental Affairs is actively investigating large national payers over their use of PA practices, including algorithms used to evaluate and deny coverage to Medicare Advantage beneficiaries. 
• Large, national MAOs under scrutiny, such as United Healthcare and CVS Aetna, announced plans to reduce prior authorizations starting in late 2023.

Bring member-level demographics into UM data analysis

Polling by the Institute for Healthcare Improvement showed the collection and stratification of outcomes by member demographics (race, ethnicity, language, etc.) to be the top action health organizations need to take to improve health equity. Health plans already closely monitor UM data and care use, which presents an opportunity to bring in member-level data and build upon outcome stratification work to meet existing CMS and NCQA requirements. Health plans can gain insight into possible disparities between groups of members in UM processes by:

• Taking prudent steps to evaluate for hidden bias in targeting algorithms, predictive modeling, and other automated processing used as part of UM
• Leveraging readily available data, such as gender, age, rural vs urban residency and then pull in race, ethnicity, and language data as more is collected
• Stratifying high-level UM outcome data, such as PA approvals, denials, volume of appeals, and appeal outcomes to gain insight into differences in the member experiences when seeking care
• Evaluating any under- or over-representation of different member groups in UM outcomes
• Bringing member-specific data into more granular utilization trending, such as type of care (physical vs behavioral health), level of care (inpatient or outpatient) and service types (surgery vs rehabilitation therapy)

Engage key stakeholders with healthcare transparency and inclusivity

Stratified UM data enables health insurers to understand and create plans to reduce observed disparities. Developing and prioritizing interventions are more meaningful when provider and member voices are included and reinforces the health plan's commitment to health equity. Actions by health plans can include:

• Seeking input from and including members from all community groups served by the health plan regarding approaches to demographic data collection. This practice establishes a foundation of partnership, inclusion, and builds trust.
• Asking members about their lived experience with UM and PA processes when seeking care. This dialogue gives the health plan and providers valuable information that informs opportunities for change.
• Sharing de-identified and stratified UM data with providers to bring visibility to differences between member groups when seeking care, setting the stage for collaboration.
• Posting stratified UM and PA reports and analysis for the public. Sharing this data clearly signals transparency, similar to the health equity report shared by Blue Cross of Massachusetts.

Employer influence in healthcare benefits

As healthcare benefit purchasers, employers are crucial partners for health payers and increasingly include health equity as part of their values and business objectives. Developing programs to address the social determinants of health impacting employees were cited as a priority focus area by large employers in a survey by the Business Group on Health. When providing clear communication around intention of data use, employers are in an ideal position to collect and share voluntarily-provided employee demographic data with health plans.

Improving health and wellness of the workforce has clear financial benefits as addressing health disparities in the US could yield an estimated savings of $42 billion from increased productivity. Employers have a vested interest in identifying if employee sub-groups experience greater burdens when accessing and using their health benefits, including going through the many hoops and hurdles of health plans' UM and PA processes.

RTI Health Advance helps across the healthcare system

Health plans' ability to look inward to measure and discover possible health disparities for members of historically under-served communities requires not only data but a commitment to meaningfully address health equity along the full healthcare system. RTI Health Advance is dedicated to advancing healthcare for everyone and has an experienced team of experts with analytic capabilities, making us an ideal partner for health plans and employers pursuing health equity. Contact us to learn more and watch for Part 2 of this article series.