AI Uncovers Hidden Bias in Amputation Rates Among Minority and Rural Patients

AI Uncovers Implicit Bias in Amputation Decisions

A groundbreaking study published in Epidemiology has employed artificial intelligence to investigate the disparities in amputation rates among patients with vascular disease. The research, led by Paula Strassle from the University of Maryland's School of Public Health, reveals that implicit bias may play a significant role in clinical decision-making, particularly affecting minority and rural patients 1

1

2

The Scope of the Problem

Peripheral Artery Disease (PAD) affects over 12 million adults in the United States, with about 10% developing Chronic Limb-Threatening Ischemia (CLTI). These conditions can lead to severe outcomes, including limb loss. The study focused on hospitalizations between 2017 and 2019 across five states, examining cases of PAD and CLTI in patients under 40 1

1

AI Model and Methodology

Researchers programmed an AI model to analyze over 70 variables contributing to differences in leg amputations. These variables included:

• Clinical factors (age, health conditions)
• Healthcare system capacity
• Legal and regulatory climate
• Physical environment (distance to emergency rooms, ZIP code median income)

This comprehensive approach allowed for a nuanced examination of intersectionality across race, sex, income, and rurality 2

2

Key Findings

Source: News-Medical

After accounting for known clinical differences and variations in hospital and neighborhood resources, the study found persistently higher amputation rates among:

• Black, Hispanic, and Native American people in rural areas
• Black and Native American people in urban areas

These findings suggest a "substantial unexplained portion" that points to implicit bias in clinical decision-making at both physician and hospital levels 1

1

Implications for Clinical Practice

Source: Medical Xpress

The study highlights the complexity of decision-making in vascular surgery. As Dr. Katharine McGinigle, a vascular surgeon and senior author of the paper, explains:

"As vascular surgeons, we have surgical guidelines, but we don't have detailed guidelines to help us make the decision between amputating someone's leg and limb-saving surgery in patients who are not medically ready" 2

2

The researchers hope their findings will lead to the development of comprehensive, evidence-based guidelines that help clinicians avoid unconscious biases and make more objective decisions 1

1

Related Stories

Researchers Call for Standardization of Race and Ethnicity Data in Medical AI to Address Bias

AI in Medicine: Study Reveals Socioeconomic Bias in Treatment Recommendations

AI screening tool boosts diabetic eye exam referrals for African Americans by 46%

Mount Sinai Researchers Develop AEquity: A Tool to Combat Bias in Healthcare AI

The Role of AI in Healthcare Equity

This study demonstrates the potential of AI in addressing healthcare disparities. By uncovering hidden biases and providing data-driven insights, AI can contribute to more equitable healthcare outcomes. As Strassle notes:

"This AI model will allow us to easily assess intersectionality across race, sex, income and rurality, and offers us the ability to indirectly study hard-to-measure causes of disparities, like implicit bias and stereotyping" 2

2

Future Directions

The researchers aim for their findings to inform health policies and guidelines that promote objective decision-making in vascular surgery. By addressing unconscious biases and unjustified differences in care quality, they hope to improve outcomes for people living with advanced vascular disease, particularly those from minority and rural communities 1

1

2