Quiet Neglect: Algorithmic Bias in Healthcare Is Hurting Older Adults

AI algorithms have the potential to vastly improve health monitoring for older adults. From detecting early warning signs of chronic disease, to using AI-enabled telemedicine to expand healthcare access in rural communities, to informing highly personalized treatment plans, the potential is driving a rapid acceleration of AI within this demographic. However, unless we address the known gaps within the data sets these algorithms draw from, we risk widening and accelerating the very health inequities these advancements aimed to solve.

It’s not news that the data foundations most healthcare algorithms are built from largely exclude the experiences of older adults — with further data gaps spanning race, gender, and income in this population. For example, demographic and health surveys typically exclude women aged 50 and over and men aged 55 or 60 and over from their remit. Further gaps in data representation among older adults of color risk perpetuating racial bias, while gaps among lower-income older adults and those from rural vs. urban communities omit critical context of lived experience, widening other biases.

Innovators, entrepreneurs, and investors have a significant opportunity to compete on equity while helping address the root cause of these healthcare data gaps. Here’s how these market leaders can do better.  

  • Bridge the data gap for marginalized older adults. We need to widen the representation of aging populations in big data generation and collection and in a manner that explicitly includes marginalized populations. One way to bridge the data gap is by prioritizing solutions that address data acquisition and/or disaggregation for underrepresented population segments. Filling the data gap can be accomplished through a variety of ways, from elevating the voices of those with lived experiences to investing in emerging data scaling strategies, such as cache database queries, database indexes, database replication, and sharding (or splitting large databases).
  • Navigate the democratization of AI. As AI in healthcare becomes more ubiquitous, its strategic importance, effects, and management need to be more defined and integrated across the healthcare sector. As new companies emerge to deliver data development, collection, solutions, and platforms, infusing equity into the landscape of health tech solutions will be critical over the next several years. Namely, we need to advance the quality and accuracy of data, and data-dependent tools, in a manner that improves health and social care outcomes for all older adults. Further, we need expanded investment in data generation and collection efforts that focus on factors, such as social determinants of health, that drive health inequities for aging populations.
  • Prioritize equity as a competitive lever. Equity is one of the defining factors of high-quality healthcare solutions and thereby is also a competitive advantage enabling personalization and tailored care that can in turn lead to better and more equitable outcomes. With the continued push for value-based care, equity will be at the core of scalable cost-effective care delivery.  Regulators and policymakers have an opportunity to accelerate this market driver by incentivizing solutions that provide measurable, scalable gains in equitable health outcomes among older adults.

Equitable AI isn’t an aspiration; it’s an absolute necessity, particularly for the millions of older Americans who remain unseen within the current frameworks and miss out on algorithmic benefits such as risk profiles and early interventions for certain diseases. Fortunately, innovators, entrepreneurs, and investors have the opportunity now to prioritize and fund robust data foundations, ensuring that the needs of older and marginalized adults are no longer overlooked and underserved.

Photo: kali9, Getty Images


Leave a Reply

Your email address will not be published. Required fields are marked *