Bridging the Gap: Dipak Patel’s AI-Driven Healthcare Language Access Initiative

Success Story

It is safe to say that every CEO in America, and perhaps the entire industrialized world, is grappling with the challenge of how to harness artificial intelligence (AI) for good while safeguarding against bad.

In the healthcare industry, in particular, AI holds the potential to achieve fantastic efficiency improvements, despite privacy and well-being concerns. Indeed, a new report from McKinsey & Company and Harvard researchers estimates that AI is expected to impact healthcare savings by 5-10%, or roughly $360 billion annually.

Beyond the financials, Dipak Patel, CEO of GLOBO Language Solutions, believes that the use of AI in healthcare linguistics holds enormous promise for patients and clinicians, helping to increase empathy, understanding, and trust at every touch point throughout the patient journey.

This is a Godsend for people who speak a language other than English at home, of which there are an estimated 68 million in the U.S. today. It is also expected to relieve administrative overload for healthcare providers and staff already impacted by a workforce labor crisis straining the healthcare system.

Patel knows firsthand how challenging the healthcare environment can be for linguistically challenged patients, as he witnessed his parents struggle with language barriers following their immigration from India. For his mother, who battled multiple system atrophy, a disease that causes muscles in the body to deteriorate, it was especially difficult communicating with her U.S. doctors.

“I would go with my mother to visit specialists, and sometimes we got an interpreter, but quite often we did not,” he recalls. “When physicians noticed my mom could speak English, they just assumed she could understand everything they were saying.

Leaving most appointments, she would ask me, ‘What happened? I don’t understand what they said.’ When we returned home, my parents would call one of our physicians in India to get an explanation, essentially dismissing what the other physician told her.”

Had his mother had adequate interpreting services, Patel believes much of this confusion could have been avoided. More importantly, he believes her diagnosis would have been confirmed sooner, improving her quality of life and creating a stronger sense of trust with her U.S. providers. As AI becomes more infused in translation and interpretation, he believes this will only improve.

“When patients need an interpreter,” Patel explains, “they might get one when they show up at the emergency department or if they let the provider know ahead of time.

However, during a hospital stay, they most likely won’t receive interpreting services as a matter of routine, such as ordering a meal from the cafeteria, or when nurses come in the middle of the night to draw blood. Similarly, they likely won’t receive language services when they go home and have questions about their bill.

“At GLOBO, we envision a world where providers will use a mix of in-person, over the phone and video interpreters, as well as AI interpreting, so patients can effectively communicate with providers and payers throughout the entire patient journey.

Not only will this facilitate better communications, but it will allow providers to build trust with their patients in a cost efficient manner.”

Patel also sees AI impacting administrative tasks such as scheduling and billing, drawing from a central data source to simplify the message for different stakeholders. He also expects it to improve communications with clinicians who are already benefiting in diagnostics to interpret imaging and other tests. As monitoring increases through “wearables” to track blood pressure, heart conditions, and more, AI will help to crunch and analyze the data, sending alerts if a patient experiences an issue.

On the drug development side, Patel envisions AI playing a role in helping to identify potential drug candidates and analyze clinical trial results, ultimately accelerating the time to market. Training is another area where he believes AI will have a strong impact, especially given today’s shortage of healthcare instructors.

What is the downside? “Everyone is of course concerned about privacy and security issues,” Patel explains, along with biases and inequality. While AI engines are in learning mode, he warns that a large language model’s data quality could throw off results.

In 2023, for instance, researchers revealed bias in diagnosing Bacterial Vaginosis (BV) in women’s health. Their findings indicated clear diagnostic differences among ethnic groups, with Hispanic women having the most false-positive diagnoses and Asian women receiving the most false-negative.

“The AI was trained on historical data from a limited number of people,” Patel explains, which resulted in imprecise and unequal representation among that specific patient cohort.

As clinical diagnostics companies work to address women’s unmet medical needs, those that persevere in the successful use of AI for medical devices and other healthcare technologies will focus on three things:

1. guarantee differentiated, balanced, and comprehensive data sets;
2. create proven ways to interpret data fairly and equally to exclude inherent biases; and
3. serve specific segments of a population with personalized support.

“Stringent regulations are not needed to address these problems because the market will self-correct,” he adds.

Just as important, Patel cautions that there is “a fear that more use of AI will erode human-to-human interaction.” However, in actuality, he believes the industry may experience the exact opposite, with AI freeing more of the clinician’s time to focus on patient care, not less.

He also expects AI to provide clinicians and healthcare administrators with more data insights that can lead to care improvements.

Patel sees five areas where insights will play an important role in healthcare:

1. Reporting and Data – To assess populations and drill down on healthcare usage
2. Prompt-based Reporting – Enabling users to query databases for quick answers to questions
3. Industry Standards – Establishing guardrails to monitor quality, which currently does not exist in the language solutions space
4. Health Literacy – Assessing a range of abilities, including reading, comprehending, and analyzing information
5. Interpreter Feedback – Providing helpful feedback to interpreters to improve the quality of live sessions (e.g. slow down when speaking)

Additionally, he envisions AI being instrumental in analyzing social determinants of health (SDoH), now an integral requirement for value-based care programs such as ACO REACH (Accountable Care Organizations Realizing Equity, Access, and Community Health).

“We know that 80 to 90% of someone’s health is dictated by what happens in the environment,” Patel explains. “What people eat, where they live, whether they have transportation, all matters.

When you monitor SDoH variables it opens up the data tremendously and ties back to what AI is really good at – taking large amounts of data and coming up with what needs to be done.

Healthcare needs AI to digest this type of information, analyze it, and do risk assessments, all of which are also essential to the growing trend toward personalized medicine.”

While Patel admits that much of the healthcare industry still perceives language solutions in any permutation as a commodity they must legally provide, he posits this could well shift to viewing it as a “superpower,” especially when augmented with AI, which is expected to drastically increase access.

He sees the ability to communicate with patients in their preferred language as foundational to creating better experiences for patients and clinicians, resulting in better outcomes.

“Those hospitals and healthcare organizations that embrace the use of linguistics, which includes AI-enabled solutions, across the entire patient journey, will be at the forefront of transforming healthcare,” he adds, noting that “more and more administrators are beginning to understand this value proposition. They are starting to see it.”

GLOBO is actively pursuing AI as part of its hybrid strategy to provide the right level of translation and interpreting for any given circumstance. “Depending on the complexity of the situation, such as challenging diagnosis, there will continue to be a need for live interpreters,” he explains.

“However, there are thousands of touchpoints in between where AI can and should be utilized to ensure patients get the information they need, in their language of choice, throughout their healthcare journey.”