Understanding the digital voice of the patient can help us improve mental health and disparities in care

We live in a time when almost 20% of U.S. adults suffer from anxiety. Close to 1 in 4 women and 1 in 7 live with an anxiety condition that interferes with their daily activities, impacts their job or school performance and even affects their personal relationships.

Simultaneously, more than 18 million U.S. adults are suffering from major depression, or about 7.9% Adult Non-Hispanic whites, 5.4% of African Americans, 5.4% of Hispanics and 4.4% of Asians. In addition to a more depressed and anxious population, the Psychology and Psychiatry associations have underlined a concerning 33% increase in the suicide rate between 1999 through 2017.

While millions of dollars are invested in public health campaigns and advertising, limited resources are allocated to carefully understand and identify cultural similarities and differences among diverse populations. Minimum efforts are dedicated to exploring and testing how individuals from each culture can be better persuaded to modify their lifestyles, prevent disease, or opportunely seek medical advice. It is common to see general population-oriented campaigns simply translated into other languages with the hope that they will contribute to reducing disparity. This is particularly critical as today’s population in America is 40% diverse and it is set to become majority-minority by 2040. This shift in the population becomes increasingly concerning given people from racial/diverse groups are less likely to receive mental health care. According to the American Psychiatry Association (APA), among adults with any mental illness, 48% of whites received mental health services, compared with 31% of blacks and Hispanics, and 22% of Asians.

This compounding effect of population shift and lower utilization rates, ultimately has an impact on healthcare cost and economic factors for all. Mental disorders are among the top most costly health conditions for adults 18 to 64 in the U.S., along with cancer and trauma-related disorders, according to APA.

With a shift in population dynamics comes the need to shift how we tackle the mental health crisis America is facing today. Today more than ever, we need cultural intelligence in the research and insights gathered, medical education, patient interventions and family/ community engagement programs to make a difference that could potentially impact more lives.

Can Digital Data & AI Algorithms help?

The internet represents one of the principal ways to receive and share information and ideas with over 4.1 billion active users on social media platforms, websites, and blogs every day.

Clearly, in this era of everything digital, online discussions represent one of the most valuable sources of unbiased insight with healthcare being a critical topic. Every day millions of patients, family members, caregivers, community members or healthcare stakeholders volunteer their comments and/or questions online, representing a powerful source of open, candid and unsolicited information about people's true relationship to their condition and the often untold social and emotional factors that surround it. In all, 80 percent of Internet users, or about 93 million Americans, have searched for health-related topics online, according to a study released by the Pew Internet & American Life Project. That’s up from 62 percent of Internet users who said they went online to research health topics in 2001.

While some claim that our reliance on social media and digital platforms can have a detrimental effect on mental health, ironically, its organic content may represent a rich source of patient and caregiver insights that can give us signals to proactively help patients with solutions personalized to their need and eventually save lives.

The emergence of leading technologies like Artificial Intelligence (AI), and Natural Language Processing (NLP), can help us proactively mine open-source digital discussions of mental health patients to truly tap into their voice, needs and cultural nuances; when no one is asking. Rethinking our methods to become digital anthropologists, can generate unique insights regarding emotional, cognitive and behavioral factors that are often missed by traditional methods or clinical case data. The collection and proper interpretation of this data can ultimately drive interventions that impact patients, their families, caregivers, as well as impact the healthcare ecosystem, with much needed emotional and cultural personalization in content and care.    

We must rethink how we design personalized mental health care with cultural intelligence?

While much literature exists about barriers to mental health treatment and clinical data reports lower utilization rates among diverse populations (i.e. Hispanics and African Americans), limited quantitative data or studies exist to understand barriers-to-care and the cultural, emotional and lifestyle factors impacting the patient-journey, particularly by ethnicity. This lack of personalized insights is particularly concerning as diverse-segments represent 100% of new population growth in the U.S. and among younger generations like Gen Z's and Millennials, they already represent almost half of the total population today. We are estimated to become a majority-minority nation by 2040 and unless we have an inclusive approach to mental health treatment, it will continue to grow into a compounding, and deadly issue, for our society.

We have an opportunity to educate and enhance treatment interventions and education informed by the voice of millions of spontaneous digital posts shared daily by patients in different stages of their health experience including their emotional responses to illness-related symptoms, how they search for information, how they rationalize their condition and decide when to seek medical care.    

This is why researchers from Medical Strategies at Johnson & Johnson's Hispanic Organization for Leadership and Achievement (HOLA) Employee resources group teamed up with Culturintel to use new technologies to tap into the open digital voice of the patient and better understand barriers to depression and mental health treatment for Hispanics versus overall patients. Also, a collaboration with The Cleveland Clinic studied factors related to suicide rates for epilepsy sufferers by generation. This effort analyzed more than 543,000 U.S. digital conversations related to depression, with 43K self-identified as Hispanic/ Latino. The findings offer unique cultural intelligence to design interventions that align to the lifestyle, emotional and cultural needs that are often hard to quantify or report in medical literature at scale.

Unlike traditional social listening, Culturintel's unique algorithm crawls and mines anywhere there is a relevant open and organic discussion happening extending to forums, topical sites, and message boards; which in the case of mental health represented 74% of all data sources analyzed, with social networks only representing 16% of the discussion volume. Ethnic segmentation is tagged based on how users organically self-identify. Also, the method is designed to be GDPR compliant protecting user privacy since personally-identifiable information is never collected nor stored. The insights reported as outputs reflect the voice of the people as a collective to identify patterns in affinity by segment through patient journey stages.

For its unique findings and the innovative use of artificial intelligence, machine learning, and leading big-data tools to turn open-source patient discussions into actionable insights, Culturintel studies about mental health were selected to present during the World Congress of Psychiatry in Lisbon, Portugal. The WPA is psychiatry’s global association representing 140 psychiatric societies in 120 countries and supporting more than 250,000 psychiatrists.

Findings to be reported at the congress will showcase how AI-powered tools, text mining, data clustering, segmentation, sentiment analysis, and other NLP applications can yield a better understanding of the patients’ path to treatment and the barriers and factors that influence their decision to seek mental health/ depression treatment or not. Unique findings from this first-ever study done at scale will be reported to the public after the congress in Portugal. More information will be published upon completion at voiceofthepatient.com.

Co-authored by Lili Gil Valletta and Carlos Gutierrez, MD, MPH. Lili Gil Valletta, award-winning entrepreneur, former pharma executive, and co-founder and CEO of the cultural intelligence? big data and marketing firms CIEN+ and Culturintel. She is a World Economic Forum Young Global Leader, member of the Harvard Women’s Leadership Board at the Harvard Kennedy School and an independent TV commentator. Dr. Gutierrez is an experienced medical professional with expertise in the global pharmaceutical industry, biotechnology, precision medicine, genetics, neuroscience, and public health; currently serves as Chief Medical Strategy Fellow at CIEN+.