Why Digital Mental Health Fails Spanish Speakers

TL;DR: A new peer-reviewed study published in JMIR Formative Research (2025) shows that Spanish-language mental health services can achieve high engagement and satisfaction at scale when designed around linguistic concordance, cultural responsiveness, and genuine accessibility. We served 6,163 users across all 50 U.S. states, delivered 36,858 telehealth sessions, and achieved a Net Promoter Score of +85. The findings challenge prevailing assumptions about how to close the Latino mental health gap — and make the case for a fundamentally different approach to solving it.

The Industry's Blind Spot: Surface-Level Access Is Not Enough

Digital mental health has experienced explosive growth over the past five years. Telehealth platforms, therapy apps, and AI-powered wellness tools have raised billions in venture capital on the premise of democratizing access to care.

Yet for Spanish-speaking adults in the United States — over 40 million people, making it the second-largest Spanish-speaking country in the world — this growth has largely bypassed them.

The numbers are stark: between 2014 and 2019, the number of facilities offering mental health treatment in Spanish decreased by 17.8%, even as the Latino population grew by 4.5% during the same period. Only 5% of psychologists offer Spanish-language services. Nearly half of Latino adults with a perceived need for mental health care report that need going unmet.

The standard industry response has been to add Spanish-language options to existing platforms — a dropdown menu here, a translated interface there. This approach treats language as a feature to be toggled rather than a fundamental dimension of care quality.

It doesn't work. And we now have peer-reviewed evidence showing what does.

The Uncomfortable Math: Why U.S.-Licensed Providers Can't Solve This

Before presenting our findings, we need to address something the industry avoids saying plainly: it is not mathematically possible to close the Latino mental health gap using U.S.-licensed providers alone. Not in our lifetimes.

Only 5-6% of the U.S. mental health workforce identifies as Latino or offers Spanish-language services. The pipeline of new providers is not growing fast enough to change this. According to the Health Resources and Services Administration, only 26.4% of the nation's mental health professional workforce needs are currently being met — and that's the overall figure, before accounting for language concordance.

Meanwhile, the Spanish-speaking population continues to grow, and unmet need continues to compound.

The conventional response — recruit more bilingual providers, expand telehealth networks, translate apps — is not wrong. It's just insufficient. These efforts will take decades to meaningfully shift supply, while millions of people go without adequate support today.

The Translation Fallacy

Most digital health companies treat Spanish-language support as a localization problem: translate the app, hire a few bilingual providers, check the box. This approach systematically underinvests in the provider supply, cultural adaptation, and user experience elements that actually drive engagement and outcomes.

The Insurance Dependency

The U.S. mental health system assumes insurance-based access. But Spanish-speaking adults are disproportionately uninsured or underinsured, and the administrative complexity of insurance-based care creates additional barriers. Models that require insurance eligibility as a prerequisite systematically exclude the populations most in need.

The Domestic Provider Constraint

Waiting for domestic workforce development to solve the provider shortage means waiting decades while need goes unmet. The math simply doesn't work.

A Different Model: Subclinical Support from Providers Who Share Language and Culture

We built Sanarai on a fundamentally different premise: connecting Spanish-speaking adults in the U.S. to licensed mental health professionals in Latin America who deliver subclinical support services via telehealth.

To be clear: we're not an app or a self-guided digital tool. We deliver live, synchronous video sessions with licensed professionals. The distinction matters, because much of the "digital mental health" category has underdelivered for Spanish-speaking populations precisely because asynchronous, app-based interventions lack the relational and cultural dimensions that drive engagement.

Our model is also not a workaround or a compromise. It's a recognition of two realities:

First, a significant proportion of mental health concerns can be addressed through supportive services that do not require diagnosis or treatment of diagnosed disease. Stress, relationship challenges, life transitions, grief, early-stage anxiety and depression — these are the concerns that drive most people to seek help, and they can often be addressed through professional support outside traditional clinical frameworks. This is the same logic underlying Employee Assistance Programs (EAPs) and the rapidly growing coaching industry, both of which have operated successfully outside the licensed clinical model for decades.

Second, linguistic and cultural concordance are not nice-to-haves — they are prerequisites for effective mental health support. Research consistently shows that Spanish-speaking adults prefer to discuss emotional concerns in Spanish, and that language-concordant care improves therapeutic alliance, engagement, and outcomes. Providers from Latin America offer this naturally, without the friction of operating in a second language or across cultural distance.

Our providers are licensed mental health professionals from Mexico and Argentina with master's degrees and a minimum of five years of clinical experience. They deliver psychosocial and emotional support services — not psychiatric diagnosis or medication management — at price points ($49-$59 per session) accessible to uninsured and underinsured populations.

The question is whether this model actually works at scale. That's what our research set out to answer.

What the Research Shows: Key Findings from the JMIR Study

In December 2025, JMIR Formative Research published "Reach, Engagement, and Acceptability of a Subclinical Telehealth Service for Spanish-Speaking Adults: Retrospective Mixed Methods Pilot Study" — a four-year evaluation of our model conducted in collaboration with researchers at UC Berkeley and UC San Francisco.

The study examined three implementation outcomes: reach, engagement, and acceptability. Here's what we found:

Reach

• 6,163 Spanish-speaking users served between August 2020 and September 2024
• Coverage across all 50 U.S. states, the District of Columbia, and Puerto Rico
• Highest concentration in states with large Latino populations and significant provider shortages: Texas (15.3%), California (14%), Florida (13.3%), New York (6.1%), and New Jersey (4.7%)
• 47% average quarter-over-quarter growth in active users throughout the evaluation period

Engagement

• 36,858 total appointments delivered, including individual sessions, couples sessions, and initial consultations
• 94% of users scheduled their first appointment within 7 days; 43% scheduled within one day
• 62.6% of users engaged in two or more sessions
• Average of 8.94 sessions per engaged user over an average of 110 days
• Average wait time of 2.87 days to see a provider — compared to the 3+ month waitlists reported by 56% of U.S. psychologists in 2023

Acceptability

• Average session satisfaction rating: 4.88 out of 5
• Net Promoter Score: +85 (indicating strong likelihood of recommending services)
• 95% of survey respondents indicated intent to schedule another session
• Qualitative interviews revealed users chose our service specifically for language concordance, cultural fit, affordability, and scheduling flexibility

Why These Numbers Matter: Context for Decision-Makers

For health system leaders, payers, and funders evaluating solutions for Spanish-speaking populations, these findings address several critical questions.

Does this model actually reach the populations that need it?

Yes. The geographic distribution mirrors the distribution of unmet need among Spanish-speaking communities. Users came from every state, with concentration in regions where HRSA data shows severe mental health workforce shortages — and where telehealth adoption by mental health facilities has historically lagged.

Do Spanish-speaking users engage with telehealth at comparable rates?

Yes, and in some cases at higher rates. The average of 8.94 sessions per engaged user compares favorably to published benchmarks: prior research shows Latino adults attend an average of 5 sessions for telephone-based mental health services and 8 sessions for in-person care. The 110-day average engagement duration suggests sustained relationships rather than one-off interactions.

Does a model using providers from Latin America deliver acceptable care?

The evidence strongly supports this. Users explicitly cited the linguistic and cultural concordance of our providers as central to their positive experience — not as a compromise, but as a feature. Qualitative interviews revealed that users perceived these providers as more culturally aligned than the limited Spanish-language options available locally.

What Users Actually Said: Insights from the Qualitative Data

The mixed-methods design provides direct insight into why users engaged with our service. Interviews with 30 users revealed four consistent themes:

1. Language Concordance Is Non-Negotiable

Users repeatedly emphasized the importance of receiving care in their primary language — not as a convenience, but as a prerequisite for meaningful engagement. Many had tried to access local services and found no Spanish-speaking providers available, or had been offered interpreter services that felt inadequate for discussing emotional concerns.

2. Cultural Fit Goes Beyond Language

Speaking Spanish is necessary but not sufficient. Users valued providers who shared cultural reference points, understood family dynamics common in Latino communities, and didn't require extensive explanation of cultural context.

3. Affordability Must Be Real, Not Theoretical

At $49-$59 per session without insurance requirements, co-pays, or deductibles, we reached users who had been priced out of traditional care. Users noted that even when their insurance nominally covered mental health, the actual out-of-pocket costs were comparable — without the language or cultural benefits.

4. Accessibility Means More Than Telehealth

Users cited ease of scheduling, short wait times, and the ability to receive care from home as critical enablers. But they also emphasized the simplicity of the user experience: a clear workflow, predictable pricing, and the ability to change providers without friction.

Implications for Payers, Health Systems, and Funders

For Commercial and Medicaid Payers

The data suggests that subclinical telehealth services can function as a high-value, high-retention benefit for Spanish-speaking populations — particularly for the large segment of mental health need that does not require diagnosis or clinical treatment.

This is not a replacement for clinical care. It's a complement — one that addresses the same population segment currently served by EAPs and coaching, but with linguistic and cultural concordance that those services rarely offer.

The engagement data (62.6% multi-session rate, 8.94 average sessions) indicates that users are receiving sustained support, not just sampling services. For payers seeking to address mental health access disparities and improve quality measures related to follow-up care, models like ours warrant serious consideration.

For Health Systems

The workforce shortage is not going away. Our study demonstrates that connecting patients to international providers via telehealth can produce satisfaction and engagement outcomes that meet or exceed domestic benchmarks.

For health systems struggling to offer Spanish-language behavioral health services, partnership models that extend support networks internationally may offer a faster path to equity than domestic hiring alone — particularly for the subclinical support needs that represent the majority of demand.

For Funders and Philanthropic Organizations

This study was supported by the California Health Care Foundation and represents exactly the kind of implementation research the field needs: rigorous evaluation of real-world service delivery, focused on populations experiencing the greatest disparities.

The findings validate continued investment in culturally and linguistically concordant models — and suggest that such models can achieve scale without sacrificing quality. They also point toward the next research priority: outcomes-based evaluation to determine whether engagement translates to meaningful improvement in mental health and wellbeing.

Limitations and What Comes Next

We're transparent about this study's limitations:

• No clinical outcomes data: The study evaluates implementation outcomes (reach, engagement, acceptability) rather than clinical effectiveness. Whether engagement translates to symptom improvement remains to be evaluated.
• Self-selection in satisfaction surveys: Only 22.5% of users responded to satisfaction surveys, which may inflate positive findings.
• Qualitative sample composition: Interview participants were disproportionately women (70%), potentially limiting generalizability.
• Descriptive design: Without a comparison group, the study cannot establish that our model outperforms alternatives — only that it achieves strong absolute outcomes.

These limitations point toward our next phase of research: outcomes-based evaluation and, ultimately, controlled trials comparing culturally concordant subclinical telehealth to standard care pathways.

The Larger Point: Solving the Problem We Actually Have

The gap in Latino mental health access is real, it's growing, and it will not be solved by incremental improvements to the current system. The provider math doesn't work. The insurance model excludes too many people. The translation-as-localization approach doesn't drive engagement.

Our research suggests that when you build differently — when you leverage international providers who share language and culture, focus on subclinical support rather than trying to replicate the full clinical model, and design for accessibility from the start — the results are strong. Users show up. They come back. They recommend the service to others.

That's not a workaround. That's what meeting people where they are actually looks like.

About the Research

Citation: Arévalo Avalos MR, Fu J, Aguilera A, Suarez L. Reach, Engagement, and Acceptability of a Subclinical Telehealth Service for Spanish-Speaking Adults: Retrospective Mixed Methods Pilot Study. JMIR Formative Research. 2025;9:e80026. doi: 10.2196/80026

Funding: California Health Care Foundation (Grant #G-33613)

Authors: Marvyn R. Arévalo Avalos, PhD (UC Berkeley School of Social Welfare); Julio Fu, MBA (Sanarai); Adrian Aguilera, PhD (UC Berkeley, UC San Francisco); Luis Suarez, MBA (Sanarai)

Sanarai is a digital health company connecting Spanish-speaking adults in the United States to licensed mental health professionals in Latin America via telehealth. Learn more at sanarai.com.