80% of trials miss enrollment targets. Learn why eligibility alone isn’t enough and how SDOH data predicts who actually successfully participates.
In theory, clinical eligibility should predict participation.
In practice, it rarely does.
A recent study on breast cancer screening uncovered something that upends traditional thinking about how we design health programs and clinical trials alike. Researchers found that social, behavioral and environmental factors—like neighborhood income, rurality, and access to care—were among the strongest predictors of who actually showed up for mammograms. Even when two women had identical clinical risk, their likelihood of attending came down to the realities of their everyday lives.
That insight should ring alarm bells for anyone involved in clinical research. Because the same pattern that shapes preventive screening behavior also drives who enrolls in—and stays engaged with—clinical trials. Yet most feasibility models, recruitment plans, and retention strategies continue to treat social context as an afterthought.
The result? Trials that look feasible on paper but stumble in the field, missing enrollment targets, over-spending budgets, and under-representing the very populations they’re meant to help.
The Participation Problem No Algorithm Alone Can Fix
Recruitment challenges aren’t new. Even before the COVID-19 pandemic, nearly 80 percent of clinical trials failed to meet their initial enrollment timelines. What’s changed is our understanding of why.
It’s not that people don’t qualify; it’s that many can’t realistically participate. Transportation, caregiving responsibilities, work schedules, food insecurity, or lack of broadband access can all become invisible barriers. And these aren’t edge cases. They’re widespread, measurable patterns embedded in social determinants of health (SDOH).
When trials overlook these realities, the consequences ripple across every operational and scientific dimension:
- Under-recruitment: Eligible patients identified through EHRs or claims data never appear for screening visits.
- High early drop-out rates: Participants disengage after enrollment due to logistical or financial strain.
- Non-representative cohorts: Results skew toward higher-income, urban, or health-literate populations, weakening generalizability.
- Site inefficiency: Some sites over-perform while others lag, often because their catchment areas differ in social vulnerability, not because of effort or skill.
In short: data models optimized for eligibility often fail at predicting attendance.
The Lesson from Screening Data
The breast-screening study, published in npj Digital Medicine, modeled who completed mammograms across diverse geographies. Clinical risk played a role—but not the defining one. Instead, the predictive power came from social and environmental factors: household income, travel distance, and local healthcare access.
When those SDOH variables were added to the model, the accuracy of predicting attendance rose substantially. When they were removed, predictions faltered even though biological risk stayed constant.
That finding mirrors what clinical research teams see every day: people who qualify on paper often don’t make it to the appointment. Eligibility isn’t the barrier…life is.
If preventive screening, which requires a single short visit, is so sensitive to social context, imagine the magnitude of those barriers for multi-month or multi-year trials that demand ongoing travel, adherence, and communication.
Beyond Eligibility: The Compounding Effect of Social Risk
A second line of research reinforces this message. A 2025 study in JACC Advances examined how obesity interacts with social disadvantage across 136,000 U.S. adults. The results revealed a sobering “double jeopardy.”
Participants facing both severe obesity and high SDOH burden defined by indicators like low income, housing instability, and food insecurity, had a 3.5-fold higher risk of premature mortality than those without those burdens. Even after controlling for clinical risk factors, social context amplified negative outcomes dramatically.
The takeaway for research organizations:
If SDOH can so profoundly predict mortality in the real world, it certainly predicts attrition in a clinical trial. Ignoring it isn’t a neutral choice—it’s an operational liability.
The Hidden Costs of Designing for Eligibility Alone
When social context is missing from trial planning, problems surface in predictable and preventable ways.
1. Flawed Feasibility Forecasts
Feasibility models built solely on incidence and prevalence data can overestimate available participants by 30–50 percent. Without accounting for social barriers like transportation or competing obligations, projected enrollment windows collapse under real-world conditions.
2. Budget Overruns and Timeline Delays
Each month of delayed enrollment costs sponsors an estimated $600,000 to $8 million depending on phase and indication. Dropouts or screen-failures compound these losses—issues often traceable to unmet social needs rather than protocol design.
3. Site Disparities
Two sites can have identical eligibility counts yet radically different performance. Often the difference lies in surrounding community factors—poverty rates, digital access, or healthcare density. Without visibility into those dynamics, CROs can’t allocate resources or support equitably.
4. Equity and Compliance Risks
Regulators are now asking for proof that diversity and inclusion are baked into trial design. If recruitment continually under-represents minority or underserved populations, sponsors face scrutiny from both the FDA and public opinion.
Each of these pain points stems from the same blind spot: data that defines who qualifies without explaining who can participate.
Designing for Attendance: What That Actually Looks Like
The good news is that this challenge is solvable. The same SDOH factors that predict screening behavior can be leveraged to forecast and improve trial participation, especially if teams integrate them early enough.
1. Smarter Site Selection
Instead of choosing sites solely based on EHR counts or investigator availability, add a layer of SDOH analysis. Mapping neighborhoods by transportation access, broadband coverage, and healthcare density helps identify where participants are able to engage.
For example, a site in a lower-income urban area may house plenty of eligible patients but face high travel friction. Building satellite sites, offering rideshare vouchers, or scheduling flexible visits can close that gap—but only if the risk is known upfront.
2. Predictive Enrollment Modeling
When SDOH variables such as housing stability, income range, or digital access—are included alongside clinical data, trial teams can forecast participation likelihood with far greater accuracy. It’s the difference between assuming everyone who qualifies will enroll and understanding that maybe only 60 percent realistically can.
That foresight allows for smarter budgeting, proactive mitigation plans, and more accurate recruitment timelines.
3. Tailored Outreach and Engagement
Understanding community-level context enables more effective communication. If broadband access is low, SMS or mail outreach may outperform email campaigns. If local food insecurity is high, emphasizing transportation stipends or meal support may build trust and remove practical barriers.
Participation increases when outreach reflects real life and not generic assumptions.
4. Proactive Retention Planning
SDOH insights don’t end at enrollment. They inform ongoing participant support. Trials can pre-flag participants at higher risk of dropout based on social risk factors and intervene early with scheduling flexibility, local check-ins, or digital monitoring alternatives.
These adjustments aren’t acts of convenience; they’re data-driven levers to preserve sample size and data integrity.
A New Kind of Feasibility Model
Traditional feasibility asks: Where can we find eligible patients?
Modern feasibility must ask: Where can eligible patients actually participate and stay?
This shift requires moving from static demographic and clinical views to dynamic, context-aware models that integrate SDOH data. The payoff is substantial: faster enrollment, lower attrition, and more representative evidence.
In one recent oncology feasibility analysis, adding community-level SDOH variables improved site-level recruitment predictions by nearly 20 percent and helped sponsors allocate transportation budgets precisely where needed. Those kinds of optimizations don’t just help equity—they help economics.
Where HealthWise Data Fits In
That’s where HealthWise Data comes into the picture.
HealthWise Data curates one of the industry’s most comprehensive SDOH and behavioral data libraries, encompassing thousands of variables across economic, environmental, and social dimensions. Unlike clinical records that capture fragments of a patient’s medical history, HealthWise Data’s consumer healthcare marketing database, featuring over 2,000+ data variables, reveal the conditions that shape participation potential:
- Transportation and mobility access – access to transportation, access to care and distance to site locations
- Technology and internet access – a crucial predictor for decentralized and hybrid trial participation.
- Neighborhood and community stability – housing turnover, safety, and social cohesion metrics that influence adherence.
- Lifestyle strain indicators – food insecurity, income levels, and economic stressors that can impact retention.
By integrating these insights with existing EHR or claims data, sponsors and clinical research organizations can finally move beyond theoretical feasibility to a realistic, actionable model of participation.
In practical terms, that means:
- Selecting sites where both clinical prevalence and social access align.
- Predicting likely no-shows before recruitment begins.
- Designing participant-support programs targeted to the actual needs of local populations.
- Meeting diversity and equity goals with quantifiable, data-driven strategies.
HealthWise Data helps turn “who qualifies” into “who can and will participate.”
What This Means for the Future of Clinical Research
The convergence of screening analytics, SDOH science, and advanced data enrichment is pushing the industry toward a more human-centered model of research design. The lesson from the breast-screening data is clear: eligibility doesn’t guarantee engagement.
As more trials incorporate real-world data and decentralized components, understanding the social fabric of participants will be as critical as understanding their biomarkers. Sponsors who embed SDOH early, especially during feasibility and not after the first recruitment shortfall, will gain measurable advantages in both efficiency and credibility.
Because at the end of the day, trials don’t fail because people don’t care. They fail because systems aren’t designed for the lives participants actually lead.
HealthWise Data: Delivering the SDOH and BDOH Data Needed to Increase Success Rates for Clinical Trials
Clinical trials are about people—people with complex lives, competing responsibilities, and unequal access to care. Focusing on eligibility alone treats participation as a binary variable: yes or no. But participation is a spectrum shaped by social reality.
By embracing SDOH data, researchers can close the gap between who qualifies and who attends, between feasibility and follow-through. The result is faster recruitment, stronger retention, and evidence that better reflects the world it’s meant to serve.
That’s the future HealthWise Data is helping the industry design, one where trials don’t just measure health outcomes, but understand the human factors that make those outcomes possible.
