We Designed an EHR Workflow to Help Identify Trafficked Kids. You Can Too.

Clinicians carry a heavy load, and we’re not in the habit of making it any heavier. But we want to make the case for one more thing worth looking for, because we’re uniquely situated to help children who desperately need it.

Nearly 88% of trafficking victims interact with a healthcare provider during their exploitation. The physical and psychological toll of exploitation is significant — victims present with infections, injuries, mental health crises, and other care needs that might appear entirely routine. Emergency departments (EDs), urgent care centers, and primary care offices are among the most consistent points of contact these children have with the outside world. That sobering statistic is an opening. We can turn clinical encounters into opportunities for identification and intervention.

Traffickers are deliberate about limiting their victims’ contact with the outside world. They accompany them to appointments, answer questions on their behalf, and pose as family members. Healthcare settings are one of the few places that contact still happens. That gives us an opportunity to intervene when others cannot.

Our team at Baptist Health in Jacksonville, Florida designed a child sex trafficking screening protocol in Epic for the pediatric emergency department (ED). The tool doesn’t ask children direct questions unlikely to elicit disclosure. Instead, it helps clinicians flag indirect warning signs — patterns in visit history, presenting symptoms, behavioral cues — and guides them through a trauma-informed response when risk is identified. If a screen is positive, the workflow directs the care team through next steps: safely separating the patient from accompanying individuals, engaging social services, and connecting to reporting resources.

We consider our data worth sharing with the broader community. From November 2023 through October 2024, we averaged more than 9,000 pediatric ED visits per month. Boys represented 38% of positive screens — considerably higher than national estimates, which have historically placed the female proportion at roughly 83%. We cannot draw firm conclusions from a single-site study, but the gap between our findings and those estimates was wide enough to raise questions about whether current assumptions undercount male victims.

Most identified children were teenagers, though infants and younger children appeared in the data as well. Cases clustered heavily in the late afternoon and evening, suggesting that shift-level vigilance may matter. And nearly all positive screens involved English-speaking patients. We interpret these patterns cautiously, as they might reflect limits of our current tool rather than the true distribution of risk. Additional studies and evaluation of different tools are warranted.

Staff used the screening protocol in more than 95% of eligible visits — a number we attribute largely to workflow design and staff training. Rather than asking clinicians to adopt a separate process, we built the screening into the electronic health record (EHR) workflows they already use every day. When the benefit of a workflow is clear and the next step is obvious, compliance tends to follow.

We want to be straightforward about what we don’t yet know. Our program did not capture the rate at which positive screens were ultimately substantiated. Understanding the ratio of true to false positives is essential, both for refining the tool and for managing a legitimate concern: some trafficking risk indicators have benign explanations. An autistic child might avoid eye contact; a teenager might be evasive for entirely ordinary reasons. Flagging a family for investigation based on a misread screen can cause real harm. We took this seriously through implementation and training, and we continue to do so.

Nevertheless, the infrastructure exists to do this well. The hurdles in front of us are not reasons to stop trying. The EHR tools are the easy part; our workflow is now freely available in Epic’s Foundation System. The harder work is organizational: training clinicians on how to use the tool appropriately, building relationships with social work and legal teams, and committing to measuring outcomes so the approach can improve over time.

We have already learned that embedding structured workflows into the EHR improves screening and identification of problems like suicide risk, domestic violence, and social determinants of health. Child sex trafficking fits that pattern. The question isn’t whether we can help. It’s how we help most effectively. Let’s work together to figure this out.