Columbia College of Dental Medicine Integrates Evidence-Based Caries Prevention Into Clinical Workflow

When Dr. Biana Roykh arrived at Columbia University College of Dental Medicine (CDM) to serve as associate dean for clinical affairs, she set out to bring a proven, evidence-based approach to caries management into everyday clinical practice. Having trained and practiced within the Caries Management by Risk Assessment (CAMBRA) model at the University of California, San Francisco (UCSF) under its creator, Dr. John Featherstone, she knew firsthand how it could transform patient care.

CAMBRA is a preventive approach to oral health that helps clinicians assess a patient's risk of developing tooth decay and personalize care. Rather than relying on a one-size-fits-all treatment plan, providers evaluate factors that increase risk, such as diet, harmful bacteria, or reduced saliva flow, alongside protective factors like fluoride use and healthy saliva production. The resulting risk assessment guides preventive therapies, treatment recommendations, and follow-up care aimed at identifying disease early and preventing cavities from progressing.

At CDM, CAMBRA was already taught throughout the curriculum, and faculty were committed to its preventive approach. Students could complete paper-based caries risk assessments, but the results were not consistently carried forward into treatment decisions or follow-up care. The college's recent transition to Epic created an opportunity to build a standardized workflow that connected those assessments to the clinical decisions that followed each patient encounter. "The philosophy was there, but it hadn't really made its way into clinical operations," Roykh said.

Roykh partnered with Dr. Sharon Perelman, DDS, associate professor and associate dean of clinical informatics and innovation at CDM. A general dentist with a master's degree in medical informatics and a certified Epic physician builder, Perelman brought the clinical and technical expertise needed to translate that vision into the electronic health record.

The project is the focus of a newly published Journal of Dental Education article, "Shifting the Paradigm: CAMBRA Adoption and Medical-Dental Integration Leveraging an Electronic Health Record," authored by Perelman and colleagues. The paper documents how Columbia integrated CAMBRA into its Epic electronic health record and redesigned the clinical process to support risk-based care, offering a framework that other academic dental institutions can adapt. Following implementation, completion of caries risk assessments increased from fewer than 5% of eligible patient visits to 89%. 

Embedding CAMBRA into the electronic health record was only the first step. Beyond digitizing the assessment, Perelman designed a clinical workflow that translated risk assessments into treatment decisions and follow-up care."My part was developing this algorithm in Epic," she said. "How do we take this piece of paper and its scoring and build it into the workflow? It wasn't just the CAMBRA scoring. It was the entire workflow that we implemented beyond that."

The timing, however, made the project particularly ambitious. Columbia had only recently completed its enterprise-wide Epic implementation when the COVID-19 pandemic began, leaving no resources available to support a year-long effort to build such a complex system. Instead, Perelman led the technical development herself, working with former Epic employees on her team and collaborating closely with Featherstone to translate and validate the CAMBRA protocol within Epic.

The team redesigned the patient encounter from start to finish. Rather than simply recording caries risk scores, the system guides students through the next steps in care. Support tools prompt providers to consider documenting the patient's diagnosis, adding the appropriate caries risk diagnosis to the problem list, prescribing evidence-based preventive therapies, and incorporating those recommendations into the treatment plan.

The system was designed to reinforce clinical judgment while helping students apply what they had learned in the classroom to patient care. "We have students learning a lot at once," Perelman said. "It wasn't to tell them what to do. It was saying, 'Because you have this score, think about these things. Do you want to incorporate them into your treatment plan?' "

The results reflected strong adoption of the new clinical process. After a mandatory confirmation step was introduced, provider agreement with the algorithm-generated caries risk level exceeded 96%, demonstrating faculty buy-in and trust.

As development continued, Perelman designed the workflow to support patients after they left the clinic as well. Because the CAMBRA protocol often calls for instructions that differ from standard package labeling, she incorporated protocol-specific directions into both the prescription label and the patient's after-visit summary in MyChart. If patients forgot how to use the medication, those instructions remained accessible through their electronic health record.

She also considered how patients would receive those prescriptions. If patients never picked up their medication, the treatment plan could break down before it began. To address that challenge, Perelman developed an internal pharmacy workflow within Epic that allowed medications to be dispensed directly in the clinic while automatically becoming part of the patient's permanent electronic medication record. At the same time, a patient's moderate or high caries risk could also become part of the broader medical record, allowing oral health information to be viewed alongside medications across Columbia's integrated health system.

For Roykh, that integration represents one of the project's greatest advances. "What value do we place on a patient going to see their cardiologist and being able to see that they're at high risk for dental disease?" she said. "It's priceless, because you're starting to have a conversation with your medical counterparts."

The new approach also made it possible to connect a patient's caries risk with the care they received and measure how that risk changed over time. By consistently documenting risk assessments, diagnoses, preventive care, counseling, and follow-up, CDM can now evaluate whether its approach is reducing patients' risk rather than simply treating cavities after they develop.

"What we're trying to do is move away from the drill and fill model and focus more on prevention and risk assessment," Perelman said. "We want to show that we're implementing this protocol, providing non-surgical interventions, and achieving good outcomes."

Turning the workflow into everyday clinical practice required collaboration across the college. As the new system was introduced, Perelman and Roykh worked alongside students, faculty, and staff, using usability testing, shadowing, and real-time feedback to identify barriers and iteratively refine the workflow.

"You really can't achieve this from an ivory tower and you can't achieve it at the grassroots level," Roykh said. "You really need that full value chain alignment."

The implementation demonstrates what is possible when clinical expertise, education, and informatics work together rather than in isolation. CDM has established itself as a leader in translating evidence-based caries prevention into practice, creating a replicable model for other academic dental institutions.