Baptist Memorial's Systemwide Incidental Nodule Program: Faster Care, Higher Follow-Up, Solid ROI
Baptist Memorial Health Care Corp. spans more than 20 hospitals, 100 imaging sites and 3,000 staffed beds across Tennessee, Mississippi and Arkansas. The system set a bold target: reduce lung cancer deaths in the Mid-South by 25% by 2030.
To get there, they had to fix two things: inconsistent identification of incidental pulmonary nodules and uneven follow-up. The result of their overhaul: more than double the nodule detections, faster routing of high-risk cases and $3.44 million in incremental direct margin from downstream services in 12 months.
The challenge
Processes were manual and fragmented. Teams combed through multiple systems, re-validating lists and duplicating work across sites. As the footprint grew, the cracks widened-no consistent way to capture findings or push patients into the right clinical pathways.
Leadership knew early detection requires consistency at scale. "Every hospital had its own process," said Parker Harris, CEO of Baptist Memorial Hospital-Tipton and system oncology service line administrator. "We needed a reliable, systemwide way to identify patients and move them into care-fast."
The plan
With an internal grant, Baptist evaluated AI options with outside advisors. The top priority: accurate, consistent identification of incidental findings across all facilities-and the operational muscle to manage the volume that better detection would create.
A prevalence study with Eon, an AI-powered longitudinal care management platform, proved the gap. Baptist believed incidental pulmonary nodules appeared in 0.362% of reports. Eon analyzed 2,000 reports and found 2.28%. That delta represented a large group at risk of being missed.
Equally important, the vendor brought an operations layer to handle the surge in identified cases. Detection plus capacity-both were required to make the program sustainable.
Execution: standardize the flow, automate the next step
Eon integrated with Baptist's Epic EHR. Radiologist-dictated reports flow to Eon via HL7 on signature. The system extracts incidental pulmonary nodules and key clinical details. Through FHIR, it pulls in risk factors like smoking history, prior cancer, PCP and preferred language to complete the profile.
Eon's Navigation Services review each case. When next steps are clear based on guidelines and Baptist-approved protocols, the system auto-generates follow-up orders and communications. Everything writes back to Epic and is recorded within Eon.
Ambiguous or higher-risk cases route to local Baptist teams via a prioritized worklist. The platform also captures related imaging and updates plans when findings progress-so patients don't sit on outdated pathways.
Bottom line: Baptist more than doubled nodule identification without adding staff, while creating one consistent workflow across hospitals and imaging centers.
Results that matter
- Incidental findings: 7,160 in 2024; on track for 18,000+ in 2025.
- Risk mix: 53% of nodules are >8mm-higher risk and need timely review.
- Speed: ~2 days from radiologist report to provider review for high-risk nodules.
- Follow-up adherence: 89% for incidental nodules; 77% for screening patients.
- Cancers found: 625 detected through cases surfaced and managed on the platform.
- Financial impact: $3.44M incremental direct margin in 12 months from PET/CT, interventional radiology, interventional pulmonology and thoracic surgery.
- Volume lift: +60% interventional pulmonary, +99% interventional radiology, +65% PET/CT and more thoracic surgeries.
- Payback: ROI in seven months.
Clinical guidance was central to the model. For reference on screening thresholds and follow-up, see the USPSTF lung cancer screening recommendation.
Management playbook: how to replicate
- Quantify the gap first. Run a prevalence study on a sample of reports to validate missed findings and size the opportunity.
- Pair AI with operational coverage. Identification without capacity creates backlogs and burnout. Resource the follow-up work.
- Integrate with the EHR. Use HL7 for report feeds and FHIR to enrich cases with risk factors and provider attribution.
- Standardize protocols. Lock clinical pathways to guidelines and system-approved rules; auto-generate next steps when clear.
- Escalate edge cases. Push ambiguous or high-risk findings to local experts through a prioritized worklist.
- Track both outcomes and margin. Measure adherence, time-to-review, cancers detected and downstream service lines.
- Design for scale. Build a single workflow that works across hospitals and imaging sites-urban and rural.
KPIs to run weekly
- Incidental nodule prevalence by site and modality
- Time from report signature to provider review (by risk tier)
- Follow-up adherence rates for incidental and screening populations
- Cancers detected attributable to the program
- Downstream volumes and margin by service line
- Escalation rate and turnaround for high-risk/ambiguous cases
Why this worked
Leadership aligned on a clear outcome-earlier detection and equitable access across the Mid-South-and backed it with consistent workflows, automation and capacity. The system didn't just find more nodules; it moved people into care faster and kept them coming back for follow-up.
If you're building a similar program, start with accurate identification, remove manual friction and make the right next step automatic. Then measure it every week.
If your leadership team is upskilling on AI selection and implementation, you can browse practical, management-focused programs here: AI courses by job and latest AI courses.
Your membership also unlocks:
