Sarasota Memorial IDs Lung Cancers Earlier with Analytics Platform

Why It Matters

Early‑stage lung cancer detection dramatically improves survival, and a scalable analytics platform shows how health systems can close care gaps without overhauling existing workflows.

Key Takeaways

• Eon platform raised incidental lung nodule alerts from 2 to 170 weekly.
• 75% of 2025 lung cancers diagnosed at Stage I/II, double national rate.
• Follow‑up adherence reached 87% for screened, 92% for high‑risk patients.
• Dedicated lung nodule clinic enabled timely evaluation of thousands of at‑risk patients.

Pulse Analysis

Lung cancer remains the leading cause of cancer death in the United States, and survival hinges on diagnosing the disease at an early stage. While low‑dose CT screening programs have improved detection, a substantial share of early tumors first appear as incidental pulmonary nodules on imaging performed for unrelated complaints. Large health systems routinely generate hundreds of thousands of radiology reports each year, creating a data‑rich but unwieldy environment where actionable findings can be buried in free‑text narratives. Harnessing that hidden intelligence requires technology that can parse language at scale without disrupting clinicians’ workflow.

Sarasota Memorial Health Care System turned to the Eon analytics platform, which applies natural‑language processing to every radiology report in real time. The system extracts nodule size, location, and risk factors, then cross‑references guideline‑based pathways to generate a prioritized alert for care teams. Because the alerts are embedded within existing electronic health‑record interfaces, clinicians receive the information without extra clicks, while a centralized dashboard tracks each patient’s follow‑up status over months. The platform also flags gaps when recommended imaging or consultations are missed, enabling care navigators to intervene proactively rather than reactively.

The impact has been measurable. After deployment, incidental findings presented to the lung‑cancer team surged from two per week to roughly 170, prompting the creation of a dedicated lung nodule clinic that now evaluates thousands of at‑risk individuals annually. In 2025, 75 % of cancers identified through the combined screening and incidental pathways were stage I or II, far exceeding the 28 % national average and driving better survival odds. High adherence rates—87 % for screened patients and 92 % for high‑risk groups—demonstrate that automated longitudinal monitoring can close care gaps, offering a replicable blueprint for other health systems seeking to leverage AI for population‑level oncology stewardship.