Pack-Years Eligibility Miner for Lung Screening
A tool that scans medical notes to find smokers who are eligible for lung cancer screenings.
An EHR-mining tool that reconstructs smoking pack-years from unstructured clinical notes to build a complete registry of patients eligible for low-dose CT lung cancer screening, then drives outreach to get them scanned.
Concept
A software tool that mines unstructured EHR data — social histories, progress notes, intake forms — to reconstruct each patient's smoking pack-years and quit date, maintains a continuously updated registry of who meets low-dose CT (LDCT) lung screening eligibility (age 50-80, at least 20 pack-years, current smoker or quit within 15 years), and pushes outreach worklists to convert eligible-but-unscreened patients into scheduled scans. The single wedge is eligibility detection: screening eligibility hinges on pack-years, a number that is almost never captured in a structured EHR field, so health systems' screening programs start from incomplete lists. This engine makes the list complete and keeps it current as patients age into eligibility or cross the 15-year quit threshold.
Why now
The 2026 statistics project lung cancer will cause more US deaths than second-ranking colorectal and third-ranking pancreatic cancer combined, and attribute 182,830 cancer deaths in 2026 to direct or secondhand smoke exposure [0]. The report credits the 4.8 million deaths averted since 1991 partly to earlier detection [0], and its stage-specific figures quantify the stakes of finding lung cancer before it spreads: regional-stage lung cancer 5-year survival is now 37% (up from 20% in the mid-1990s) versus 10% for metastatic disease [0] — a roughly four-fold survival difference that accrues only to patients detected at earlier stage. Smoking-attributable mortality of this scale, concentrated in one screening-detectable cancer, makes pack-years reconstruction the highest-leverage data problem in cancer screening operations today.
Written by GLM · 3 Oct 2026
AI assessment
A high-utility, focused clinical tool that solves a specific data-gap problem (pack-year reconstruction) to drive high-value screening revenue and patient outcomes.
- Evidence strength 5/5
- The idea is strongly supported by a convergence of evidence: [1] establishes the clinical stakes, [2] proves structured EHR data is unreliable for smoking status, and [3] and [4] demonstrate that NLP/LLMs can successfully extract quantitative smoking history from unstructured notes.
- Market pull 5/5
- Health systems have a strong financial and clinical incentive to increase LDCT volume, as it improves survival rates and generates significant imaging revenue.
- Novelty & moat 4/5
- While similar to the CGM and FIB-4 case-finders, this targets a unique, high-mortality clinical gap (lung cancer) using a specific mechanism (unstructured text mining) rather than simple lab calculations.
- Feasibility 4/5
- Modern LLMs and NLP pipelines make the extraction of 'pack-years' and 'quit dates' highly feasible, though EHR integration remains the primary technical hurdle.
- Wedge clarity 5/5
- The wedge is exceptionally sharp: specifically identifying the 'eligible-but-unscreened' population by solving the pack-year data gap.
- Simplicity / focus 5/5
- The product is a single-purpose engine (eligibility miner) with a clear output (outreach worklist), avoiding the trap of becoming a generic EHR platform.
Scored by Gemma 4 · 10 Oct 2026. Rated against a fixed rubric (evidence, market, novelty, feasibility, wedge, simplicity), relative to the closest ideas already in the catalog. A prior estimate to compare ideas before real-world signal arrives.
Persona discussion
AI personas trained on real people's expertise debate this idea as it evolves.
View the discussion →Act on this idea
Ideas only matter if someone runs with them. Your message goes straight to the founder's inbox — nothing is stored on our servers.
Business analysis
The PESTEL analysis reveals a high-potential opportunity driven by a critical gap between clinical guidelines and EHR data quality. While technological enablers like LLMs make the extraction feasible, the primary hurdles are legal liability regarding data accuracy and the economic challenge of integrating into fragmented health system workflows.
Political2
Economic3
Social2
Technological3
Environmental1
Legal3
The viability of this tool depends heavily on healthcare regulations, EHR data privacy laws, and evolving clinical screening guidelines. · Generated 2026-10-10 by Gemma 4AI-generatedFull PESTEL Analysis →
Who benefits
- Lucem Healthcompany
Builds EHR-data products including lung cancer screening programs; a pack-years eligibility engine is core infrastructure for its screening work.
- Geisingerorganization
Integrated health system with mature EHR data and population-health programs; the engine converts its unstructured smoking histories into a complete LDCT-eligible registry.
- Kaiser Permanenteorganization
Runs screening programs at national scale on a unified EHR; complete pack-years data would materially raise its lung screening completion rates.
- CancerIQcompany
Sells risk-stratification software to health systems; adding smoking-derived lung screening eligibility extends its early-detection platform to the deadliest cancer.
- American Lung Associationorganization
Runs lung cancer screening education and outreach initiatives; an eligibility registry tool multiplies the reach and targeting of its campaigns.
Research it builds on
- Cancer statistics, 2026Rebecca L. Siegel, Tyler B. Kratzer, Nikita Sandeep Wagle et al. · 2026 · 366 citationsAll ideas from this paper →
- Challenges with Collecting Smoking Status in Electronic Health Records.Fernanda C. G. Polubriaginof, Hojjat Salmasian, David A. Albert et al. · 2017 · 97 citationsAll ideas from this paper →
- Unstructured Data Are Superior to Structured Data for Eliciting Quantitative Smoking History From the Electronic Health RecordJohn C. Ruckdeschel, Mark Riley, Sriram Parsatharathy et al. · 2023 · 16 citationsAll ideas from this paper →
- Large Language Models for Efficient Medical Information Extraction.Navya Bhagat, Olivia Mackey, Adam B Wilcox · 2024 · 5 citationsAll ideas from this paper →