Truace
Subscribe

Real-world before-and-after evaluation of AI support for lung cancer diagnosis at three US lung nodule clinics

Background Health systems and payers require evidence that artificial intelligence (AI)-enabled decision support improves care delivery. Integrating AI into lung nodule management pathways may streamline workflows, improve identification and triage of patients with pulmonary nodules, and enable earlier lung cancer diagnosis. Yet real-world evidence of clinical utility remains limited. This study evaluated the impact of implementing commercially available AI software to support the lung nodule and lung cancer car…

By the Truvace desk2 min read1 sourcePeer-reviewedTRV-2026-1921· cite & verify
Real-world before-and-after evaluation of AI support for lung cancer diagnosis at three US lung nodule clinics
Pulmonary adenocarcinoma, left lung, CT scan Case 239 (7603361542) by Yale Rosen from USA. CC BY-SA 2.0 · https://creativecommons.org/licenses/by-sa/2.0

In brief

Researchers compared matched 12.5-month periods before and after commercially available AI software was introduced into lung nodule and lung cancer care pathways at three US health systems. The number of new patients seen rose from 138 to 232, and monthly patients undergoing an invasive procedure climbed from 4.9 to 8.0, both statistically significant.

The findings suggest AI-assisted identification and triage may widen access to specialist review, which matters to health systems and payers seeking evidence of clinical utility. However, the diagnostic outcomes that matter most, such as cancers found and time to diagnosis, did not change significantly, so the benefit for patients remains unproven.

Main points

  1. Median monthly new patients rose from 9 to 16 across the three clinics after implementation.
  2. Monthly cancers moved from 3.0 to 3.8, which was not statistically significant (p = 0.3242).
  3. Monthly early-stage (I-II) cancers rose from 1.4 to 2.4, also not statistically significant (p = 0.125).
  4. Median time to diagnosis was essentially unchanged at 58 versus 57 days, though the interquartile range narrowed from 125 to 58 days.

The gain

After AI software was implemented, monthly new patients at three lung nodule clinics rose from 11.0 to 18.6 and monthly invasive procedures rose from 4.9 to 8.0.

The rundown

The study used a real-world before-and-after design at three US lung nodule clinics, with matched pre- and post-implementation periods of 12.5 months each. Primary outcomes were new patient numbers and patients undergoing at least one invasive procedure.

Secondary outcomes included total lung cancers, early-stage cancers, and time from baseline CT to diagnosis. Analyses used Wilcoxon signed-rank and Mann-Whitney U tests, reporting medians and interquartile ranges for continuous measures.

The authors conclude the software was associated with significantly increased throughput and procedural volume. The narrower spread in time to diagnosis hints at more consistent timelines, though the median did not shift meaningfully.

Sources

  1. Peer-reviewedPLOS One2026-10-14

The debate