Claims-based LASSO algorithm for identifying interstitial lung disease in Japanese cancer patients

A LASSO model identified interstitial lung disease (ILD) in Japanese cancer patients: sensitivity 33.5% (external 19.8%), specificity 99.3% (99.4%), positive predictive value (PPV) 76.7% (65.5%). High PPV aids case confirmation despite low sensitivity.

• The best-performing model was a reduced LASSO model using confirmed ILD diagnosis codes, KL-6/serum surfactant protein-D codes, age, and sex.
• External validation showed similar high specificity (99.4%) and positive predictive value (65.5%), supporting generalizability.
• Sensitivity was low (33.5% internal, 19.8% external), limiting use for screening, but the high PPV supports confident case confirmation in retrospective comparative studies.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.