LactCheck is a drug safety tool that helps breastfeeding mothers and healthcare providers understand how medications transfer into breast milk.
Unlike simple lookup tools, LactCheck uses molecular structure analysis to estimate drug transfer for medications that haven't been directly studied in clinical lactation trials. This is particularly valuable for newer medications or less-common drugs where published data is scarce.
LactCheck uses machine learning to predict two key metrics:
| Model | Spearman ρ | R² | MAE | Training Size |
|---|---|---|---|---|
| RID — Full Range | 0.29 | 0.05 | 2.90% | n=446 |
| RID — Filtered (≤20%) | 0.48 | 0.17 | 0.67% | n=429 |
| M:P — Filtered (≤3.0) | 0.575 | 0.33 | 0.47 | n=239 |
| M:P — Full Range | 0.52 | 0.07 | 1.01 | n=290 |
Filtered models cover the typical clinical range (RID ≤20%, M:P ≤3.0), where most medications fall. Full-range models extend predictions to extreme values, covering RID from 0.002% to 67% and M:P from 0.001 to 37.0 in training data.
Verified data comes from published clinical lactation studies. These are marked with a green "Verified" badge on LactCheck.
Predicted data is generated by our machine learning model for drugs not yet studied in clinical lactation trials. Predictions are based on the drug's molecular structure and known physicochemical properties.
LactCheck includes a structural novelty detection system that identifies drugs whose molecular structure differs substantially from the drugs in our training database.
Metabolite flags: We also flag drugs with active metabolites (e.g., prodrugs like codeine that convert to morphine, or fluoxetine whose active metabolite norfluoxetine has a 4-16 day half-life). These drugs may transfer differently into breast milk than their parent compound alone would predict.
This tool is not a substitute for professional medical advice. LactCheck provides estimates based on published data and machine learning. It cannot account for your individual medical situation, other medications you may be taking, or your baby's specific health needs.
Always consult with your healthcare provider before making any decisions about medication use during breastfeeding. This is especially important for:
LactCheck is built by Benevolence Labs. We believe that new mothers deserve access to the same quantitative safety data that healthcare providers use. Not just vague categorizations like "probably safe."
Our mission is to close the information gap in pediatric pharmacology by applying modern machine learning to the problem of medication safety during lactation.
Questions about LactCheck or this research? Contact us at [email protected].