Duloxetine Delayed-Release

• CMS State Drug Utilization Data (SDUD), data.medicaid.gov

CMS's Medicaid Amount Reimbursed for this exact NDC, summed across every state and quarter of the period shown, fee-for-service and managed care together — the program's gross payment to pharmacies before manufacturer rebates (statutory rebates alone recover 23.1% or more of a brand's price). Cells CMS suppresses (fewer than 11 prescriptions in a state-quarter) are absent from the file, so totals for small drugs run low. The prescription count is CMS's Number of Prescriptions. To check: filter the SDUD file for the year to this NDC and sum the column.

Pipeline and calculations built by Claude (Fable 5) from the public datasets above. Educational only — verify against primary sources before acting.

• FDA NDC Directory (openFDA) — labelers active in the last 30 days, repackagers excluded via DailyMed SPL roles
• FDA drug shortages and enforcement reports (openFDA)
• CMS NADAC weekly file — price now vs the price in effect a year ago

Fragility v1 adds fixed points per condition for this NDC's molecule × dosage form: active labelers minus repackagers (1 → +30, 2 → +20, 3 → +10, 4–5 → +5; unknown count → 0); shortage (this NDC listed Current +25, another presentation of the same molecule/form +15, a resolution within 12 months +8); Class I/II recalls in 12 months (this NDC +15, same molecule/form +8); NADAC down 40%+ year over year +15 or 20%+ +8 (a Medicare-negotiated price cut also trips this); sterile injectable form or intravenous route +10; a to-be-discontinued FDA listing +10; capped at 100. Every point shows as a line above. To check: count the molecule's labelers in the NDC Directory, then look it up in FDA's shortage list, openFDA enforcement, and two NADAC files a year apart.

Pipeline and calculations built by Claude (Fable 5) from the public datasets above. Educational only — verify against primary sources before acting.

• FDA NDC Directory labeler counts, CMS NADAC price trend, openFDA shortages and enforcement, FDA inspection classifications, Wayback archive of FDA's shortage list (labels)

A logistic regression over molecule × dosage-form markets, trained on a monthly grid of historical dates since 2019 with labels equal to whether a new FDA shortage posting appeared within the following 90 days; eight features knowable on each date (labeler count, sole-source flag, NADAC year-over-year change, Class I/II recalls, injectable form, discontinuation filings, open shortage episodes, recent OAI inspection). The probability shown is the model's output for today's feature values; the reasons are the features pushing this market's probability up, listed only when the feature actually holds here. Held-out performance and coefficients are on the fragility page.

Pipeline and calculations built by Claude (Fable 5) from the public datasets above. Educational only — verify against primary sources before acting.

• DailyMed SPL labels (human prescription full release, monthly) — establishment DUNS and operations per label
• FDA Drug Establishments Current Registration Site (DECRS) — DUNS to name and country
• FDA Data Dashboard inspection classifications (Drug Quality Assurance)

The label for this product's NDC-9 names each establishment and its operations; those DUNS numbers are matched to FDA's registration list for name and country. Only rows with a MANUFACTURE or API MANUFACTURE operation are places the drug is made; repack, relabel, pack, label and analysis rows are shown as what they are. The OAI badge means a drug-quality (GMP) inspection at that facility ended in Official Action Indicated within 24 months. Labels list plants, not volumes — a drug with two named plants may be made almost entirely at one. To check: open the label on DailyMed and its establishment section.

Pipeline and calculations built by Claude (Fable 5) from the public datasets above. Educational only — verify against primary sources before acting.