Aggregate surveillance data only. No patient information and no treatment recommendations.

Methods and definitions

How NovAMR works

A displayed result begins with a published aggregate report and keeps its original labels, metric, period, location, and source link.

From original document to public row

  1. Identify the reportRecord the publisher, reporting period, population or care setting, and stable source link.
  2. Extract the printed resultKeep the original organism and antibiotic labels, percentage, metric, count, and page or table locator.
  3. Normalize for searchAdd canonical names without replacing the source wording.
  4. Verify before releaseCompare candidate rows with the original report before they enter a dated public release.
  5. Calculate from eligible rowsFilters, reported metric, denominator class, source year, and geography determine which rows support a summary.

One result, from paper to map

The printed value and the public record stay side by side. Follow the same reported measurement from Stanford's antibiogram into NovAMR.

Original reportStanford Health Care · 2025
Excerpt from Stanford Health Care's 2025 Adult Emergency Department E. coli urine antibiogram. The Ampicillin cell reports 47.5% susceptible among 1,730 tested isolates.View full excerpt
Printed in the Ampicillin column47.5% susceptible · 1,730 tested
NovAMR evidence rowCurrent public Explorer
Current NovAMR Explorer row for Stanford report R004, showing E. coli, Ampicillin, 47.5% reported susceptibility, 1,730 reported isolates, 2025, urine specimen scope, and its original-report link.View Explorer capture
Stored as source-reported susceptibility47.5% · n = 1,730
Organism
Escherichia coli
Antibiotic
Ampicillin
Report scope
E. coli isolated from urine cultures
Source locator
Page 1, E. coli row, Ampicillin column
Record and denominator details

Raw labels ESCHERICHIA COLI and AMPICILLIN are retained. The parenthetical count is stored as a drug-specific reported n. Provenance: R004 · IV1256AAABBEFF88359ECF.

Source excerpt: Stanford Health Care. The original report remains the publisher's work. Regional summaries in NovAMR combine eligible rows and are not the same as this individual source measurement.

Counts remain in separate evidence classes

Drug-tested n
A count printed for the organism-antibiotic result. This can support the primary weighted median.
Organism-level n
A source-wide or organism-row count. It is reported separately as an approximation and is not blended into drug-tested weighting.
No compatible n
A row without a usable denominator. These rows receive an unweighted median, range, row count, and source count.

Reported metrics remain distinct

Susceptibility uses a source-reported susceptible value. Resistance uses only a source-reported resistant value. NovAMR does not calculate resistance as 100 minus susceptibility because intermediate and other categories may exist.

Time and geography are explicit

Latest available per source selects each source's most recent eligible period. Map areas use real administrative boundaries where available; local nearest-source polygons are comparison regions, not clinical catchments or official service areas. Source-reported coordinates and estimated centers remain labeled separately.