Coverage is uneven
Publicly available antibiograms are not a representative sample of hospitals, laboratories, patients, countries, or health systems. A blank area means no matching mapped evidence in the current release, not an absence of resistance.
Reports are heterogeneous
Facilities use different testing methods, breakpoints, suppression rules, isolate-selection rules, specimen groupings, care settings, and reporting periods. Side-by-side values may not be directly comparable.
Counts have different scopes
A printed n may refer to a drug-tested cell, an organism row, a table, or an entire report. NovAMR separates these evidence classes, but source ambiguity can remain.
Years are not synchronized
Latest available per source can combine different source years. It is useful for current coverage, but it is not a synchronized cross-sectional survey.
Geography is approximate
A source location may represent a facility, laboratory network, referral population, or reporting jurisdiction. Generated nearest-source regions are visual comparison areas, not patient catchments.
Missing metrics remain missing
Resistance is not inferred from susceptibility. As a result, a resistance view can contain fewer calculation rows than the same susceptibility query.
Source access can change
Institutions may move, replace, or remove files. NovAMR retains source metadata and links, but cannot guarantee that every external document remains online.
Not clinical guidance
Aggregate antibiograms cannot account for an individual patient's infection site, allergies, prior cultures, dosing, local formulary, or current breakpoint guidance.
Use results as a trail to evidence
Open the contributing rows and original reports before drawing a conclusion. For clinical decisions, use current local guidance and a qualified healthcare professional.