Sampling Bias

One of the key Challenges to Integrated Surveillance

Problem

AMR data often comes mainly from hospital patients or diagnostic submissions, which may not represent the wider community or animal reservoirs.

Why This Matters

Hospital-biased sampling means:

  • Community-acquired resistance may be underestimated
  • The data does not reflect AMR prevalence in the general healthy population
  • Animal reservoirs and environmental sources are under-represented
  • Trends may be skewed by changes in hospital admission patterns or diagnostic practices

Proposed Solution

  • Sample both healthy and sick individuals across human, animal, and environmental sectors
  • Monitor sewage as a composite sample of the entire population of a city (see Environmental AMR Surveillance)
  • Extend surveillance beyond clinical settings to include community-based sampling

Nordic examples

  • Sweden conducts routine surveillance for ESBL-producing E. coli from urine and bloodstream infections, providing insight into community-acquired resistance
  • Several Nordic countries are piloting wastewater monitoring as a population-level sampling approach
  • Norway extends animal surveillance to include wild animals, not just livestock

Benefit

Provides a more comprehensive and unbiased picture of AMR prevalence across populations and ecosystems.