An empirical comparison of isolate-based and sample-based definitions of antimicrobial resistance and their effect on estimates of prevalence.

•A literature search demonstrated the widespread use of isolate-level measurement of resistance.•Faecal samples were tested for antimicrobial resistant E. coli at both the sample-level and the isolate-level.•Sample-level measurement was more likely to demonstrate resistance.•Using simulation we back predict isolate-level results from an historic study that only used sample-level measurement.•We question the widespread use of single isolate tests as demonstrated by our literature search.