Traditional Salmonella Typhimurium typing tools (phage typing and MLVA) are sufficient to resolve well-defined outbreak events only.

Between 1991 and 2014 the per capita notification rate of salmonellosis in Australia increased from 31.9 to 69.7 cases per 100,000 people. Salmonella Typhimurium accounted for nearly half the human cases until the end of 2014. In this study, we used cluster analysis tools to compare S. Typhimurium isolates from a chicken-meat study with those reported to the National Enteric Pathogen Surveillance System (NEPSS) from the coincident human and non-human populations. There was limited phage type diversity within all populations and a lack of specificity of MLVA profiling within phage types. The chicken-meat study isolates were not significantly clustered with the human cases and at least 7 non-human sources, based on typing profiles (PT/MLVA combination), could be implicated as a source of human cases during the same period. In the absence of a strong surveillance system representative of all putative sources, MLVA and phage typing alone or in combination are insufficient to identify the source of human cases. • Phage typing and MLVA are insufficient for source attribution of S. Typhimurium. • Limited phage type diversity reveals lack of MLVA specificity. • Cluster analysis identified cattle as an important putative source. • Passive surveillance is insufficient to identify all non-human sources of Salmonella. [ABSTRACT FROM AUTHOR]