Abstract Background Early identification of MRSA by diagnostic medical microbiology laboratories enables improved antimicrobial choice and outcomes. The Cepheid Xpert® MRSA/SA BC test rapidly identifies Staphylococcus aureus bloodstream infections through spa gene detection and methicillin resistance via mecA gene detection. Recent emergence of S. aureus with deletions in the spa gene has resulted in false-negative results for this test, leading to misidentification of infections with this organism, particularly MRSA ST45. Objectives To investigate the emergence and prevalence of ST45 MRSA in New South Wales (NSW), Australia. Methods WGS read data from six NSW hospitals were collected for 131 ST45 MRSA isolates and analysed. Results Of the 131 ST45 MRSA investigated, 88.5% (116/131) contained a deletion in the spa gene that appeared to have arisen once in approximately 2010 followed by clonal expansion. Given the successful establishment of this ‘spa-deletion’ MRSA clone, the Cepheid Xpert® MRSA/SA BC test became unreliable for confirming S. aureus bacteraemia in NSW. Subsequently, the algorithm used by this test has been updated and evaluated to take into account the presence of S. aureus with either a spa deletion or SCCmec target variations. Conclusions This study highlighted the applied use of WGS for assessing diagnostic assays and informing necessary changes to ensure the viability of the Cepheid Xpert® MRSA/SA BC test in the context of the new ‘spa-deletion’ MRSA clone. It demonstrated how continued surveillance through WGS can reveal evolutionary events that may impact diagnostic assays, allowing corrective modifications to be made in real time.