Abstract
SummaryObjectivesThe magnitude of impact caused by low blood culture utilization on estimates of the proportions and incidence rates of antimicrobial-resistant (AMR) bacterial infections is largely unknown.MethodsWe used routine electronic databases of microbiology, hospital admission and drug prescription at Sunpasitthiprasong Hospital, Ubon Ratchathani, Thailand, from 2011 to 2015, and bootstrap simulations.ResultsThe proportions of Escherichia coli and Klebsiella pneumoniae bacteraemias caused by 3rd generation cephalosporin resistant isolates (3GCREC and 3GCRKP) were estimated to increase by 13 and 24 percentage points (from 44% to 57% and from 51% to 75%), respectively, if blood culture utilization rate was reduced from 82 to 26 blood culture specimens per 1,000 patient-days. Among patients with hospital-origin bloodstream infections, the proportion of 3GCREC and 3GCRKP whose first positive blood culture was taken within ±1 calendar day of the start of a parenteral antibiotic at the study hospital was substantially lower than those whose first positive blood culture was taken later into parenteral antibiotic treatment (30% versus 79%, p<0.001; and 37% versus 86%, p<0.001). Similar effects were observed for methicillin-resistant Staphylococcus aureus, carbapenem-resistant Acinetobacter spp. and carbapenem-resistant Pseudomonas aeruginosa.ConclusionImpacts of low blood culture utilization rate on the estimated proportions and incidence rates of AMR infections could be high. We recommend that AMR surveillance reports should additionally include blood culture utilization rate and stratification by exposure to a parenteral antibiotic at the hospital.
Highlights
Antimicrobial-resistance (AMR) surveillance reports are commonly used to monitor trends, inform recommendations for empirical therapy, estimate the burden of AMR, and assess the impact of local, national and global interventions.[1,2,3] Such reportsImpact of low blood culture usage infection, respectively
We focused on rates of blood culture utilization in relation to bloodstream infection (BSI) caused by Escherichia coli, Klebsiella pneumoniae, Staphylococcus aureus, Acinetobacter spp. and Pseudomonas aeruginosa because these are the top five pathogens attributable to deaths caused by AMR infections in Thailand[8] and the EU and the European Economic Area.[9]
We illustrate that estimates of the proportions and incidence rates of AMR infections could be considerably changed due to low blood culture utilization rates in low and middleincome countries (LMICs) settings, where patients are frequently treated empirically and blood culture is frequently sampled after empirical treatment failure
Summary
Antimicrobial-resistance (AMR) surveillance reports are commonly used to monitor trends, inform recommendations for empirical therapy, estimate the burden of AMR, and assess the impact of local, national and global interventions.[1,2,3] Such reportsImpact of low blood culture usage infection, respectively. Estimated parameters include the proportions of patients with bloodstream infection (BSI) caused by AMR isolates (using an isolate-based surveillance approach), and incidence rates of patients with bloodstream infection (BSI) caused by AMR isolates in the tested population (using a sample-based surveillance approach) stratified by origin of infection (community or hospital).[1]. It is well recognised that a low blood culture utilization rate can bias AMR surveillance data,[1,4,6] but the magnitude of impact from this on estimates of proportions and incidence rates of AMR infections is largely unknown.[7] In hospitals in low and middleincome countries (LMICs), patients with severe infectious diseases are frequently treated empirically and blood culture is frequently sampled after empirical treatment failure. We quantify the magnitude of effect caused by low blood culture utilization rates in LMICs on estimates of proportions and incidence rates of AMR infections. We develop and evaluate a new parameter to represent blood culture utilization in LMICs
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
