CMI: Serious clinical consequences of a semi-automated system for antimicrobial susceptibility testing of carbapenem-resistant Enterobacteriaceae

Background: Carbapenem-resistant Enterobacteriaceae (CRE) is considered a global health emergency
.
Infections caused by CRE are associated with high mortality, relapse rates, and microbiological failure
.
There are differences in the reliability of antimicrobial susceptibility tests (ASTs) performed by semi-automated systems for several drugs commonly used to treat CRE
.
Both the Clinical and Laboratory Standards Institute (CLSI) and the European Committee on Antimicrobial Susceptibility Testing (EUCAST) recommend the use of the reference microdilution method (BMD) for susceptibility testing of some antibiotics of last resort, such as colistin and tigecycline Vinegar, or fosfomycin agar dilution to avoid this problem
.
Furthermore, reference methods are not always used for routine drug susceptibility testing in diagnostic laboratories due to the additional workload required compared to semi-automated systems
.

OBJECTIVE: In this study, we investigated the discrepancy rate of drug susceptibility testing for CRE between the semi-automated system and the reference system, and the impact of the so-called error of the semi-automatic system in guiding targeted therapy for bloodstream infection (BSI) in CRE
.

Methods: A multicenter retrospective study of patients with monomicrobial BSI caused by CRE from January 2013 to December 2016
.
Local testing of non-repeated isolates in index blood cultures using a semi-automated system centrally in referring laboratories and retesting by reference broth microdilution or agar dilution methods
.

RESULTS: We included 366 patients with CRE-BSI; 220 (60%) were male and the median age (IQR) was 67 years (54-76 years)
.
When compared with the results of the reference method, the results of the semi-automatic system exhibit a very large error (VME, ie false magnetic susceptibility) and a large error variable ratio (ME, ie false resistance)
.
Fosfomycin (14%) and colistin (13.
9%) had the highest incidence of VME, while gentamicin (21%), fosfomycin (7.
7%) and tigecycline (34%) had the highest incidence of VME highest rate
.
Overall, VME and ME led clinicians to ineffective treatment in 25 of 341 patients (7%)
.
According to Kaplan-Meier survival curves, receiving ineffective treatment supported by a misleading drug susceptibility test was associated with higher 30-day mortality compared with receiving active treatment [56% vs 26% (p=0.
002)], after adjusting for COX This difference was confirmed after confounders in the regression model [aHR 2.
91 (95% 1.
62-5.
22) p<0.
001]
.

Table 1 Based on EUCAST breakpoints* of meropenem, fosfomycin, amikacin, gentamicin, colistin, and tigecycline*, 366 strains of carbapenem-resistant strains collected from 3 large tertiary teaching hospitals Minimal inhibitory distribution (MIC), MIC50, MIC90 of Enterobacteriaceae

Figure 1 Kaplan-Meier curve analysis of the impact of inappropriate targeted therapy due to misleading results from automated tests compared to aggressive therapy based on retrospective reference testing on the same strains
.
Compare with log-rank test

Conclusion: ME and VME are relatively common semi-automatic drug susceptibility testing systems
.
VME is associated with inappropriate use of antibiotics and poor prognosis
.

Original source: Bartoletti M, Antonelli A, Bussini L, et al.
Clinical consequences of very major and errors with semi-automated testing systems for antimicrobial susceptibility of carbapenem-resistant enterobacterales.
Clin Microbiol Infect 2022 Mar 17