Accuracy and applicability of different phenotypic methods for carbapenemase detection in Enterobacteriaceae: a systematic review and meta-analysis

Journal Pre-proof Accuracy and applicability of different phenotypic methods for carbapenemase detection in Enterobacteriaceae: a systematic review and meta-analysis Hong Zhong, Meng-Lu Wu, Wen-Juan Feng, Shi-Feng Huang, Ping Yang

PII:

S2213-7165(19)30264-4

DOI:

https://doi.org/10.1016/j.jgar.2019.10.010

Reference:

JGAR 1067

To appear in:

Journal of Global Antimicrobial Resistance

Received Date:

27 June 2019

Revised Date:

9 October 2019

Accepted Date:

10 October 2019

Please cite this article as: Zhong H, Wu M-Lu, Feng W-Juan, Huang S-Feng, Yang P, Accuracy and applicability of different phenotypic methods for carbapenemase detection in Enterobacteriaceae: a systematic review and meta-analysis, Journal of Global Antimicrobial Resistance (2019), doi: https://doi.org/10.1016/j.jgar.2019.10.010

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

Accuracy and applicability of different phenotypic methods for carbapenemase detection in Enterobacteriaceae: a systematic review and meta-analysis Hong Zhong, Meng-Lu Wu, Wen-Juan Feng, Shi-Feng Huang, Ping Yang1 1

Department of Laboratory Medicine, the First Affiliated Hospital of Chongqing

Medical University, Chongqing, China. No.1 Friendship Road, Yuzhong District, Chongqing 400016, China.

ro of

*Correspondence

Professor Ping Yang, Department of Clinical Laboratory, the First Affiliated Hospital of Chongqing Medical University, Chongqing, 400016, China

-p

No.1 Friendship Road, Yuzhong District, Chongqing, 400016, P.R.China

re

E-mail address: [email protected]

lP

Tel: 86-023-89012513; Fax: 86-023-89012513



na

Highlights

We did a meta-analysis and systematic review to assess the accuracy and

ur

applicability of MHT, Carba NP, mCIM, and MALDI-TOF MS for CPE detection.

Carba NP, mCIM and MALDI-TOF MS all demonstrated high accuracies

Jo



in CPE detection, while MHT was not recommended due to some clear drawbacks.



We recommended the selection of carbapenemase detection tests in the order of mCIM, Carba NP and MALDI-TOF MS according to their 1

Jo

ur

na

lP

re

-p

ro of

simplicities, costs and the equipments and skills involved.

2

Abstract Objectives: Infections caused by carbapenem-resistant Enterobacteriaceae (CRE) are major public health threats, and the most important mechanism of resistance to carbapenems in CRE is the production of cabarpenemases. Early identification of carbapenemase-producing Enterobacteriaceae (CPE) leads to improved clinical outcomes. This systematic review was to assess the accuracy and applicability of the

ro of

modified Hodge’s test (MHT), carbapenemase Nordmann-Poirel test (Carba NP),

modified carbapenem inactivation method (mCIM) and matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF-MS) for CPE detection.

-p

Methods: The meta-analysis included pooled sensitivity, specificity, diagnostic odds

re

ratio, summary receiver operating characteristic curve, and area under the curve. Results: 67 studies were included for the analysis. The pooled sensitivities of MHT,

lP

Carba NP, mCIM and MALDI-TOF MS were 92% (95% CI: 87%-95%), 97% (95% CI:

na

94%-98%), 99% (95% CI: 99%-100%), and 99% (95% CI: 96%-100%), respectively; the pooled specificities were 93% (95% CI: 86%-97%), 100% (95% CI: 99%-100%),

ur

99% (95% CI: 96%-100%), and 99% (95% CI: 96%-100%), respectively; the pooled diagnostic odds ratios were 98.156 (95% CI: 48.175-199.995), 1277.710 (95%CI:

Jo

751.391-2172.692), 3597.352 (95% CI: 1287.575-10000), and 1781.360 (95% CI: 651.827-4868.228), respectively; the pooled AUC values of the SROC were 0.97, 1, 1, and 1, respectively. Conclusions: Carba NP, mCIM and MALDI-TOF MS all demonstrated high accuracies in CPE detection, while MHT was not recommended due to some clear drawbacks. We recommended the selection of carbapenemase detection tests in the order of mCIM, 3

Carba NP and MALDI-TOF MS according to their simplicities, costs and the equipments and skills involved. Keywords: carbapenemase-producing Enterobacteriaceae, phenotypic detection, modified Hodge test, Carba NP, modified carbapenem inactivation method, MADLI-

Jo

ur

na

lP

re

-p

ro of

TOF-MS

4

Introduction Carbapenem-resistant Enterobacteriaceae (CRE) causes serious infections and contributes to significant morbidity and mortality in health-care settings [1]. Over the past years, the spread of CRE has been widely reported around the world [2-5]. The mechanisms of carbapenem resistance are varied, including the alteration of penicillin binding proteins (PBPs), the production of an extended spectrum β-lactamases (ESBLs)

ro of

and/or overexpression of AmpC β-lactamases combined with the loss/deficiency of

porins of the outer membrane, the upregulation of the active efflux system, or, more commonly, the production of carbapenemases [6].

-p

Carbapenemases can be divided into different Ambler classes including Ambler class

re

A (e.g. KPC, GES, SME), Ambler class B (e.g. NDM-1, IMP, VIM, GIM), and Ambler class D (e.g.OXA-48a, OXA-181) [7]. Hence, CRE is categorized as carbapenemase-

lP

producing Enterobacteriaceae (CPE) and non-CPE. What is noteworthy is that CPE is

na

recognized as posing far greater threats to patients than non-CPE, as they may be more virulent and easier to horizontally transmit [8]. Therefore, we focused on seeking

ur

accurate and immediate CPE detection methods. Since CPE emerged into the public consciousness, studies on methods for their

Jo

phenotypic detection have been growing steadily [9]. These phenotypic methods include the modified Hodge’s test (MHT), Carba NP and its variants, the carbapenem inactivation method (CIM) and its variants, the agar-based culture medium containing antibiotic, combined disc method, the automated identification systems, and the matrixassisted laser desorption ionization–time of flight mass spectrometry ((MALDI-TOF5

MS), etc.[10]. However, there is still no single method that meets all the clinical demands perfectly and enables a breakthrough [11]. This systematic review was initiated to summarize four frequently-used phenotypic testing methods of CPE, including MHT, Carba NP, and mCIM, which were recommended as confirmatory test by the Clinical and Laboratory Standards Institute (CLSI) in 2009, 2015, and 2017 [1214], respectively, and MALDI-TOF MS. And we strived to complete the following tasks:

ro of

1. to evaluate the accuracy of these methods; 2. to confirm the applicability of these

methods based on economic factors, human resources, and clinical efficiencies; 3. to propose targeted and personalized selection of CPE detection methods in the context of

Jo

ur

na

lP

re

-p

different carbapenemase gene epidemics.

6

Methods Search strategy: We carried out a comprehensive literature search of PubMed ， EMBASE, Google Scholar, the Cochrane Library, China National Knowledge Infrastructure (CNKI), Sinomed, wan fang database and VIP database until March, 2019.

Keywords

included

“carbapenem-resistant

Enterobacteriaceae”,

OR

“carbapenemase-producing Enterobacteriaceae”, AND “MHT”, OR “Carba NP” OR

ro of

“mCIM” OR “MALDI-TOF MS”. Documentation retrieval was not restricted by test

method, country of publication, date of publication, and research design. In addition, studies from references were searched manually.

-p

Selection criteria: We included studies that evaluated the diagnostic accuracies of

re

phenotypic CPE detection methods, including MHT, Carba NP (non-commercial reagent), mCIM, and MALDI-TOF MS. Exclusion criteria included: 1) studies only

lP

aimed at one species or genus of Enterobacteriaceae; 2) studies only involved one

na

genotype of carbapenemase gene; 3) studies that were not only confined in Enterobacteriaceae, for instance, carbapenemase-producing Acinetobacter baumannii,

ur

carbapenemase-producing Pseudomonas aeruginosa. Data extraction: We designed a data extraction form and followed these steps to

Jo

implement the data extraction: the first author, nationality, year of publication, study design, true positives, false positives, false negatives, true negatives, detection method, and gold standard. Two authors independently screened the literature and extracted data to ensure that the screening core criteria and data gathering were consistent. If the

7

opinions were different, further discussion was carried out or a third person was invited until consensus was reached. Assessment of study quality: The quality of the included studies was evaluated according to Quality Assessment of Diagnostic Accuracy Studies -2 (QUADAS-2). Two authors independently performed the analyses, and consensuses were reached on all decisions.

ro of

Statistical analysis and data synthesis: The pooled sensitivity, specificity, diagnostic

odds ratio, SROC curve, and AUC along with 95% confidence intervals (CIs) for MHT, Carba NP, mCIM, and MALDI-TOF MS in CPE detection were calculated by Stata

-p

(version 15.1 Stata Corp LP, College Station, TX, USA). The I2 and Q-test were used

re

to determine study heterogeneity, and logistic meta-regression was used to identify the potential causes. Pooled results were calculated using a fixed effects model when I2

lP

≤50%, otherwise, a random-effects model was applied. The relative ratio of diagnostic

na

odds ratio was calculated to indicate the indirect comparison of the accuracy of different diagnostic methods by meta-regression analysis using Meta-Disc (version 1.4 Cochrane

ur

Colloquium, Barcelona, Spain) [15]. Publication bias was examined with the Deeks’ funnel plot and Duval's trim and fill method. Sensitivity analysis was conducted by

Jo

Stata 15.1 to evaluate the robustness of the meta-analysis. P value<0.05 indicates significance.

8

Results Description of the eligible studies The literature search processes were shown as in Figure 1. After searching from the different sources, 1359 potentially relevant publications were obtained. Excluding 1122 articles based on title and abstract, or duplication, we initially screened out 237 full-text articles. Of these eligible articles, 183 publications were further excluded, among which

ro of

29 were excluded as they were reviews, letters or case reports; 34 were excluded because their subjects exceeded the scope of CRE; 32 were discarded because they

aimed at only one genus of Enterobacteriaceae or one genotype of Carbapenemase

-p

gene; 88 were rejected because of the absence of a reference method (n=11) or the

re

targeted methods (n=77). 54 articles were ultimately included in the analysis: 19 studies on Carba NP, 4 studies on mCIM, 10 studies on MHT, 7 studies on MS, 3 studies on

lP

the comparison of mCIM vs Carba NP, 1 study on the comparison of Carba NP vs

na

MALDI-TOF MS, and 10 studies on the comparison of Carba NP vs MHT[16-69]. Additionally, in the 54 articles, some authors used both retrospective and prospective

ur

designs or compared with the variants of the same method to conduct research in one paper, so we finally incorporated 67 studies. Table S1 shows the basic characteristics of

Jo

the included studies. Quality assessment

We incorporated 67 studies and completed the assessment of study quality using the QUADAS-2 stool. Figure 2 shows the details of the results. Substantial risk of bias in the patient selection was observed as many of these studies were case–control ones and 9

were lack of consecutive or random samples of enrolled patients. And the high risk of bias in index testing was mainly due to the lack of blinded design and the differences in the conduction or interpretation of the index tests. Evaluation of diagnostic performances of different methods When grouped by methodology, there were 23 studies on MHT, 42 studies on Carba NP, 9 studies on mCIM, and 11 studies on MALDI-TOF MS. According to the I2 and

ro of

Q-test of each method , the pooled effect sizes of MHT, Carba NP, and MALDI-TOF MS were calculated by using random-effects model because of the high heterogeneity,

while that of mCIM was computed by using fixed-effects model (Figure 3-7). The

-p

pooled sensitivity and specificity of MHT were 92% (95%CI: 87%-95%) and 93%

re

(95%CI: 86%-97%), respectively (Figure 3). The pooled sensitivity and specificity of Carba NP were 97% (95%CI: 94%-98%) and 100% (95%CI: 99%-100%), respectively

lP

(Figure 4). Meanwhile, the pooled sensitivity and specificity of mCIM were 99%

na

(95%CI: 99%-100%) and 99% (95%CI: 96%-100%), respectively (Figure 5), and those of MALDI-TOF MS were 99% (95%CI: 96%-100%) and 99% (95%CI: 96%-100%),

ur

respectively (Figure 6).

The pooled diagnostic odds ratios for MHT, Carba NP, mCIM, and MALDI-TOF MS

Jo

were 98.156 (95%CI:48.175-199.995), 1277.710 (95%CI: 751.391-2172.692), 3597.352 (95%CI:1287.575-10000), and 1781.360(95%CI: 651.827-4868.228), respectively. Figure 7 showed the SROC curves of the four different methods in CPE identification. In addition, the pooled AUC values of the SROC for MHT, Carba NP, mCIM and MALDI-TOF MS were 0.97, 1, 1, and 1, respectively. 10

The meta-regression of these four methods on diagnostic odds ratios by pairwise comparison showed that Carba NP, mCIM and MALDI-TOF MS had better diagnostic performances than MHT (p=0.000 for MHT vs Carba NP, p=0.0002 for MADLI-TOFMS vs MHT, p=0.0000 for mCIM vs MHT). And no significant differences were noted among the pairwise comparisons of Carba NP, mCIM and MALDI-TOF MS. Comprehensive comparisons of different methods

ro of

We also summarized the characteristics of these four methods in accuracy, organisms, turnaround time, cost, special requirements, reagents, limitations, inspectors and interpretation of results (Table 1).

-p

Investigations of heterogeneity

re

There were substantial heterogeneities of pooled effect sizes in each method group except mCIM (Figure 3-6). The meta-regression evaluating the effect of confounding

lP

factors on diagnostic odds ratios were performed according to country (North America

na

and Asia or Africa and Europe), blind design or not, carbapenems (meropenem or nonmeropenem for MHT; imipenem or non-imipenem for MALDI-TOF MS),

ur

carbapenemase genes (class B or class A carbapenemase dominated for MHT; class D carbapenemase dominated or not for Carba NP), system (Bruker Daltonics or the

Jo

bioMérieux), sample selection (consecutive or randomized sample enrollment or not), method (CLSI standard test or variants of Carba NP) (Table 2). For Carba NP, an apparent effect was noted in the “method” covariate (p=0.02<0.05). And for MHT, carbapenemase gene subtypes were the main source of heterogeneity (p=0.004<0.05).

11

However, for MALDI-TOF MS, the meta-regression failed to show any statistically significant effects on diagnostic odds ratio. Assessment of reporting bias: The Deeks’ tunnel plot was associated with a p-value of 0.02 for MHT, 0.00 for Carba NP, 0.01 for mCIM, and 0.05 for MADLI-TOF-MS, respectively, suggesting asymmetry for these funnel plots (Figure S1). Then the trim and fill method was applied to assess the potential effects that missing studies may have

ro of

had on the observed results. Results showed that there was no significant influence of

the missing studies on the effect size as demonstrated by imputing the missing studies in the meta-analysis (no new studies added for mCIM and MADLI-TOF-MS).

-p

Sensitivity analyses

re

We investigated the influence of a single study on the overall meta-analysis estimate by sensitivity analyses. Figure S2 showed that there was no individual study that had an

Jo

ur

na

lP

obvious influence, and thus indicated that the results were robust.

12

Discussions MHT was the first method recommended for detecting carbapenemases in Enterobacteriaceae for epidemiological or infection control purposes in the Clinical Laboratory Standard Institute guidelines 2009 [13]. However, it has been removed from the recent guideline (CLSI 2018) [70]. Even though MHT was an inexpensive and easyto-perform tool for CPE detection, there were some limitations with it: the sensitivity

ro of

of MHT was low for NDM-producers, and the specificity was poor due to false positives in the strains with extended-spectrum β-lactamases (ESBLs) or AmpC overexpression

combined with porin loss of outer membrane [16]. The pooled sensitivity, specificity

-p

and diagnostic odds ratio of MHT were not as good as the other three methods in this

re

meta-analysis.

Carba NP, a quick biochemical method for detecting carbapenemase producers based

lP

on the hydrolysis of carbapenems, showed the optimum pooled specificity (100%) on

na

account of excellent ability to differentiate carbapenemase producers from carbapenemresistant strains with other resistance mechanisms [71]. However, the test had an

ur

unsatisfactory sensitivity for class D carbapenemase due to their low enzyme activities [17, 18]. Additionally, some researchers found low sensitivity of Carba NP in

Jo

Enterobacteriaceae harboring GES or in mucoid colonies [22, 72]. The carbapenem inactivation method (CIM) is a growth-based assay which proves carbapenemase existence based on bacterial growth in the presence of carbapenems. As compared to CIM, although mCIM requires both a prolonged incubation time of four hours, and tryptic soy broth to substitute water during the inactivation step [19], its 13

diagnostic efficacy is enhanced consequently. The pooled sensitivity and specificity of mCIM were 99% and 99%, respectively in this meta-analysis. MALDI-TOF MS, originally used for rapid bacterial identification, has been demonstrated as a good method for detecting carbapenemases recently. The pooled sensitivity, specificity and AUC value of the SROC were respectively 99%, 99% and 1, as a par as mCIM. We found no significant differences in the diagnostic odds ratios of

ro of

Carba NP, mCIM, and MALDI-TOF MS.

There was substantial heterogeneity of pooled effect size in each method group except

mCIM. We evaluated the effects of various interfering factors on pooled effect size for

-p

each method. For MHT, carbapenemase gene subtypes were found to be the main source

re

of heterogeneity: the diagnostic performance of MHT in Ambler class B carbapenemase, especially NDM-type, was inferior to its performance in Ambler class A, which were

lP

consistent with the results from previous studies [16]. For Carba NP, we found an

na

apparent effect of the “method” covariate, which means that the result of Carba NP performed in accordance with the CLSI standard, was better than that of Carba NP done

ur

with the agar-based culture medium containing antibiotic or with paper strip. However, we failed to demonstrate statistically significant effects of “carbapenemase gene

Jo

subtypes”, even though Ambler class D carbapenemases were consistently deemed as a difficult subtype for Carba NP. For MALDI-TOF MS, the results of meta-regression based on potential sources of heterogeneity, including carbapenems, system, blinded design, and country, failed to show any statistically significant effects. In spite of that, attention should still be paid on the selection of carbapenems for this approach. 14

Imipenem is often considered a better choice for the following reasons: Imipenem has only two molecular peaks of interest ， whereas ertapenem has eight to ten ， and meropenem has four to six, and fewer peaks facilitate interpretation of results and avoid reading errors of mass spectrum

[20, 73]; In addition, imipenem has superior

sensitivity for OXA-48, and a higher hydrolysis rate which allows for a fast and accurate detection than ertapenem [21].

ro of

In term of costs, human resources, turnaround time, etc., these four methods showed

their advantages and disadvantages. Both MHT and mCIM were cost-saving and of no special reagent and equipment needs, but the over-night incubation led to longer

-p

turnaround time [12, 70, 74]. Carba NP showed a relatively higher cost due to the need

re

of freshly prepared reagents with a shorter expiration date [70, 74]. For MADLI-TOFMS, although the instrument, thorough work flow and well-trained staffs would

na

cost [10, 20, 75].

lP

increase the expense, the spectrometer for routine bacterial identification could split the

The Deeks’ tunnel plot analyses showed asymmetries for Carba NP, MHT, mCIM, and

ur

MADLI-TOF-MS in detecting CPE (p=0.00, 0.02, 0.01 and 0.05, respectively). Then we used trim and fill method to reappraise and adjust the potential bias [76]. No

Jo

apparent difference in the final estimate of the pooled effect size was noticed when compared to the original counterparts. At the same time, combination with the sensitivity analysis by investigating the influence of a single study on the overall metaanalysis estimate suggested that the results were receivable and robust. Funnel plot can detect any effect that is related to sample size, and an asymmetrical graph means a lot 15

more than publication bias [77]. We speculated that the asymmetrical tunnel plot may be due to unsatisfactory study quality and limitations of languages (only available for English and Chinese). Early identification of CRE and differentiation between CPE and non-CPE lead to improved clinical outcomes as well as time and cost savings in the context of the increasing prevalence of carbapenemases [9, 78]. Rather than using PCR-based

ro of

methods, which are cost-intensive and unable to detect novel unidentified genes [10], phenotypical methods look increasingly more likely to be recommended for routine

clinical use. Based on the comprehensive analysis of the four different phenotypic

-p

carbapenemase detection methods in Enterobacteriaceae, we recommended the

re

selection of carbapenemase detection tests in the order of mCIM, Carba NP and MALDI-TOF MS based on their respective diagnostic efficiencies.

lP

The findings of this meta-analysis should be interpreted with caution due to a few

na

limitations. Firstly, we omitted unpublished studies and limited the language of literatures, and therefore the results were susceptible to certain types of bias, such as

ur

publication bias and language bias. Secondly, there were significant heterogeneities in some of the results, still not well explained after exploration of some possible causes.

Jo

Thirdly, indirect comparison of the diagnostic accuracies of these four methods lacked preciseness. Further studies to analyze the comprehensive diagnostic capabilities of these technologies by including studies on direct comparison or network meta-analysis are required. Fourthly, this meta-analysis did not provide the pooled effect size of these

16

methods in certain carbapenemase genes, and the performances of the above methods at the molecular level were inconclusive.

Declarations Funding: This study was supported in part by the National Natural Science Foundation of China (Grant No. 81772239 and 31500749), the Science Foundation of Chongqing

ro of

(No. cstc 2016jcyjA0248 and No. cstc2015jcyjA10102), the Science and Technology Research Program of Chongqing Municipal Education Commission (Grant No.

KJ1500235 and KJ1702022), and the Medical Research Program of Chongqing Health

-p

and Family Planning Commission (No. 2018MSXM009 and 2016MSXM001).

Jo

ur

na

lP

Competing Interests: None.

re

Ethical Approval: Not required.

17

References: [1] Falagas ME, Tansarli GS, Karageorgopoulos DE, Vardakas KZ. Deaths attributable to

carbapenem-resistant

Enterobacteriaceae

infections.

Emerg

Infect

Dis.

2014;20:1170-5. [2] ElMahallawy HA, Zafer MM, Amin MA, Ragab MM, Al-Agamy MH. Spread of carbapenem resistant Enterobacteriaceae at tertiary care cancer hospital in Egypt. Infect

ro of

Dis (Lond). 2018;50:560-4.

[3] Mitgang EA, Hartley DM, Malchione MD, Koch M, Goodman JL. Review and

mapping of carbapenem-resistant Enterobacteriaceae in Africa: Using diverse data to

-p

inform surveillance gaps. Int J Antimicrob Agents. 2018;52:372-84.

re

[4] Zhang Y, Wang Q, Yin Y, Chen H, Jin L, Gu B, et al. Epidemiology of CarbapenemResistant Enterobacteriaceae Infections: Report from the China CRE Network.

lP

Antimicrobial Agents and Chemotherapy. 2018;62:e01882-17.

na

[5] Jean SS, Lee NY, Tang HJ, Lu MC, Ko WC, Hsueh PR. Carbapenem-Resistant Enterobacteriaceae Infections: Taiwan Aspects. Front Microbiol. 2018;9:2888.

ur

[6] Kopotsa, K., Osei Sekyere, J. & Mbelle, N. M. Plasmid Evolution in Carbapenemase-Producing Enterobacteriaceae: A Review. Ann. N. Y. Acad. Sci. 2019

Jo

Aug 30. doi: 10.1111/nyas.14223. [Epub ahead of print] [7] Somboro AM, Osei Sekyere J, Amoako DG, Essack SY, Bester LA. Diversity and proliferation of metallo-β-lactamases: a clarion call for clinically effective metallo-βlactamase inhibitors. Appl Environ Microbiol. 2018; 31; 84(18). pii: e00698-18.

18

[8] Tamma PD, Goodman KE, Harris AD, Tekle T, Roberts A, Taiwo A, et al. Comparing the Outcomes of Patients With Carbapenemase-Producing and NonCarbapenemase-Producing Carbapenem-Resistant Enterobacteriaceae Bacteremia. Clin Infect Dis. 2017;64:257-64. [9] Osei Sekyere J, Govinden U, Essack SY. Review of established and innovative detection methods for carbapenemase-producing Gram-negative bacteria. J Appl

ro of

Microbiol. 2015; 119(5): 1219–33.

[10] Hammoudi D, Moubareck CA, Sarkis DK. How to detect carbapenemase producers? A literature review of phenotypic and molecular methods. J Microbiol

-p

Methods. 2014;107:106-18.

re

[11] Lutgring JD, Limbago BM. The Problem of Carbapenemase-ProducingCarbapenem-Resistant-Enterobacteriaceae Detection. J Clin Microbiol. 2016;54:529-

lP

34.

na

[12] CLSI. Performance standards for antimierobial susceptibility testing; Twentyseventh information supplement. CLSI document M100-S27 [S]. Wayne, PA:SLSI,

ur

2017.

[13] CLSI. Performance standards for antimierobial susceptibility testing; Nineteenth

Jo

information supplement. CLSI document M100-S19[S]. Wayne, PA: CLSI, 2009. [14] CLSI.Performance standards for antimierobial susceptibility testing; Twenty-fifth information supplement. CLSI document M100-S25[S]. Wayne, PA: CLSI, 2015. [15] Jingling W, Jipin L, Long G, Xiaoyan L, Junhua Z, Jinhui T. Comparation of accuracy of different diagnostic tests: an introduction of network meta-analysis 19

methods. CHINESE JOURNAL OF EVIDENCE-BASED MEDICINE. 2017;17:98792. [16] Kuchibiro T, Komatsu M, Yamasaki K, Nakamura T, Nishio H, Nishi I, et al. Evaluation of the modified carbapenem inactivation method for the detection of carbapenemase-producing Enterobacteriaceae. J Infect Chemother. 2018;24:262-6. [17] Ong CH, Ratnayake L, Ang MLT, Lin RTP, Chan DSG. Diagnostic Accuracy of

ro of

BD Phoenix CPO Detect for Carbapenemase Production in 190 Enterobacteriaceae Isolates. J Clin Microbiol. 2018;56:e01043-18.

[18] Laolerd W, Akeda Y, Preeyanon L, Ratthawongjirakul P, Santanirand P.

-p

Carbapenemase-Producing Carbapenem-Resistant Enterobacteriaceae from Bangkok,

re

Thailand, and Their Detection by the Carba NP and Modified Carbapenem Inactivation Method Tests. Microb Drug Resist. 2018;24:1006-11.

lP

[19] Butler-Wu SM, Abbott AN. Is This the Carbapenemase Test We've Been Waiting

na

for? A Multicenter Evaluation of the Modified Carbapenem Inactivation Method. J Clin Microbiol. 2017;55:2309-12.

ur

[20] Choquet M, Guiheneuf R, Castelain S, Cattoir V, Auzou M, Pluquet E, et al. Comparison of MALDI-ToF MS with the Rapidec Carba NP test for the detection of

Jo

carbapenemase-producing Enterobacteriaceae. Eur J Clin Microbiol Infect Dis. 2018;37:149-55. [21] Monteferrante CG, Sultan S, Ten Kate MT, Dekker LJ, Sparbier K, Peer M, et al. Evaluation of different pretreatment protocols to detect accurately clinical

20

carbapenemase-producing Enterobacteriaceae by MALDI-TOF. J Antimicrob Chemother. 2016;71:2856-67. [22] Sekyere J, Govinden U, Essack S. Comparison of Existing Phenotypic and Genotypic Tests for the Detection of NDM and GES Carbapenemase- Producing Enterobacteriaceae. Journal of Pure and Applied Microbiology. 2016;10:2585-91. [23] Muntean MM, Muntean AA, Gauthier L, Creton E, Cotellon G, Popa MI, et al.

ro of

Evaluation of the rapid carbapenem inactivation method (rCIM): a phenotypic

screening test for carbapenemase-producing Enterobacteriaceae. J Antimicrob Chemother. 2018;73:900-8.

-p

[24] Yamada K, Kashiwa M, Arai K, Nagano N, Saito R. Evaluation of the modified

re

carbapenem inactivation method and sodium mercaptoacetate-combination method for the detection of metallo-beta-lactamase production by carbapenemase-producing

lP

Enterobacteriaceae. J Microbiol Methods. 2017;132:112-5.

na

[25] Yan L, Zhai Q, Gu B. Evaluation of the mCIM for the detection of carbapenemaseproducing strains. Chinese Journal of Laboratory Medicine. 2018;41:321-3.

ur

[26] Yulan M, Wenjie S, Jihong L, Shiying F, Qian S, Lin Z, et al. Evaluation for effectiveness: Mcim and eCIM for screening carbapenemase in Enterobacteriaceae.

Jo

Chinese Journal of Clinical Laboratory Science. 2018;36:15-9. [27] Yang J, Niu C, Zhao J. Evaluation of clinical application value of CIM in phenotypic detection of carbapenemase in Enterobacteriaceae. Chinese Journal of Laboratory Medicine. 2016;39:969-71.

21

[28] Wang C. Comparison on detection methods of phenotype of carbapenemases. Chinese Journal of Mordern Drug Application. 2017;7:93-5. [29] Yang Y, Min Z, Fu F. Study on the clinical application of two experimental methods for carbapenemase-producing Enterobacteriaceae. System Medicine. 2017;2:9-12. [30] Cao L, Li X, Zhou W, Zhou H, Zhu J, Zhang Z, et al. Comparison of carbapenem inactivation method and Carba NP test for screening of carbapenemase-producing

ro of

Enterobacteriaceae. Chinese Journal of Clinical Laboratory Science. 2016;34:646-9.

[31] Hu R, Yang Z, Hu X. Performance evalucation of Carba NP test and CIM in

detecting carbapenemase among Enterobacteriaceae. Occup and Health 2016;32:3208-

-p

13.

re

[32] Hu X, Hu R, Liu X. Clinical evaluation of Carba NP paper strip test in detection of carbapenemase-producing Enterobacteriaceae. J Cl in Pathol Res. 2017;37:2043-9.

lP

[33] Yang Y, Chen J, Sun Q, Di Lin, Cheng J. Evaluation of the Carba NP test to detect

na

carbapenemase-producing Enterobacteriaceae. Chinese Journal of Laboratory Medicine. 2016;39:99-102.

ur

[34] Ma W, Xu X, Sun K, Wang L, Zhang L. Methodological evaluation of Carba NP test and modified Hodge test in the detection of carbapenemase producing

Jo

Enterobacteriaceae strains. Chinese Journal of Microbiology and Immunology. 2016;36:64-5.

[35] Chen K, He J, Feng J, Li Y, Zhang J. Evaluation of the carbapenem inactivation method for the detection of carbapenemase-producing strains. Chinese Journal of Clinical Laboratory Science 2017;40:64-6. 22

[36] Segawa T, Matsui M, Suzuki M, Tsutsui A, Kuroda M, Shibayama K, et al. Utilizing the Carba NP test as an indicator of expression level of carbapenemase genes in Enterobacteriaceae. J Microbiol Methods. 2017;133:35-9. [37] Surre J, Canard I, Bourne-Branchu P, Courbiere E, Franceschi C, Chatellier S, et al. Enhanced detection of carbapenemase-producing Enterobacteriaceae by an optimized phenol red assay. Diagn Microbiol Infect Dis. 2018;90:11-7.

ro of

[38] Hartl R, Widhalm S, Kerschner H, Apfalter P, Gattringer R. Cilastatin does not

affect Carba NP test performance for detection of carbapenemase production in Enterobacteriaceae. Wien Klin Wochenschr. 2017;129:29-32.

-p

[39] Gauthier L, Bonnin RA, Dortet L, Naas T. Retrospective and prospective

re

evaluation of the Carbapenem inactivation method for the detection of carbapenemaseproducing Enterobacteriaceae. PLoS One. 2017;12:e0170769.

lP

[40] Boutal H, Naas T, Devilliers K, Oueslati S, Dortet L, Bernabeu S, et al.

na

Development and Validation of a Lateral Flow Immunoassay for Rapid Detection of NDM-Producing Enterobacteriaceae. J Clin Microbiol. 2017;55:2018-29.

ur

[41] Yamada K, Kashiwa M, Arai K, Nagano N, Saito R. Comparison of the ModifiedHodge test, Carba NP test, and carbapenem inactivation method as screening methods

Jo

for carbapenemase-producing Enterobacteriaceae. J Microbiol Methods. 2016;128:4851.

[42] Srisrattakarn A, Lulitanond A, Wilailuckana C, Charoensri N, Wonglakorn L, Piyapatthanakul S, et al. Modification and evaluation of the Carba NP test by use of

23

paper

strip

for

simple

and

rapid

detection

of

carbapenemase-producing

Enterobacteriaceae. World J Microbiol Biotechnol. 2016;32:117. [43] Lifshitz Z, Adler A, Carmeli Y. Comparative Study of a Novel Biochemical Assay, the

Rapidec

Carba

NP

Test,

for

Detecting

Carbapenemase-Producing

Enterobacteriaceae. J Clin Microbiol. 2016;54:453-6. [44] Garcia-Fernandez S, Morosini MI, Gijon D, Beatobe L, Ruiz-Garbajosa P,

ro of

Dominguez L, et al. Detection of Carbapenemase Production in a Collection of Enterobacteriaceae with Characterized Resistance Mechanisms from Clinical and

Environmental Origins by Use of Both Carba NP and Blue-Carba Tests. J Clin

-p

Microbiol. 2016;54:464-6.

re

[45] Dortet L, Jousset A, Sainte-Rose V, Cuzon G, Naas T. Prospective evaluation of the OXA-48 K-SeT assay, an immunochromatographic test for the rapid detection of

lP

OXA-48-type carbapenemases. J Antimicrob Chemother. 2016;71:1834-40.

na

[46] Bogaerts P, Yunus S, Massart M, Huang TD, Glupczynski Y. Evaluation of the BYG Carba Test, a New Electrochemical Assay for Rapid Laboratory Detection of

ur

Carbapenemase-Producing Enterobacteriaceae. J Clin Microbiol. 2016;54:349-58. [47] Yusuf E, Van Der Meeren S, Schallier A, Pierard D. Comparison of the Carba NP

Jo

test with the Rapid CARB Screen Kit for the detection of carbapenemase-producing Enterobacteriaceae and Pseudomonas aeruginosa. Eur J Clin Microbiol Infect Dis. 2014;33:2237-40. [48] Knox J, Jadhav S, Sevior D, Agyekum A, Whipp M, Waring L, et al. Phenotypic detection of carbapenemase-producing Enterobacteriaceae by use of matrix-assisted 24

laser desorption ionization-time of flight mass spectrometry and the Carba NP test. J Clin Microbiol. 2014;52:4075-7. [49] Dortet L, Brechard L, Poirel L, Nordmann P. Rapid detection of carbapenemaseproducing Enterobacteriaceae from blood cultures. Clin Microbiol Infect. 2014;20:3404. [50] Huang TD, Berhin C, Bogaerts P, Glupczynski Y. Comparative Evaluation of Two

ro of

Chromogenic Tests for Rapid Detection of Carbapenemase in Enterobacteriaceae and

in Pseudomonas aeruginosa Isolates. Journal of Clinical Microbiology. 2014;52:30603.

-p

[51] Bernabeu S, Dortet L, Naas T. Evaluation of the beta-CARBA test, a colorimetric

re

test for the rapid detection of carbapenemase activity in Gram-negative bacilli. J Antimicrob Chemother. 2017;72:1646-58.

lP

[52] Shalom O, Adler A. Comparative study of 3 carbapenem-hydrolysis methods for

na

the confirmation of carbapenemase production in Enterobacteriaceae. Diagn Microbiol Infect Dis. 2018;90:73-6.

ur

[53] Pires J, Tinguely R, Thomas B, Luzzaro F, Endimiani A. Comparison of the inhouse made Carba-NP and Blue-Carba tests: Considerations for better detection of

Jo

carbapenemase-producing Enterobacteriaceae. J Microbiol Methods. 2016;122:33-7. [54] Xing S, Men K. Application value of three methods in detecting carbapenemresistant Enterobacteriaceae. Jiangsu Med J. 2018;44:1182-5.

25

[55] Bir R, Mohapatra S, Kumar A, Tyagi S, Sood S, Das BK, et al. Comparative evaluation of in-house Carba NP test with other phenotypic tests for rapid detection of carbapenem-resistant Enterobacteriaceae. J Clin Lab Anal. 2019;33:e22652. [56] Yu J, Liu Y, Yu J, Li Y, Liu J. Efficiency evaluation of carbapenem inactivation method for screening carbapenemase-producing Enterobacteriaceae. Laboratory Medicine. 2017;32:628-32.

ro of

[57] Yang Q, Zheng R, Wang H, Chen M, Xu Y, Wu W, et al. Performance of modified Hodge test on the detection of carbapenemase among Enterobacteriaceae. Chinese Journal of Clinical Laboratory Science. 2010;33:1122-7.

-p

[58] Lu Y, Cheng L, Zhao Z. Methology camparison among three tests for detection of

re

carbapenemases. Maternal and Child Care of China. 2016;31:2173-5. [59] Doyle D, Peirano G, Lascols C, Lloyd T, Church DL, Pitout JD. Laboratory

na

2012;50:3877-80.

lP

detection of Enterobacteriaceae that produce carbapenemases. J Clin Microbiol.

[60] Bartolini A, Frasson I, Cavallaro A, Richter SN, Palù G. Comparison of phenotypic

ur

methods for the detection of carbapenem non-susceptible Enterobacteriaceae. Gut Pathogens. 2014;6:13.

Jo

[61] Saito R, Koyano S, Dorin M, Higurashi Y, Misawa Y, Nagano N, et al. Evaluation of a simple phenotypic method for the detection of carbapenemase-producing Enterobacteriaceae. J Microbiol Methods. 2015;108:45-8.

26

[62] Yan YZ, Sun KD, Pan LH, Fan HQ, Yang HZ, Lu YC, et al. A screening strategy for phenotypic detection of carbapenemase in the clinical laboratory. Can J Microbiol. 2014;60:211-5. [63] Ambretti S, Gaibani P, Berlingeri A, Cordovana M, Tamburini MV, Bua G, et al. Evaluation of phenotypic and genotypic approaches for the detection of class A and class B carbapenemases in Enterobacteriaceae. Microb Drug Resist. 2013;19:212-5.

ro of

[64] Bae IK, Kang HK, Jang IH, Lee W, Kim K, Kim JO, et al. Detection of

Carbapenemases in Clinical Enterobacteriaceae Isolates Using the VITEK AST-N202 Card. Infect Chemother. 2015;47:167-74.

-p

[65] Kost K, Yi J, Rogers B, Jerris R. Comparison of clinical methods for detecting

re

carbapenem-resistant Enterobacteriaceae. Pract Lab Med. 2017;8:18-25. [66] Lasserre C, De Saint Martin L, Cuzon G, Bogaerts P, Lamar E, Glupczynski Y, et

lP

al. Efficient Detection of Carbapenemase Activity in Enterobacteriaceae by Matrix-

na

Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry in Less Than 30 Minutes. J Clin Microbiol. 2015;53:2163-71.

ur

[67] Dortet L, Tande D, de Briel D, Bernabeu S, Lasserre C, Gregorowicz G, et al. MALDI-TOF for the rapid detection of carbapenemase-producing Enterobacteriaceae:

Jo

comparison of the commercialized MBT STAR(R)-Carba IVD Kit with two in-house MALDI-TOF techniques and the RAPIDEC(R) CARBA NP. J Antimicrob Chemother. 2018;73:2352-9.

27

[68] Yu J, Liu J, Li Y, Yu J, Zhu W, Liu Y, et al. Rapid detection of carbapenemase activity of Enterobacteriaceae isolated from positive blood cultures by MALDI-TOF MS. Ann Clin Microbiol Antimicrob. 2018;17:22. [69] Feng L, Li C. Detection of carbapenemase-producing Enterobacteriaceae by matrix-assisted laser desorption/ ionization time of flight mass spectrometry (MALDITOF MS). Chin J Microbiol Immunol. 2018;38:218-25.

ro of

[70]CLSI. Performance standards for antimierobial susceptibility testing; Twentyeighth information supplement. CLSI document M100-S28[S]. Wayne, PA: CLSI, 2018

[71] Nordmann P, Poirel L, Dortet L. Rapid detection of carbapenemase-producing

-p

Enterobacteriaceae. Emerg Infect Dis. 2012;18:1503-7.

re

[72] Osei Sekyere J, Govinden U, Essack SY. Comparison of Existing Phenotypic and Genotypic Tests for the Detection of NDM and GES Carbapenemase- Producing

lP

Enterobacteriaceae. J Pure Appl Microbio. 2016; 10: 2585–91.

na

[73] Sparbier K, Schubert S, Weller U, Boogen C, Kostrzewa M. Matrix-assisted laser desorption ionization-time of flight mass spectrometry-based functional assay for rapid

37.

ur

detection of resistance against beta-lactam antibiotics. J Clin Microbiol. 2012;50:927-

Jo

[74] Tamma PD, Opene BNA, Gluck A, Chambers KK, Carroll KC, Simner PJ. Comparison of 11 Phenotypic Assays for Accurate Detection of CarbapenemaseProducing Enterobacteriaceae. Journal of Clinical Microbiology. 2017;55:1046-55.

28

[75] Calderaro A, Buttrini M, Piergianni M, Montecchini S, Martinelli M, Covan S, et al. Evaluation of a modified meropenem hydrolysis assay on a large cohort of KPC and VIM carbapenemase-producing Enterobacteriaceae. PLoS One. 2017;12:e0174908. [76] HR R, AJ S, M B. Publication Bias in Meta-Analysis Prevention, Assessment and Adjustments Chichester: John Wiley ＆ Sons Ltd. 2005:p.127-44. [77] Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and

J Clin Epidemiol. 2005;58:882-93.

ro of

other sample size effects in systematic reviews of diagnostic test accuracy was assessed.

[78] Lapointe-Shaw L, Voruganti T, Kohler P, Thein HH, Sander B, McGeer A. Costanalysis

of

universal

screening

for

carbapenemase-producing

-p

effectiveness

re

Enterobacteriaceae in hospital inpatients. Eur J Clin Microbiol Infect Dis.

Jo

ur

na

lP

2017;36:1047-55.

29

Jo

ur

na

lP

re

-p

ro of

Yang et al., Figure 1

30

Jo

ur

na

lP

re

-p

ro of

Yang et al., Figure 1

31

re

-p

ro of

Yang et al., Figure 3

Jo

ur

na

lP

Yang et al., Figure 4

32

ro of -p re

Jo

ur

na

lP

Yang et al., Figure 5

33

ro of -p re

Jo

ur

na

lP

Yang et al., Figure 6

34

ro of -p re

Jo

ur

na

lP

Yang et al., Figure 7

35

36

ro of

-p

re

lP

na

ur

Jo

Table 1 Comprehensive comparison of different methods Parameter Accuracy

(%;

Carba NP (13,17)

MHT (12,17)

mCIM (13,17)

MADLI-TOF-MS (11)

97;100

92;93

99;99

99;99

Enterobacterceae

Enterobacterceae

Enterobacterceae

that

, P. aruginosa that

are not susceptible to

are

one

sensitivity, specificity) organisms

Enterobacterceae,

P.

aeruginosa,

and

are

not

Acinetobacter spp. that are

susceptible

to

not susceptible to one or

one

more carbapenems

carbapenems

or

more

not

susceptible one

or

to

or

that

more

carbapenems

more

carbapenems 16-20h

22-28h

Cost

$2

<$1

<$1

Special

pH meter

none

none

time

requirements Commercially

available

no

special

special

α-cyano-4

regents or media

regents or media

hydroxycinnamic acid

reagent in Tris HCl buffer;

necessary

necessary

(HCCA)；

False-positive

Only applies to

Lack of standardization.

instant

results

Enterobacteriace

It’s

have a limited shelf life.

occur in isolates

ae.

establish the detection

False-negative

that produce

Over-night

system and determine

ESBL or AmpC

incubation.

the cut-off ratio prior to

preparation

and

lP

results

re

Some of reagent need

occur in

OXA ， GES

carbapenemases

and

enzymes

along

with porin loss.

Invalid results occur in

False-negative

some isolates.

results occur in

na

mucoid colony.

ur Jo

MALDI-TOF-MS

bacterial protein extraction

solution A; solution B; Limitations

no

<$1

-p

reagents

0.5-1h

ro of

2h

Turnaround

necessary

to

routine implementation.

isolates producing NDM carbapenemase). Over-night incubation

Inspectors

Simply-trained

Simply-trained

Simply-trained

Well-trained

Interpretation

Color change; subjective

Zone

Zone

The peak intensity ratio

of

results;

diameter;

subjective

diameter;

subjective

calculated by software; objective

authenticity

37

Table 2 Meta-regression Evaluation of the Effects of Confounding Factors on Diagnostic Odds Ratio Coefficient

RDOR (95%CI)

p-value

Blinded design

0.925

2.52(0.81;7.88)

0.08

Method

1.875

6.52(1.33;25.99)

0.02

Sample selection

-0.305

0.74(0.23;2.33)

0.59

Country

-0.125

0.88(0.24;3.28)

0.85

Gene

-3.02

0.74(0.19;2.81)

0.65

Carbapenems

0.57

1.77(0.37;8.56)

0.45

Gene

2.24

9.42(2.27;39.11)

0.004

Sample selection

-0.49

0.61(0.12;3.15)

0.54

Country

1.245

3.47(0.40;29.84)

0.24

Carbapenems

0.56

1.76(0.00;1611.89)

0.84

System

0.57

1.77(0.00;182350.63)

0.93

Blinded design

-0.74

0.48(0.00;529.09)

0.80

Country

-0.60

0.55(0.00;88.51)

0.77

Parameters Carba NP

ro of

MHT

Jo

ur

na

lP

re

-p

MALDI-TOF-MS

38