https://orcid.org/0000-0003-3876-8229
Abstract
Carbapenemase-producing Enterobacterales (CPE) are of international concern. Screening for CPE can encompass single- or two-step approaches, using culture, PCR or a combination. Each approach has benefits, but none are without disadvantage.
To reflect on the challenges and implications of PCR-positive/culture-negative CPE screening results and assess if PCR cycle threshold (Ct) value can be helpful in predicting positive culture results.
Risk factor-based CPE screening swabs were tested using a two-step algorithm: PCR followed by culture of PCR-positive specimens. Data on all PCR-positive specimens between 1 August 2022 and 31 May 2024 were extracted. ORs were estimated using receiver operating characteristic (ROC) curves to compare Ct values and culture. Youden’s index was calculated to establish the optimal Ct cut-off value. Compliance with the CPE screening pathway for newly identified CPE PCR-positive patients was assessed.
Of 61 268 CPE screens, 292 were PCR positive (0.5%), with 298 genes identified. Of these, 81.5% were culture confirmed. ROC analysis showed an AUC of 0.82 and Youden’s index yielding a Ct cut-off value of 33.7. Repeat CPE screens were obtained from 33 new PCR-positive, culture-negative inpatients. Further investigation was not possible for 17 new PCR-positive, culture-negative patients (11%).
Molecular platforms alone cannot detect species or antimicrobial resistance. A molecular-followed-by-culture algorithm can reduce workloads associated with culture, resulting in comprehensive data, including antimicrobial susceptibility results, to support informed clinical, policy and epidemiological decision-making.
Introduction
Carbapenemase-producing Enterobacterales (CPE) are a significant threat to public health, with recommended control measures including screening for CPE carriage to prevent cross-transmission.1 The WHO and the ECDC recommend risk factor-based screening.1,2 In Australia and the UK, screening approaches based on CPE epidemiology are advised.3,4 Universal admission screening may detect more CPE carriers earlier, thus reducing their exposure to other patients. However, such an approach would not be viable where there is a low CPE prevalence and/or available resources are limited.5,6
Regardless of CPE screening approaches, the laboratory methodology requires significant consideration. Laboratories should be able to detect carbapenem resistance in a timely manner and with high sensitivity at the point of care.7 There are several approaches available, including single- or two-step algorithms using culture, PCR or a combination of both. While each approach has benefits, none are without disadvantage.8,9 Where a single-step PCR approach is used, criteria for reporting a positive specimen need careful consideration.6
In 2011, CPE became a notifiable disease in Ireland. In 2017, CPE was declared a National Public Health Emergency.10 The local CPE epidemiology and laboratory testing algorithm from 2011 to 2019 has been described elsewhere.11 In August 2022, the laboratory algorithm for CPE screening was updated to a two-step screening approach, PCR on all specimens, followed by culture of positive specimens only, with the aim of improving workflow and achieving efficiencies.
Here we reflect on the challenges and implications of PCR-positive/culture-negative CPE screening results and assess if the PCR cycle threshold (Ct) value can be helpful in predicting positive CPE culture results.
Methods
Study design
A retrospective observational cohort study investigating CPE laboratory data between 1 August 2022 and 31 July 2024 was undertaken following local Ethics Committee approval (REC reference 23/26), and STROBE guidelines were followed.12
Setting
Beaumont Hospital is an 820-bed adult tertiary referral centre providing specialist services such as neurosurgery, kidney transplantation and cochlear implantation, along with acute care services to the local catchment area of 290 000 people. Most accommodation is multi-occupancy, with 136 single rooms, 77% with en suite facilities, and 12 airborne isolation rooms.
Population
A risk factor-based admission CPE screening policy is used (Supplementary material, available as Supplementary data at JAC Online). Screening compliance is monitored monthly. Once newly identified as CPE colonized, a patient’s record is electronically flagged indefinitely. On readmission, the flag identifies the patient’s CPE status, the immediate need for transmission-based precautions, including single-room placement. The screening pathway and actions taken are outlined in Figure 1.
CPE screening pathway and actions arising from a PCR-positive and culture-negative result.
Data source and variables
The LightMix® Modular carbapenemase panel was used for the molecular detection of carbapenemase genes directly from rectal swabs. The procedure was performed using MagNA Pure 96 (for extraction) and Light Cycler 480 II using Roche FLOW software®, which included targets for OXA-48, NDM, KPC, IMP and VIM genes. Ct threshold values were determined during local validation processes prior to implementation. Specimens with a Ct of ≥30 and ≤35 were confirmed using the Xpert® Carba-R assay. A Ct of >35 was reported as ‘CPE DNA not detected’ and no further follow-up of that patient was required. PCR-positive specimens were inoculated directly on CPE selective chromogenic agar, chromID® CARBA SMART or Colorex mSuperCARBA. Growth of Enterobacterales was identified using MALDI-TOF ID MS and the presence of CPE confirmed using a lateral flow immunoassay, NG-test CARBA 5.
Statistical methods
Data on all PCR-positive CPE screening specimens between 1 August 2022 and 31 July 2024 were extracted from the laboratory information system. Ct values were accessed from Flow® software. To study the association of the Ct value and culture result, the ORs, both crude and adjusted by age, sex and specialty, and their 95% intervals, were estimated using logistic regression. Receiver operating characteristic (ROC) curves were then calculated comparing Ct values and culture outcome. Youden’s index was calculated to establish the optimal Ct cut-off value to confirm/rule out culture positivity.13 All analyses were performed using Stata version 17.0. Associations with a P value of <0.05 were considered statistically significant.
Results
Of 61 268 CPE screens, 292 (0.5%) were PCR positive. Most positive screens had one CPE gene detected (n = 289; 99%). Three screens had three genes detected. A total of 298 CPE genes were detected, consisting of OXA-48 (n = 248; 83%), NDM (n = 26; 9%), KPC (n = 16; 5%), IMP (n = 5; 2%) and VIM (n = 3; 1.0%).
Most screens (81.5%) were subsequently CPE-confirmed by culture and lateral flow immunoassay NG-test CARBA 5. The proportion that were culture positive was similar across carbapenemases OXA-48, NDM and KPC (average 82%). Only two of the five IMP carbapenemase positives were culture positive and none of the three VIM carbapenemase positives were confirmed by culture. Escherichia coli (42%), Klebsiella species (20%), Enterobacter species (14%) and Citrobacter species (5%) were the most common CPE cultured.
The 292 specimens were taken from 177 patients. On review of each patient’s CPE status, 13% were previously known to be CPE colonized. Of the 154 new PCR-positive patents, 121 (79%) were CPE culture-confirmed and 33 (21%) were culture negative. Of the 33 culture-negative patients, four were confirmed as CPE on repeat screening and five were reported as probable CPE (i.e. second screen was PCR positive, but culture negative).
Seven patients had CPE culture-negative stool specimens and CPE colonization ruled out. Eight patients had follow-up (i.e. a second screen was PCR negative but a stool specimen for culture was not sent before discharge) and nine were discharged before any repeat specimen could be taken. Overall the CPE status was determined for 16 of the 33 patients and the CPE status of the other 17 patients was indeterminate at discharge.
The interquartile range of Ct values for culture-negative specimens was 5.36 (p25 = 28.3, median = 31.8, p75 = 33.7) while for culture-positive specimens, it was 9.89 (p25 = 19.3, median = 24.8, p75 = 29.2). Adjusted logistic regression showed the Ct value to be significantly associated with a positive culture result (P < 0.001, OR 0.76, 95% CI 0.7–0.8). The ROC curve for CPE Ct as a test for detection of culture-positive specimens is shown in Figure 2, with an AUC of 0.82. Youden’s index yielded a Ct cut-off value of 33.7. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for Cts were 95.5%, 50.9%, 89% and 73%, respectively.
ROC analysis for prediction of a positive CPE culture result according to PCR Ct value.
Discussion
There are varying approaches to CPE screening, with local policies informed by guidelines, epidemiology and resources. The use of PCR has benefits including rapid results, less labour and higher sensitivity.14 It has been reported that implementation of a Ct cut-off to help differentiate false-positive or epidemiologically insignificant findings for carbapenemase-producing organisms (CPOs) is clinically useful, with Ct values of ≤35 confirmed by culture in 96% of cases.15 This study specifically involved CPE only, and lower Ct values were significantly associated with a positive culture result, with a Ct of ≤33.7 more likely to result in a positive CPE culture. This difference in Ct values may relate to differences in the molecular platforms used, other aspects of laboratory methodology, the population screened, or the organisms being screened for, i.e. CPOs and not CPE.
In our study we found the performance metrics of PCR plus culture for CPE screening effective in identifying true positive cases. This reliability was underpinned by the PPV, providing reassurance in terms of appropriate use of isolation resources, and even possibly on the choice of empirical antibiotic treatment in the event of infection. Confirmatory testing by lateral flow immunoassay on culture results reduces the potential impact of false positives. In our study only 0.5% of samples underwent confirmatory steps. A cost-effectiveness analysis may be helpful in certain circumstances to assess the costs of culturing CPE positives compared with costs associated with isolating patients that are not actually CPE colonized, to justify this approach.
Current molecular platforms do not report the bacterial species carrying the carbapenemase gene, or the presence of additional antimicrobial resistance (AMR).16 Even though WGS adds more comprehensive data on resistance genes and mutations, the lack of AMR data remains. There are several concerns that arise in the absence of culture confirmation. The lack of an antibiogram impedes selecting effective treatment should CPE infection arise. Empirical prescribing guidelines are best underpinned by local antibiogram data.17 Reported rates of CPE bloodstream infection among CPE-colonized patients range between 5.8% and 12.0%.18,19 Also, the monitoring and tracking of emerging strains of CPE, such as hypervirulent Klebsiella pneumoniae ST23 CPE are dependent on isolate analysis.20 Molecular approaches should be used in tandem with traditional culture methods to ensure molecular results can be translated to phenotypic findings on which future treatment and local antibiotic policies can be based.
The addition of PCR to CPE detection provides greater comprehensiveness to developing a standardized CPE case definition, which is central to any robust surveillance, and associated response system. A consensus on a case definition incorporating the different CPE testing algorithms, i.e. molecular positive/culture negative or molecular only, would provide a standardized framework from which true incidence can be reported. In the interim, a local surveillance pathway was developed whereby a definitive CPE status was obtained for 89% of all new CPE PCR-positive patients, ensuring a clear pathway for patients on readmission and efficient use of limited isolation room resources. The 17 patients (11%) with indeterminate CPE status at discharge are targeted for screening according to a risk factor-based assessment (Supplementary material) should they be readmitted. However, these are less of a priority for patient isolation if single rooms are not available.
Limitations of this study include it being a single-centre study, and as information on the quality of rectal samples was not assessed, poor sampling cannot be out ruled as a potential source of false-negative results. However, the strengths of this study are the robustness of analysed data with prospective follow-up of patients via the CPE screening pathway.
While there are additional considerations when using a two-step CPE screening algorithm of molecular then culture, we found only 0.5% of screens in our study required a culture step; all others were PCR negative. This approach provides comprehensive microbiological and epidemiological information, reduces the amount of medical scientist time, and ensures a better quality of care for our patients.
Acknowledgements
We are grateful for the assistance of the staff of the Surveillance, Microbiology, Infection Prevention and Control Departments. We are also grateful for the assistance of John Heritage, University Teaching Fellow and Senior Lecturer in Microbiology (retired), Faculty of Biological Sciences, University of Leeds, and patient representative on the European Study Group on Clostridioides difficile (ESGCD).
Funding
M.S. conducted this work with financial support from the Irish Research Council (IRC) (EBPPG/2022/196).
Transparency declarations
H.H. has received a research grant from Pfizer in recent years, but not related to CPE. He has also received professional fees from the Scottish Hospital Enquiry and the Bons Secours Hospital Group (Ireland). F.F. has received research grant support and been a recipient of a consultancy fee from Tillots Pharma. All other authors: none to declare.
Supplementary data
Supplementary material is available as Supplementary data at JAC Online.
