Perinatal colonization with extended spectrum beta-lactamase-producing and carbapenem-resistant Gram-negative bacteria: a hospital-based cohort study

Please read the article at:

https://doi.org/10.1186/s13756-024-01366-9

Disclaimer: Below is an interpretation of the article thus do not necessarily represent the views of the authors.

Introduction

Neonates are a high-risk population for antimicrobial bacteria infections, associated with 200,000 deaths annually, due to their immature microbiome and immune systems. Colonisation of multi-drug resistant (MDR) bacteria may predispose the neonates to resistant infections.

South Asia is a hotspot for multi-drug resistance due to overcrowding and understaffing, limited health-care infrastructures, poor antimicrobial stewardship and lack of diagnostics.

Styczynski et. al. aimed to determine the burden of phenotypic Extended-Spectrum Beta-Lactamase production (p-ESBL) and Carbapenem-Resistant (CRB) Gram-Negative bacteria colonisation among mothers and their neonates within a tertiary care hospital in Faridpur, Bangladesh.

Methods

177 pregnant women were recruited, alongside their neonates, between February-March and August-October 2020; COVID restrictions imposing the four month hiatus. Swabs were taken of the mother and neonates' rectum and the mother's vagina pre and post-delivery. High-contact surfaces were also swabbed. Simultaneously, mothers were surveyed on potential risk factors.

Within 24hrs of collection, swabs were cultured in Chromagar-ESBL and Chromagar mSuperCARBA, with an extra enriching step in tryptic soy broth for environmental swabs prior.

Results

The maternal median age was 25, with one-third having no prior pregnancies and two-thirds of deliveries at or post-term. 79% of deliveries were C-sections resulting in longer hospital stays (4 days vs. 1.6 days).

96% of mothers were prescribed perinatal prophylactic antibiotics for approx. 10 days versus 7 days among vaginal births. The most common prescriptions were metronidazole (89%), flucloxacillin (69%) and 3rd-Generation Cephalosporins (3GCs, 67%). All three were prescribed as a combination prophylactic to 48% of mothers. In contrast, only 5% of neonates were provided antibiotics.

On admission, mothers presented approx. 17% vaginal and 71% rectal colonisation of p-ESBL. At discharge, these increased to 88% and 98% respectively. Vaginal CRB colonisation also rose from 5% to 74% while rectal rose from 15% to 87%. Neonatal rectal swabs were 89% p-ESBL and 72% CRB positive.

High-contact surfaces were repeatedly sampled throughout the study period amounting to 290 swabs, which were 77% p-ESBL and 69% CRB positive. p-ESBL and CRB environmental presence was also higher during COVID.

Statistical analysis revealed significant associations between post-maternal CRB colonisation with C-sections and delivery complications. Neonate p-ESBL and CRB colonisation was also associated with C-sections and post-delivery maternal colonisation. Pre-delivery p-ESBL and CRB colonisation was associated with prior hospitalisation during pregnancy. Other associations were identified, though some were non-significant likely due to low subset sizes.

Discussion

The authors observed high prevalence of perinatal p-ESBL and CRB colonisation for mothers and neonates, especially by discharge. They attribute this mostly to the high rates of C-section births and the prescription of antibiotics as prophylaxis. Additionally, neonatal colonisation most closely matched prevalence in mothers by discharge.

In Bangladesh, another study observed 82% prevalence of ESBL-E. coli among healthy neonates. In contrast, higher income countries like Israel and Sweden presented were 5-14%. ESBL burden is similarly high in other LMICs, though CRB is non-prevalent whereas Styczynski et. al. saw high CRB colonisation. This was despite no carbapenem use reported, suggesting use of other beta-lactams promoted the phenotype.

The study was conducted in a tertiary care hospital, therefore it receives referrals for high risk pregnancies. This may contribute to the 67% C-section rate, beyond the 10-15% recommended by the WHO. The hospital also prescribed prolonged courses of prophylactic antibiotics, against WHO recommendations of a single pre-operation dose. These recommendations come from non-inferiority observations against prolonged and multi-dose regimens in terms of neonatal outcomes. Prior hospitalisation was a risk factor for perinatal maternal colonisation which may be exacerbated if pre-natal practices are similarly non-congruent with antibiotic stewardship.

Other risk factors were identified though were often collinear mode of delivery. Neonates acquire their early microbiomes primarily through the vaginal canal of the mother. C-sections disrupt this thus resulting in less commensal colonisation, presumably providing exploitable niches for opportunistic bacteria, like MDR strains. This is more concerning considering most C-sections in Bangladesh were not out of medical necessity

Neonatal microbiome disruptions can also occur if even only the mother takes antibiotics. The evident overuse of antibiotics in Bangladesh is due to lack of AMR awareness, understanding of antibiotic function and the overemphasis of prophylactic use. Indeed, the study setting lacks a microbiology lab nor an Infectious Diseases consultant, a major gap in expertise for a referral hospital. The authors posit that antibiotics are used as a replacement to developing infrastructure and programs for antimicrobial stewardship. In LMICs, hospital-acquired infections are ~20x more than higher income countries which may make antibiotics more attractive as an immediate fix over longer-term investments. Indeed, the effect of maternal antibiotic usage on neonatal colonisation could not be assessed because almost all mothers were prescribed antibiotics.

Conclusion

Styczynski et. al. conclude that deliveries at the hospital were overmedicalised, with healthcare professionals opting for C-sections far too often and prescribing antibiotics excessively.

Thoughts

This was a very informative exploratory study that highlights the importance of antibiotic stewardship regarding maternal and neonatal health. Most impressive was how aware the authors were of their study's limitations and the lucidity they were expressed. As such, they have already addressed much of my caveats and critique.

Among them were not identifying which bacteria were cultured. Though the setting was clinical, the bulk of the microbial community would be environmental and non-pathogenic human-associated bacteria. Antibiotic resistance is abundant outside of pathogenic bacteria, including p-ESBL, and many of these phenotypes are derived from likely non-mobile genes on non-pathogenic bacteria. Selection for Enterobacterales maximises the likelihood that potentially pathogenic bacteria are captured, while minimising the capture rate of non-pathogenic bacteria. It is a trade-off and though the p-ESBL and CRB prevalence may be accurate, their potential clinical impact may be over-estimated. A follow-up study on neonatal outcomes would be illuminating, if possible.

The study was also in a unique position to investigate the differences between pre and during COVID19. Understandably, this was not within the scope of the study and could not have been planned for. However, its inclusion arguably distracts from the main objectives and not enough evidence was provided for the emergent hypotheses. Throughout the article, the authors used a p-value of <0.05 to mark significance. However, when discussing neonatal colonisation pre and during COVID19, they described the difference as "only marginally significant" with p < 0.10. Yet, in the discussion, they linked the increase in environmental contamination to an increase in neonatal colonisation during COVID19. P values are arbitrary, but I think if a value is chosen, it should be adhered to when constructing conclusions or caveats should be explicitly stated in the immediate text. In reality, regarding colonisation, only pre-delivery p-ESBL and CRB rectal colonisation significantly increased during COVID19.

Intuitively, one could expect an increase in hospital admittance and workload for staff during COVID19. The authors speculate that perhaps decreased routine cleaning by staff concerned about COVID19 contributed to the increase in environmental contamination. This would be an acceptable hypothesis however, without data, they claimed that it "supports the need for enhanced environmental cleaning and prevention efforts". This advice is always prudent, however without data showing that cleaning reduced during COVID19, it cannot be considered supportive. Cleaning could have increased instead or there could be greater risk factors.

Overall, I learned a lot about the interconnectedness of maternal and neonatal health. Furthermore, the sources cited will be formative in my understanding of this field.