To the Editor— Carbapenem-resistant Enterobacteriaceae (CRE) have emerged as a worldwide problem; they have been associated with antibiotic use, long-term and acute-care hospitalization; and they have spread to endemic areas.Reference Friedman, Carmeli, Walton and Schwaber 1 The spread of CRE in communities is a public health threat because CRE infections have limited treatment options and increased mortality. 2
In our hospital in 2016, a premature baby developed sepsis on day of life 29 and died within 24 hours. A blood culture grew CRE Klebsiella pneumoniae. Subsequent investigations into the source of the organism in the neonatal intensive care unit (NICU) did not find any CRE isolates from multiple environmental surface cultures, including isolettes, monitors, weighing scales, sinks, etc. We therefore decided to investigate the prevalence of CRE rectal carriage in our maternity population. Our hypothesis was that the baby was colonized at birth from exposure to maternal colonization with the organism.
We screened 100 consecutive pregnant women delivering babies at The Brooklyn Hospital Center (TBHC) to investigate whether CRE had become a significant clinical issue in this population.
METHODS
Pregnant women admitted in labor were approached and requested to have a perirectal sample taken for surveillance purposes. A convenience sample of 100 sequential perirectal specimens was taken. The study plan was submitted to the TBHC Institutional Review Board as a quality assurance/quality improvement study and was given waived status. The study was anonymous and required only verbal consent. We also administered an epidemiology questionnaire consisting of 15 questions related to travel history, hospitalizations, surgery, and antibiotics during pregnancy. Perirectal swabs were refrigerated and processed within 24 hours of collection. We used the Centers for Disease Control and Prevention (CDC) laboratory protocol for detection of CRE from rectal swabs. 3 Samples that screened positive for CRE were identified using the Vitek system (bioMèrieux, Marcy-l'Étoile, France). Carbapenem minimum inhibitory concentrations (MICs) for these isolates were determined using the Etest method (bioMèrieux). Carbapenem-resistant isolates were screened by polymearase chain reaction (PCR) for bla KPC, bla NDM and bla OXA48 as previously described.Reference Landman, Babu and Shah 4
RESULTS
We identified 2 specimens that grew CRE organisms (both Klebsiella pneumoniae), for a prevalence of CRE colonization in 2% of the population with a confidence interval of 0.2%–7.0% using the Clopper-Pearson method. The MICs of ertapenem and imipenem were >32 μg/mL for both isolates. In addition, PCR testing revealed the presence of bla KPC in both isolates; other carbapenemase genes were not detected. The 2 women colonized with CRE had no history of travel or antibiotics during pregnancy, but 1 of them had been hospitalized in the previous 6 months (Table 1).
TABLE 1 Carbapenem-Resistant Enterobacteriaceae Prevalence Among Pregnant Women in a Brooklyn, New York, Hospital
NOTE. CRE, carbapenem-resistant Enterobacteriaceae.
DISCUSSION
The rate of asymptomatic CRE colonization in this cohort of pregnant women is concerning. New York State has been collecting data on the prevalence of CRE isolates in New York hospitals since 2013. While nearly all hospitals in New York State have had at least 1 CRE isolate, the majority of CRE isolates are found in New York City, particularly the borough of Brooklyn. Publicly reported data from our hospital in 2014 revealed 27 CRE community-onset isolates from all sites in 99,800 patient days and an unadjusted community onset CRE rate of 1.43 per 1,000 admissions. 5 These isolates are from clinical cultures and not the result of routine screening of the population. The rate of asymptomatic colonization could be considerably higher, which could be consistent with the higher rate of carriage in our maternal population.
Around the world, prevalence rates of CRE colonization have a wide range. A recent study in the United States in 7 different communities found a range from 0.82 (Oregon) to 4.80 (Georgia) per 100,000 population in 2013.Reference Guh, Bulens and Mu 6 Comparison data on the prevalence of CRE around the world are not readily available for most countries. However, CRE have been identified in Europe, India and Pakistan, Israel, and the Middle East, South and Central America, China, and Africa.Reference Van Duin and Doi 7
Many studies describe CRE prevalence and CRE outbreaks in hospitals. Risk factors for CRE carriage include healthcare exposures such as short-stay and long-term acute-care hospitalization, long-term care facilities, surgical procedures, indwelling devices, and travel to endemic areas. Recent publications have identified CRE carriage and outbreaks in NICUs, particularly in China.Reference Jin, Shao, Li, Fan, Bai and Wang 8 The prevalence of CRE carriage in a study in Turkey was 2.6% among NICU patients and 3.6% among PICU patients.Reference Akturk, Sutcu and Somer 9 Subsequent infection with CRE organisms occurred in 18% of colonized NICU patients and 39% of pediatric intensive care units (PICU) patients, with a 16.5% mortality rate.
Although much of the literature describes CRE carriage in association with exposure to health care, antibiotics, and comorbid conditions. Once CRE have been introduced into a hospital population, they may subsequently spread to the general population. A laboratory-based surveillance study in Colorado 10 found that 6 of 10 patients identified with CRE between 2014 and 2016 were community associated and not healthcare associated. One of these patients was a pregnant woman.
Although a CRE prevalence rate of 2% in pregnant women may not seem significant, this information has already changed our empiric antibiotic therapy for very sick neonates with suspected gram-negative infections. In addition, the question has arisen of whether we should routinely screen women or a NICU babies for CRE carriage. Identifying CRE carriage would enable us to apply infection prevention isolation techniques to prevent hospital spread. However, CRE screening of all women would be expensive, and contact isolation of exposed babies would be burdensome for the busy staff.
ACKNOWLEDGMENTS
Financial support: No financial support was provided relevant to this article.
Potential conflicts of interest: All authors report no conflicts of interest relevant to this article.
