Secondary Logo

Journal Logo

Maternal-Neonatal Reports

Long-term Follow-up of Preterm Infants Having Been Colonized With Extended Spectrum Beta-lactamase–producing Enterobacterales Over the First 6 Years of Life

Eberhart, Martin MD*; Grisold, Andrea MD, MBA; Lavorato, Michela MD*,‡; Resch, Elisabeth MD*,§; Trobisch, Andreas MD*,§; Resch, Bernhard MD*,§

Author Information
The Pediatric Infectious Disease Journal: September 2021 - Volume 40 - Issue 9 - p 835-837
doi: 10.1097/INF.0000000000003212
  • Free


Intestinal colonization with extended spectrum beta-lactamase (ESBL) producing Enterobacterales represents a kind of dysbiosis of the neonatal microbiome persisting after colonization in the neonatal period.1,2 ESBL-producing Enterobacterales (ESBL-E) arise by mutation from classical beta-lactamase enzymes and are characterized by resistances mainly against broad-spectrum cephalosporins and monobactams.3 ESBL-E became a growing issue in health care systems worldwide and especially in intensive care units, and data from the United States has demonstrated that clinical isolates from children tripled between 1999 and 2011.4 Preterm infants ≤32 weeks of gestation are at high-risk of nosocomial infections because of immaturity of the immune defense mechanisms, and preventive measures have been established to avoid invasive infections by ESBL-E.5,6

In a previously published retrospective case-control study, we were able to demonstrate that ESBL-E colonization of preterm infants ≤32 weeks of gestation (from stool surveillance cultures) at the neonatal intensive care unit (NICU) was not associated with an increased risk of necrotizing enterocolitis.7 From this cohort, we extracted a matched case-control group and retrospectively followed them over the first 6 years of life with focus on gastrointestinal diseases, infections and neurodevelopmental impairment. We hypothesized that the case group might have more gastrointestinal diseases or infections compared with controls without neonatal ESBL-E colonization of the gut. ESBL-E are associated with dysbiosis that might lead to more frequent occurrence of gastrointestinal tract diseases and infections. The developing brain of the preterm infant is highly vulnerable and might be harmed by lymphocyte- or cytokine-mediated injury of the brain via the gut-brain axis.8


Between 2005 and 2013, 146 preterm infants with a gestational age ≤32 weeks of gestation who had been colonized with ESBL-E during neonatal hospitalization were retrospectively followed up to the age of 6 years until 2019 (cases with ESBL colonization). The control group was matched by 1:1 according to year of birth, gestational age, birth weight and gender (control infants ≤32 weeks without ESBL colonization) as published recently.7 The local ethics committee of the Medical University of Graz, Austria (number 32-052 ex 19/20), provided ethical approval of this retrospective case-control cohort study. We collected the total number of outpatient visits and inpatient hospital stays regarding bacterial and viral infections including ESBL-E bacteremias of both groups, number of children with febrile seizures, appendicitis, ileus, volvulus, gastritis, enteritis/diarrhea and gastroesophageal reflux disease and number of children with neurodevelopmental impairment. Routine outpatient visits for the well-being of the children were excluded. Vaccination rates were documented at the neurodevelopmental follow-up visits.

For neurodevelopmental outcome, infants were routinely examined at the Outpatient Clinic of Neurodevelopmental Follow-up of our department. Two experienced and certified evaluators specialized in neuropediatrics and developmental pediatrics examined the children, who were followed up in our neurodevelopmental follow-up program until school age. Assessment of outcome at the age of 5–6 years (not corrected for prematurity) was made using the developmental tests as described by Kaufman and Kaufman,9 neurologic examinations as described by Amiel-Tison and Stewart10 and Touwen11 and, if possible, the Wechsler Preschool and Primary Scales of Intelligence (WPPSI-4). Children were diagnosed as having developed normally, having mild developmental delay (WPPSI score 80-89) and having moderate or severe intellectual disability (WPPSI scores 70–79 or < 70, respectively). Isolated or mildly abnormal neurologic examination findings without associated functional impairment when suspicious but not abnormal at the time of examination were classified as mild developmental delay. Children classified as abnormal had definite abnormalities in tone, impairment of gross motor function and delayed motor milestones.12 We classified infants as having motor developmental delay or definite diagnosis of cerebral palsy. Additional parameters included disorder of speech development, visual impairment, hearing impairment, behavioral problems and diagnosis of attention-deficit hyperactivity disorder.

Our pediatric department covers a region with 8–10.000 births per year, and there is no other hospital which offers treatment or follow-up of pediatric patients in this area of Southern Styria in the Southeast of Austria with, whose regional capital Graz is the second largest city of Austria.

Weight gain was assessed using the last visit between the age of 5 and 6 years. Z scores were calculated by use of a percentiles calculator provided at

Antibiotic regimen at the NICU did not change over study time, including cefuroxime and ampicillin as standard antibiotics for suspected or proven early-onset sepsis. Mean length of standard antibiotic therapy was counted in days. Other antibiotic therapies were recorded individually. Furthermore, all preterm infants received enteral probiotics (Lactobacillus casei rhamnosus), enteral gentamicin and enteral antimycotics (nystatin) as a multimodal prophylaxis against necrotizing enterocolitis.14 ESBL-E–positive infants were isolated or cohorted and cared for using gloves, coats and facemasks.

Statistical analyses were done using Excel© (Microsoft Office, Excel 2013, Richmond, WA) and SPSS© (IBM SPSS Statistics 22, Somers, NY). For categorical data χ2 or Fisher’s exact tests and for numerical data t test or Mann-Whitney U test were used as appropriate. Normality assumption was checked using the Shapiro-Wilk test. Statistical significance was set at P < 0.05. To evaluate the study hypothesis, we compared all gastrointestinal diagnoses as listed above, rates of viral and bacterial infections and rates of neurodevelopmental impairment between both cohorts.


During the study period, we included 149 very preterm infants (cases) and 149 matched controls from the original cohort of 217 infants in each group. In the neonatal period, 3 (2%) children in the case and 3 (2%) in the control group died; thus, we finally compared 146 cases with 146 controls, respectively, for further follow-up. None of the children died during the follow-up period. One-hundred thirty-five of 146 cases (92.5%) had been discharged while still colonized with ESBL-E. None of the infants had invasive infection, and more than a third of the cases were twins or triplets. Infants colonized with ESBL-E stayed longer at the NICU (68 vs. 63 days; P = 0.016). Other perinatal data did not differ between cases and controls (Table 1). Rates of intraventricular hemorrhage grade 3 and 4 and cystic periventricular leukomalacia did not differ between groups nor did the rate of bronchopulmonary dysplasia, retinopathia of prematurity and rates of early and late-onset sepsis or necrotizing enterocolitis. Antibiotics used during the neonatal stay are shown in Table 2. Vaccination rates did not differ between groups. Six-year follow-up revealed neither differences regarding gastrointestinal tract diseases nor infections (Table 3) or neurodevelopmental impairment (Table 4). There was a trend to more gastroesophageal reflux disease (18.5% vs. 11.6%; P = 0.06) and a better neurodevelopmental outcome (76% vs. 67%; P = 0.058) both in the case group.

TABLE 1. - Perinatal Data of the Cases (Preterm Infants ≤32 Weeks of Gestation Colonized With ESBL-E) and Matched Controls (Without ESBL-E Colonization) Born 2005–2013
Parameter Cases Controls P
Number 146 146
Gestational age (weeks) 28 ± 2.5 28 ± 2.6 0.41
 <28 weeks of gestational age 52 (36) 57 (39) 0.27
Birth weight (grams) 1180 ± 433 1196 ± 406 0.38
 <1500 grams 113 (77) 111 (76) 0.34
Small for gestational age 22 (15.1) 18 (14.1) 0.37
Multiple birth 61 (42) 31 (23) < 0.001
Maternal age (yr) 30.4 ± 6.4 29.8 ± 5.9 0.16
Cesarean section 129 (88) 121 (82) 0.32
APGAR at 1 min 6.4 ± 2.0 6.7 ± 2.1 0.15
APGAR at 5 min 8.3 ± 1.3 8.2 ± 1.4 0.41
APGAR at 10 min 8.8 ± 1.0 8.8 ± 1.0 0.34
Umbilical artery pH 7.29 ± 0.1 7.24 ± 0.6 0.18
Length of NICU stay 68.1 ± 40.7 62.6 ± 54.4 0.016
Data are given as n (%) or mean ± standard deviation.

TABLE 2. - Antibiotics During Neonatal Stay of 146 Cases (Preterm Infants ≤32 Weeks of Gestation Colonized With ESBL-E) Born 2005–2013 and Matched Controls (Without ESBL-E Colonization) Followed Up Until the Age of 6 Years
Antibiotics Cases Controls P
Standard antibiotic regimen 100 (69) 94 (64) 0.229
 Duration (d) 5.0 ± 3.5 5.8 ± 3.8 0.493
Imipenem 64 (44) 51 (36) 0.273
Teicoplanin 27 (18.5) 23 (16) 0.066
Linezolid 9 (6.7) 2 (1.4) 0.054
Erythromcin 5 (3.4) 6 (4.1) 0.472
Meropenem 6 (4.1) 8 (5.5) 0.333
Clarithromycin 4 (2.7) 5 (3.4) 0.179
Other 8 (5.5) 6 (4.1) 0.293
Standard antibiotics are cefuroxime and ampicillin. Data are given as n (%) or mean ± standard deviation.

TABLE 3. - Follow-up of 146 Cases (Preterm Infants ≤32 Weeks of Gestation Colonized With ESBL-E) Born 2005–2013 and Matched Controls (Without ESBL-E Colonization) Followed Until the Age of 6 Years
Parameter Cases Controls P
N = 146 N = 146
Outpatient visits (n per cohort) 252 209 0.15
Hospitalizations 155 168 0.33
Rehospitalization (total days per cohort) 772 822 0.38
Diarrhea, acute (n) 36 (25) 29 (20) 0.29
Gastritis (n) 19 (13) 11 (7.5) 0.11
Diarrhea and vomiting, acute (n) 53 (36) 40 (27) 0.11
Appendicitis (n) 0 1 0.16
Ileus (n) 0 0
Volvulus (n) 1 0 0.16
GERD (n) 27 (19) 18 (12) 0.06
Viral respiratory infections (n per cohort) 199 187 0.39
Bacterial infections (having received antibiotics; n per cohort) 67 55 0.25
Febrile convulsions (n) 5 (3.4) 10 (6.8) 0.17
Data are given as n (%), n per cohort or mean ± standard deviation.
GERD indicates gastroesophageal reflux disease.

TABLE 4. - Neurodevelopmental Follow-up and Weight Gain of 146 Cases (Preterm Infants ≤32 Weeks of Gestation Colonized With ESBL-E) Born 2005–2013 and Matched Controls (Without ESBL-E Colonization) Followed Until the Age of 6 Years
Cases Controls P
Parameter n = 146 n = 146
Lost to follow-up 6 (4.1) 9 (6.2) 0.202
Infants with follow-up 140 (96) 137 (94)
Normal developmental outcome* 106 (76) 92 (67) 0.058
Mild developmental delay* 16 (11) 19 (14) 0.335
Moderate intellectual disability* 7 (5) 13 (9.5) 0.075
Severe intellectual disability* 3 (2.1) 3 (2.2) 0.489
Motor developmental delay 4 (2.9) 7 (5.1) 0.169
Cerebral palsy 10 (7.1) 7 (5.1) 0.241
Disorder of speech development 7 (5) 12 (8.8) 0.108
Visual impairment 2 (1.4) 0 0.081
Hearing impairment 0 1 (0.7) 0.156
Behavioral problems§ 4 (2.9) 3 (2.2) 0.362
Attention-deficit hyperactivity disorder§ 3 (2.1) 4 (2.9) 0.341
Weight (kg) 17.5 ± 3.4 17.7 ± 3.1 0.36
Z score −1.5 ± 1.5 −1.4 ± 1.4 0.21
Data are given as n (%).
*Kaufman et al9 and WPPSI-4 (see METHODS section).
†Amiel-Tison and Stewart10 and Touwen11.
‡According to ophthalmologist and/or ENT specialist.
§According to pediatric psychiatrist.
ADHD indicates attention-deficit hyperactivity disorder; ENT, ear-nose-throat.


We could not confirm our hypothesis that former ESBL-E colonized preterm infants ≤32 weeks of gestation had more gastrointestinal diseases, infections or a worse neurodevelopmental outcome during a 6-year follow-up until school age.

It is well known that significant proportions of ESBL-E carriers remain colonized up to 1 year in the healthcare setting, and long-term effects of decolonization therapies remain unclear.2 Nordberg et al15 found that after 2 years, 3 of 26 children were still carriers of ESBL-E and at 5 years of age, only 1 child was colonized with ESBL-E (median length of 12.5 months and range 5–68 months).

The duration of colonization is of importance to assess the risk of later invasive infection by ESBL-E in the carrier and the risk of dissemination in the community.16 Furthermore, follow-up studies play a role in understanding the dynamics of ESBL-E outbreaks. A study from 2 Swiss intensive care units reported that ESBL-E colonization on admission was associated with subsequent ESBL-E infection with a hazard ratio of 26 (95% CI, 2.4–271).17 Our follow-up study revealed no invasive ESBL-E infection in former preterm infants. Overall health status compared with matched controls did not differ suggesting no further influence of a possible dysbiosis of the gut microbiome. Rates of antibiotics and duration did not differ nor did baseline characteristics between groups. Weight was comparable between groups, as were rates of neurodevelopmental impairment. Recent research using a germ-free mouse model found that intestinal microbiota might influence brain development by an association of poor growth phenotype-associated microbiota with increased neuroinflammation—marked by increased nitric oxide synthase 1, as well as alteration in insulin-like growth factor 1 pathway including decreased circulating and brain insulin-like growth factor 1, decreased circulating insulin-like growth factor binding protein 3 and increased insulin-like growth factor binding protein 3 brain messenger RNA expression.18 Nevertheless, neurodevelopmental follow-up did not reveal higher rates of disabilities in our ESBL-E colonized children.

Limitations of the study include the retrospective design. Even careful data collection does not reach the level of a prospective study. Regarding neurodevelopmental outcome, we did not have data on the socioeconomic status of the parents that strongly influences development of the children. One strength of the study includes the unique catchment area of the pediatric department resulting in homogeneous data acquisition.

In conclusion, we did not find differences regarding gastrointestinal diseases or infections or long-term outcome of preterm infants ≤32 weeks of gestation with or without ESBL-E colonization.


1. Pammi M, Cope J, Tarr PI, et al. Intestinal dysbiosis in preterm infants preceding necrotizing enterocolitis: a systematic review and meta-analysis. Microbiome. 2017;5:31.
2. Bar-Yoseph H, Hussein K, Braun E, et al. Natural history and decolonization strategies for ESBL/carbapenem-resistant Enterobacteriaceae carriage: systematic review and meta-analysis. J Antimicrob Chemother. 2016;71:2729–2739.
3. Coque TM, Baquero F, Canton R. Increasing prevalence of ESBL-producing Enterobacteriaceae in Europe. Euro Surveill. 2008;13:19044.
4. Logan LK, Braykov NP, Weinstein RA, et al.; Centers for Disease Control and Prevention Epicenters Prevention Program. Extended-spectrum β-lactamase-producing and third-generation cephalosporin-resistant enterobacteriaceae in children: trends in the United States, 1999-2011. J Pediatric Infect Dis Soc. 2014;3:320–328.
5. Eichenwald EC, Stark AR. Management and outcomes of very low birth weight. N Engl J Med. 2008;358:1700–1711.
6. Christoph J, Dame C, Eckmanns T, et al. Prevention of nosocomial infections of preterm infants below 1500 grams at the NICU – KRINKO recommendations from 2007 and 2012 [German]. Epid Bull. 2013;42 (suppl):1–56.
7. Eberhart M, Grisold A, Lavorato M, et al. Extended-spectrum beta-lactamase (ESBL) producing enterobacterales in stool surveillance cultures of preterm infants are no risk factor for necrotizing enterocolitis: a retrospective case-control study over 12 years. Infection. 2020;48:853–860.
8. Zhou Q, Niño DF, Yamaguchi Y, et al. Necrotizing enterocolitis induces T lymphocyte-mediated injury in the developing mammalian brain. Sci Transl Med. 2021;13:eaay6621.
9. Kaufman AS, Kaufman NL. Kaufmann – Assessment Battery for Children. Deutschsprachige Fassung von Peter Melchers und Ulrich Preuß. Swets & Zeitlinger B.V.; 1991.
10. Amiel-Tison C, Stewart A. Follow up studies during the first five years of life: a pervasive assessment of neurological function. Arch Dis Child. 1989;64(4 Spec No):496–502.
11. Touwen BCL. Neurological Development in Infancy. William Heineman Medical Books Ltd; 1976.
12. Adams-Chapman I, Heyne RJ, DeMauro SB, et al.; Follow-Up Study of the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. Neurodevelopmental impairment among extremely preterm infants in the neonatal research network. Pediatrics. 2018;141:e20173091.
13. Hesse V, Schnabel O, Judis E, et al. Longitudinal study on growth charts of German children at the age of 0 – 6 years. Part 1 [German]. Monatsschr Kinderheilkd. 2016:164:478–496.
    14. Schmolzer G, Urlesberger B, Haim M, et al. Multi-modal approach to prophylaxis of necrotizing enterocolitis: clinical report and review of literature. Pediatr Surg Int. 2006;22:573–580.
    15. Nordberg V, Jonsson K, Giske CG, et al. Neonatal intestinal colonization with extended-spectrum β-lactamase-producing enterobacteriaceae-a 5-year follow-up study. Clin Microbiol Infect. 2018;24:1004–1009.
    16. Martin RM, Cao J, Brisse S, et al. Molecular epidemiology of colonizing and infecting isolates of Klebsiella pneumoniae. mSphere. 2016;1:e00261–e00216.
    17. Emmanuel Martinez A, Widmer A, Frei R, et al. ESBL-colonization at ICU admission: impact on subsequent infection, carbapenem-consumption, and outcome. Infect Control Hosp Epidemiol. 2019;40:408–413.
    18. Lu J, Lu L, Yu Y, et al. Effects of intestinal microbiota on brain development in humanized gnotobiotic mice. Sci Rep. 2018;8:5443.

    enterobacterales; extended spectrum beta-lactamase; follow-up; preterm infants

    Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.