1. INTRODUCTION

Very low birth weight (VLBW) infants (birth weight < 1500 g) often receive empirical antibiotics in their first few days of life because of the high fatality rates associated with early onset neonatal sepsis1 and the limitations of diagnostic tools for accurate sepsis detection.2,3 However, antibiotic treatment is often prolonged despite of negative culture results,4 as the perceived risk of infection may be obscured by noninfectious conditions resembling sepsis. This practice has raised concerns regarding the adverse effects of prolonged antibiotic use.

Several observational studies have demonstrated that prolonged antibiotic exposure for premature infants early in life is associated with increased rates of necrotizing enterocolitis (NEC).5–9 In addition, although little is known regarding the lung microbiota in preterm infants,10 clinical studies have demonstrated a correlation between the duration of antibiotic exposure and bronchopulmonary dysplasia (BPD),7,11 indicating that commensal bacteria in the lungs may play a role in immune homeostasis and, therefore, the pathogenesis of BPD.12,13 In our study, we implemented care protocols for preterm infants that differed considerably from those employed in other studies. For example, the preferred first-line antibiotics were a combination of ampicillin and cefotaxime, which are more broad-spectrum than those in other studies.8 Moreover, probiotics were routinely administered to clinically stable preterm infants as recommended by the American Gastroenterological Association14; by contrast, probiotics have not been universally adopted in other studies.15 Because numerous systematic reviews have reported that probiotics significantly reduce the risk of NEC,16,17 investigating whether the increased risk of NEC due to antibiotic exposure can be offset by the protective effects of probiotic use is essential.

In this study, we investigated the association between antibiotic use in VLBW infants in the first 14 days of life and prematurity-related adverse outcomes (ie, NEC and BPD) as well as whether the use of probiotics could influence the outcomes. We conducted a retrospective study using multivariable analyses to identify the risk factors for NEC and BPD.

2. METHODS 2.1. Study cohort

We conducted a retrospective study collecting data for VLBW infants who were admitted to Taipei Veterans General Hospital between January 1, 2016, and March 31, 2020. Infants (1) with major congenital anomalies (2) who died or developed NEC (modified Bell stage ≥ IIA) before 14 days of age, and (3) with culture-proven sepsis (defined as bacteremia or bacterial meningitis occurring in the first 14 days of life) were excluded. For BPD analysis, infants who died before 36 weeks postmenstrual age were excluded because they would not meet the criteria for BPD.18 This study was approved by the Institutional Review Board of Taipei Veterans General Hospital (IRB number 2021-01-027CC).

2.2. Protocols for premature infant care

In our study, clinicians followed protocols for premature infant care. The protocols that were relevant to our study were as follows. First, ampicillin and cefotaxime were used empirically for early onset sepsis, and teicoplanin and meropenem were reserved for ill-appearing infants. Second, early introduction of trophic feeding and a preference for human milk over formula were the principles of enteral feeding. Third, probiotics, either 250 mg of Lactobacillus casei daily or a 250-mg combination of Lactobacillus acidophilus and Bifidobaterium bifidum daily, were administered to all infants if they had started enteral feeding. The indication for and timing of surfactant administration were consistent with the 2014 recommendations of the American Academy of Pediatrics.19

2.3. Definitions

Antibiotic treatment was recorded for all infants from birth to 14 days of age. The length of antibiotic treatments was calculated as the number of calendar days that an infant received one or more kinds of antibiotics. Maternal pregnancy and delivery data and infant data from birth until discharge or death were collected. Chorioamnionitis was defined according to the criteria for confirmed intraamniotic infection endorsed by the American College of Obstetricians and Gynecologists.20 Small for gestational age was defined as a birth weight lower than the 10th percentile according to the 2013 Fenton Preterm Growth Chart.21 The Score for Neonatal Acute Physiology with Perinatal Extension-II (SNAPPE-II) was employed to capture illness severity of neonates.22 Probiotic exposure was defined as feeding with probiotics for at least 3 consecutive days before 14 days of age. Human milk exposure was defined as feeding exclusively or partially with human milk for at least 3 consecutive days before 14 days of age. Days of oxygen supply was defined as the number of days that the neonates were supplied with a fraction of inspired oxygen (FiO2) >40%.

The primary outcomes were NEC and BPD. NEC was classified according to the modified Bell criteria, and only NEC stage ≥IIA was included in analyses.23 BPD was categorized according to the revised definition proposed by the National Institute of Child Health and Human Development in 2016,18 and BPD of all severities was included in analyses.

2.4. Statistical analyses

To compare the characteristics of the neonates, the neonates were divided into 2 equal groups according to the median number of days of antibiotic exposure (ie, 0–8 vs 9–14 days). A group comparison of the categorical variables was conducted using the χ2 or Fisher’s exact test, as appropriate, with numbers and percentages provided for each measure. Mann–Whitney U tests were used for the comparisons of continuous variables, with medians and interquartile ranges (IQRs) provided for each measure.

Logistic regression models were used to assess risk factors, including the duration of antibiotic exposure and other variables shown previously to be associated with NEC or BPD.16,17,24–28 These variables included chorioamnionitis, the SNAPPE-II, human milk feeding, probiotic use, days of mechanical ventilation, days of oxygen supplementation, days of postnatal steroid administration, and the presence of hemodynamically significant patent ductus arteriosus. We ran 2 logistic regression models: one treated the length of antibiotic exposure as a continuous variable, and the other treated it as a categorical variable (ie, 0–8 vs 9–14 days). Gestational age, birth weight, and Apgar score at the 5th minute were not included in the models to prevent collinearity due to their linear correlations with the SNAPPE-II. The logistic regression model fit was assessed using the Hosmer–Lemeshow statistic, with P > 0.20 providing adequate fit of the model to the data. Risk was reported as an adjusted odds ratio (aOR) along with a 95% confidence interval (CI).

3. RESULTS

Of the 185 VLBW infants admitted to the medical center during the study period, 132 met the inclusion criteria and were included in the analyses (Fig. 1). The demographic and clinical characteristics of the infants and their mothers are summarized in Table 1. The median number of days of antibiotic exposure during the first 14 days of life was 8. Regarding adverse outcomes, 12 infants (9%) developed NEC, 90 infants (69%) had BPD, and 4 infants (3%) died.

Table 1 - Demographic and clinical characteristics of the enrolled infants and their mothers All Antibiotic exposure p a 0–8 d 9–14 d Number of infants 132 70 62 Antenatal information Maternal age, y, median (IQR) 35 (32–38) 36 (32–38) 35 (31–38) 0.259 Smoke, No. (%) 2 (2) 0 (0) 2 (3.2) 0.219 Chorioamnionitis, No. (%) 21 (19) 9 (13) 12 (19) 0.308 Preeclampsia, No. (%) 34 (26) 20 (29) 14 (23) 0.432 Antenatal steroid, No. (%) 80 (61) 41 (59) 39 (63) 0.611 Delivery information Rupture of membrane > 24 h, No. (%) 27 (20) 14 (20) 13 (21) 0.891 Cesarean section, No. (%) 116 (88) 62 (89) 54 (87) 0.796 Infant characteristic Gestational age, wks, median (IQR) 29 (27–32) 30 (29–33) 28 (26–29) <0.001 Birth weight, g, median (IQR) 1157 (900–1342) 1304 (1186–1422) 988 (809–1126) <0.001 Small for gestational age, No. (%) 30 (23) 22 (31) 8 (13) 0.011 Apgar score at the 5th minute, median (IQR) 8 (7–8) 8 (7–9) 7 (6–8) <0.001 Male, No. (%) 67 (51) 35 (50) 32 (52) 0.853 Twin, No. (%) 31 (23) 16 (23) 15 (24) 0.857 Infant information to day 14 SNAPPE-II, median (IQR) 10 (5–20) 5 (0–12) 15 (5–30) 0.001 Use of surfactant, No. (%) 53 (40) 18 (26) 35 (57) <0.001 Human milk feeding, No. (%) 113 (86) 63 (90) 50 (81) 0.126 Use of probiotics, No. (%) 88 (67) 48 (69) 40 (65) 0.622 Presence ofhemodynamically significantPDA, No. (%) 50 (38) 15 (21) 35 (56) <0.001 Days of mechanical ventilation, median (IQR) 2 (0–17) 0 (0–3) 14 (0–26) <0.001 Days of oxygen supplementation >FiO2 40%, median (IQR) 0 (0–0) 0 (0–0) 0 (0–1) 0.001 Days of steroid administration, median (IQR) 1 (0–2) 0 (0–1) 2 (0–3) <0.001 Days of antibiotic exposure, median (IQR) 8 (7–12) 7 (5–7) 12 (10–14) <0.001 Adverse outcomes NEC, No. (%) 12 (9) 3 (4) 9 (15) 0.041 BPD, No. (%) 90 (69) 33 (47) 57 (95) <0.001 Death, No. (%) 4 (3) 0 (0) 4 (7) 0.046

BPD = bronchopulmonary dysplasia; FiO2 = fraction of inspired oxygen; IQR = interquartile range; NEC = necrotizing enterocolitis; NSAID = nonsteroidal anti-inflammatory drug; PDA = patent ductus arteriosus; SNAPPE-II = Score for Neonatal Acute Physiology with Perinatal Extension-II.

aComparison of characteristics and outcomes by the duration of antibiotic exposure.

Fig. 1:

Study population. BPD = bronchopulmonary dysplasia; NICU = neonatal intensive care unit; NEC = necrotizing enterocolitis; PMA = postmenstrual age; VLBW = very low birth weight.

The enrolled subjects were grouped according to the length of early antibiotic exposure (0–8 vs 9–14 days) for comparative analysis (Table 1). Compared with the group with antibiotic exposure for 0 to 8 days, the group with antibiotic exposure for 9 to 14 days had smaller gestational ages, lower birth weights, lower Apgar scores at the 5th minute, higher SNAPPE-II, higher likelihood of requiring surfactant treatment, longer dependence on mechanical ventilation, and more oxygen and steroid administration. Infants with antibiotic exposure for 9 to 14 days also had significantly higher rates of adverse outcomes, including NEC, BPD, and death.

After adjustment for confounding variables, infants with longer antibiotic exposure (9–14 days) had higher odds of BPD than those with shorter antibiotic exposure (aOR, 10.104; 95% CI, 2.281–44.758; Table 2). When tested as a continuous variable, each additional day of antibiotic use was associated with higher odds of NEC (aOR, 1.278; 95% CI, 1.025–1.593) and BPD (aOR, 1.630; 95% CI, 1.233–2.156).

Table 2 - Multivariable logistic regression examining neonates’ outcomes aOR 95% CI p NECa Antibiotic exposure per day 1.278 1.025–1.593 0.03 Antibiotic exposure 9–14 d 3.436 0.818–14.431 0.092 BPDb,c Antibiotic exposure per day 1.630 1.233–2.156 0.001 Antibiotic exposure 9–14 d 10.104 2.281–44.758 0.002

Two different logistic regression models were calculated: one treated the length of antibiotic exposure as a continuous variable, while the other as a categorical variable (ie, 0–8 vs 9–14 days).

aOR = adjusted odds ratio; BPD = bronchopulmonary dysplasia; CI = confidence interval; NEC = necrotizing enterocolitis.

aAdjusted for SNAPPE-II, human milk feeding, and use of probiotics.

bTwo neonates died before 36 weeks postmenstrual age, so they were excluded from the analysis as they could not meet the criteria for BPD.

cAdjusted for chorioamnionitis, SNAPPE-II, days of mechanical ventilation, days of oxygen supplementation, days of steroid administration, and presence of hemodynamically significant PDA.

4. DISCUSSION

In this study, we demonstrated that longer antibiotic exposure was associated with increased risks of developing NEC and BPD in VLBW infants. Notably, our study revealed that the association remained after adjusting for probiotic use in the NEC analysis, indicating that the probiotics did not offset the adverse effects of prolonged antibiotic exposure.

The increased risk of NEC may be explained by the reduced diversity of microbiota caused by perinatal antibiotics, which may lead to overgrowth of pathogenic microbes over commensal species.29,30 This change became significant at 10 days of age and most profound at 30 days of age,31 coinciding with the general timing of the presentation of NEC.32 In addition, antibiotic therapy can alter and weaken the innate and adaptive immune system, resulting in increased cytokine-mediated inflammatory responses that can lead to gut injury.33 However, the pathogenesis for an increased risk of BPD has not been directly verified. Only weak evidence suggests that airway microbial dysbiosis is associated with BPD progression and severity.10

Our findings are consistent with those of previous studies. Cotton et al used a multicenter database to demonstrate that prolonged initial antibiotic treatment for extremely low birth weight infants (birth weight < 1000 g) was associated with increased incidence of NEC or death.6 Ting et al used antibiotic use rate, defined as the number of antibiotic exposure days divided by the total length of hospital stay, to demonstrate the associations between antibiotic exposure and adverse outcomes (i.e., periventricular leukomalacia, BPD, retinopathy and mortality).34 Cantey et al used a single-center database to investigate the correlations between early antibiotic exposure and NEC, BPD and death.8,11

A strength of our study is the inclusion of probiotic use as a variable for the NEC analysis. Systematic reviews have reported that probiotics significantly reduced the risk of NEC.16,17 However, whether probiotics have the benefit of preventing the development of NEC in preterm infants receiving antibiotics has not been widely studied. Our study revealed that probiotics did not reduce the adverse outcomes caused by increased antibiotic exposure. A possible explanation is that broad-spectrum antibiotics, which probiotics might be susceptible to, were used in our study. In addition, the length of the initial antibiotic treatment was longer in our study than in other studies, which might have undermined the effects of probiotics. Ting et al7 reported that 42% of VLBW infants are administered antibiotics for >3 days, while all our enrolled infants received antibiotics for >3 days. Cantey et al reported a median number of days of antibiotic therapy was 5.5,8 whereas the median was 8 in our study.

A second strength is that our study employed data spanning a 5-year period from a medical center, ensuring consistency in the principles of medical practice. Third, treating the length of antibiotic exposure as a continuous variable may provide valuable information for refining antibiotic therapy.

This study has several limitations. First, the infants at higher risk of adverse outcomes are likely the same infants who receive longer antibiotic treatment. However, we attempted to minimize this limitation by adjusting for SNAPPE-II. Second, the cohort size was relatively small. Future studies should involve a larger cohort and a prospective design.

In conclusion, our findings indicated that a longer duration of early antibiotic exposure contributed to increased risks of developing NEC and BPD in VLBW infants. The associations remained after adjustment for variables known to affect the incidence of outcomes. Notably, the potentially protective effect of probiotics was limited. Our findings suggest that rigorous antibiotic stewardship in VLBW infants to avoid adverse consequences is imperative.

ACKNOWLEDGMENTS

This manuscript was edited by Wallace Academic Editing.

REFERENCES 1. Stoll BJ, Hansen N. Infections in VLBW infants: studies from the NICHD Neonatal Research Network. Semin Perinatol. 2003;27:293–301. 2. Zea-Vera A, Ochoa TJ. Challenges in the diagnosis and management of neonatal sepsis. J Trop Pediatr. 2015;61:1–13. 3. Chen IL, Huang HC, Ou-Yang MC, Chen FS, Chung MY, Chen CC. A novel method to detect bacterial infection in premature infants: using a combination of inflammatory markers in blood and saliva. J Microbiol Immunol Infect. 2020;53:892–9. 4. Cordero L, Ayers LW. Duration of empiric antibiotics for suspected early-onset sepsis in extremely low birth weight infants. Infect Control Hosp Epidemiol. 2003;24:662–6. 5. Alexander VN, Northrup V, Bizzarro MJ. Antibiotic exposure in the newborn intensive care unit and the risk of necrotizing enterocolitis. J Pediatr. 2011;159:392–7. 6. Cotten CM, Taylor S, Stoll B, Goldberg RN, Hansen NI, Sánchez PJ, et al.; NICHD Neonatal Research Network. Prolonged duration of initial empirical antibiotic treatment is associated with increased rates of necrotizing enterocolitis and death for extremely low birth weight infants. Pediatrics. 2009;123:58–66. 7. Ting JY, Roberts A, Sherlock R, Ojah C, Cieslak Z, Dunn M, et al.; Canadian Neonatal Network Investigators. Duration of initial empirical antibiotic therapy and outcomes in very low birth weight infants. Pediatrics. 2019;143:e20182286. 8. Cantey JB, Pyle AK, Wozniak PS, Hynan LS, Sánchez PJ. Early antibiotic exposure and adverse outcomes in preterm, very low birth weight infants. J Pediatr. 2018;203:62–7. 9. Rina P, Zeng Y, Ying J, Qu Y, Mu D. Association of initial empirical antibiotic therapy with increased risk of necrotizing enterocolitis. Eur J Pediatr. 2020;179:1047–56. 10. Pammi M, Lal CV, Wagner BD, Mourani PM, Lohmann P, Luna RA, et al. Airway microbiome and development of bronchopulmonary dysplasia in preterm infants: a systematic review. J Pediatr. 2019;204:126–133.e2. 11. Cantey JB, Huffman LW, Subramanian A, Marshall AS, Ballard AR, Lefevre C, et al. Antibiotic exposure and risk for death or bronchopulmonary dysplasia in very low birth weight infants. J Pediatr. 2017;181:289–293.e1. 12. Tirone C, Pezza L, Paladini A, Tana M, Aurilia C, Lio A, et al. Gut and lung microbiota in preterm infants: immunological modulation and implication in neonatal outcomes. Front Immunol. 2019;10:2910. 13. Kuo CH, Kuo HF, Huang CH, Yang SN, Lee MS, Hung CH. Early life exposure to antibiotics and the risk of childhood allergic diseases: an update from the perspective of the hygiene hypothesis. J Microbiol Immunol Infect. 2013;46:320–9. 14. Su GL, Ko CW, Bercik P, Falck-Ytter Y, Sultan S, Weizman AV, et al. AGA clinical practice guidelines on the role of probiotics in the management of gastrointestinal disorders. Gastroenterology. 2020;159:697–705. 15. Athalye-Jape G, Patole S. Probiotics for preterm infants–time to end all controversies. Microb Biotechnol. 2019;12:249–53. 16. Dermyshi E, Wang Y, Yan C, Hong W, Qiu G, Gong X, et al. The “Golden Age” of probiotics: a systematic review and meta-analysis of randomized and observational studies in preterm infants. Neonatology. 2017;112:9–23. 17. Sawh SC, Deshpande S, Jansen S, Reynaert CJ, Jones PM. Prevention of necrotizing enterocolitis with probiotics: a systematic review and meta-analysis. PeerJ. 2016;4:e2429. 18. Higgins RD, Jobe AH, Koso-Thomas M, Bancalari E, Viscardi RM, Hartert TV, et al. Bronchopulmonary dysplasia: executive summary of a workshop. J Pediatr. 2018;197:300–8. 19. Polin RA, Carlo WA; Committee on Fetus and Newborn; American Academy of Pediatrics. Surfactant replacement therapy for preterm and term neonates with respiratory distress. Pediatrics. 2014;133:156–63. 20. Committee opinion No. 712: intrapartum management of intraamniotic infection. Obstet Gynecol. 2017;130:e95–e101. 21. Fenton TR, Kim JH. A systematic review and meta-analysis to revise the Fenton growth chart for preterm infants. BMC Pediatr. 2013;13:1–13. 22. Richardson DK, Corcoran JD, Escobar GJ, Lee SK. SNAP-II and SNAPPE-II: Simplified newborn illness severity and mortality risk scores. J Pediatr. 2001;138:92–100. 23. Walsh MC, Kliegman RM. Necrotizing enterocolitis: treatment based on staging criteria. Pediatr Clin North Am. 1986;33:179–201. 24. Watterberg KL, Demers LM, Scott SM, Murphy S. Chorioamnionitis and early lung inflammation in infants in whom bronchopulmonary dysplasia develops. Pediatrics. 1996;97:210–5. 25. Wai KC, Kohn MA, Ballard RA, Truog WE, Black DM, Asselin JM, et al.; Trial of Late Surfactant (TOLSURF) Study Group. Early cumulative supplemental oxygen predicts bronchopulmonary dysplasia in high risk extremely low gestational age newborns. J Pediatr. 2016;177:97–102.e2. 26. Le Doare K, Holder B, Bassett A, Pannaraj PS. Mother’s milk: a purposeful contribution to the development of the infant microbiota and immunity. Front Immunol. 2018;9:361. 27. Laughon MM, Langer JC, Bose CL, Smith PB, Ambalavanan N, Kennedy KA, et al.; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. Prediction of bronchopulmonary dysplasia by postnatal age in extremely premature infants. Am J Respir Crit Care Med. 2011;183:1715–22. 28. Ucar S, Varma M, Ethemoglu M, Acar N. The efficacy of Snappe-II in predicting morbidity and mortality in extremely low birth weight infants. Arch Dis Child. 2014;99:A468. 29. Arboleya S, Sánchez B, Solís G, Fernández N, Suárez M, Hernández-Barranco AM, et al. Impact of prematurity and perinatal antibiotics on the developing intestinal microbiota: a functional inference study. Int J Mol Sci. 2016;17:E649. 30. Chernikova DA, Koestler DC, Hoen AG, Housman ML, Hibberd PL, Moore JH, et al. Fetal exposures and perinatal influences on the stool microbiota of premature infants. J Matern Fetal Neonatal Med. 2016;29:99–105. 31. Arboleya S, Sánchez B, Milani C, Duranti S, Solís G, Fernández N, et al. Intestinal microbiota development in preterm neonates and effect of perinatal antibiotics. J Pediatr. 2015;166:538–44. 32. Yee WH, Soraisham AS, Shah VS, Aziz K, Yoon W, Lee SK; Canadian Neonatal Network. Incidence and timing of presentation of necrotizing enterocolitis in preterm infants. Pediatrics. 2012;129:e298–304. 33. Sharma R, Tepas JJ 3rd, Hudak ML, Mollitt DL, Wludyka PS, Teng RJ, et al. Neonatal gut barrier and multiple organ failure: role of endotoxin and proinflammatory cytokines in sepsis and necrotizing enterocolitis. J Pediatr Surg. 2007;42:454–61. 34. Ting JY, Synnes A, Roberts A, Deshpandey A, Dow K, Yoon EW, et al.; Canadian Neonatal Network Investigators. Association between antibiotic use and neonatal mortality and morbidities in very low-birth-weight infants without culture-proven sepsis or necrotizing enterocolitis. JAMA Pediatr. 2016;170:1181–7.