Journal of Tropical Pediatrics, 2018, 64, 24–30

doi: 10.1093/tropej/fmx023
Advance Access Publication Date: 18 April 2017
Original paper

Early Total Enteral Feeding in Stable Very

Low Birth Weight Infants: A Before and
After Study

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019

by Sushma Nangia,1 Amit Bishnoi,2 Ankita Goel,2 Piali Mandal,2
Soumya Tiwari,2 and Arvind Saili1
1
Department of Neonatology, Lady Hardinge Medical College and Kalawati Saran Children Hospital, New Delhi 110001, India
2
Department of Pediatrics, Lady Hardinge Medical College and Kalawati Saran Children Hospital, New Delhi 110001, India
*Correspondence: Sushma Nangia, Department of Neonatology, Lady Hardinge Medical College and Kalawati Saran Children Hospital, New
Delhi 110001, India. Tel: 011-23344161-70, Ext 331, Mobile: 91-9810838181, E-mail <[email protected]>.

ABSTRACT
Background: Fear of necrotizing enterocolitis (NEC) has perpetuated delayed initiation and slow
advancement of enteral feeding in very low birth weight (VLBW) infants with inherent risks of par-
enteral alimentation. The objective of this study was to assess effect of early total enteral feeding
(ETEF) on day of achievement of full enteral feeds, feed intolerance, NEC and sepsis.
Methods: In total, 208 stable VLBW neonates (28–34 weeks) admitted during 6 month periods of
three consecutive years were enrolled. First phase (n ¼ 73) constituted the ‘before’ phase with
standard practice of initial intravenous fluid therapy and slow enteral feeding. The second prospect-
ive phase (n ¼ 51) consisted of implementation of ETEF with infants receiving full enteral feeds as
per day’s fluid requirement since Day 1 of life. The third phase (n ¼ 84) was chosen to assess the
sustainability of change in practice.
Results: Day of achievement of full feeds was significantly earlier in Phases 2 and 3 compared with
Phase 1 (8.97 and 5.47 vs. 14.44 days, respectively, p ¼ 0.0001). Incidence of feed intolerance was
comparable between Phases 1 and 2 (22 vs. 14%, p ¼ 0.28), with marked reduction in incidence of
NEC (14 vs. 4%, p ¼ 0.028). There was a significant decrease in sepsis, duration of parenteral fluid
and antibiotic therapy as well as hospital stay with comparable mortality.
Conclusion: In stable preterm VLBW infants, ETEF is safe and has the benefit of optimizing nutri-
tion with decrease in sepsis, NEC and hospital stay.

K E Y W O R D S : early total enteral feeding, necrotizing enterocolitis, sepsis, very low birth weight

INTRODUCTION parallel the corresponding intrauterine period [2].

Optimal nutrition has been identified as a fundamen- This goal remains elusive to best neonatal centres
tal factor in reducing mortality and long-term mor- around the world with almost 90% having growth
bidities like extrauterine growth restriction and poor delay at 36 weeks’ corrected age and 40% at 18–
neurodevelopmental outcome in preterm very 22 months of age [3]. In most centres, aggressive
low birth weight (VLBW) infants (birth weight early nutritional rehabilitation of preterms is
<1500 g) [1, 4]. According to American Academy of achieved by total parenteral nutrition (TPN) with
Pediatrics, postnatal growth of preterms should delayed initiation of enteral feeding. However,

C The Author [2017]. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
V  24
 Early Total Enteral Feeding in VLBW Neonates  25

complications like sepsis related to invasive catheters support (CPAP or ventilation)] admitted to NICU
along with gut translocation of bacteria, gut atrophy, were included in the study [19, 20]. Exclusion crite-
thrombosis or bleeding and cholestasis offset the ria were documented as absence or reversal of end
benefits of TPN [5]. diastolic flow in umbilical arteries and presence of
Improved understanding of preterm gastrointes- gross congenital malformations.
tinal functions along with increased recognition of Study design: The study was divided into three
beneficial effects of human milk has resulted in tro- phases. In Phase 1, retrospective data were collected
phic feeding along with TPN being put forth as a from case files of eligible infants, admitted between

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019

feasible solution [6–8]. More recently, studies with January and June 2010, who were treated with stand-
variable time of initiation of feeds and different feed- ard practice of initial intravenous (IV) therapy (10%
ing regimens have been conducted in preterm infants dextrose for first 2 days and N/5 saline in 10% dex-
[9–14]. Ostertag et al. [15] found no difference in trose from Day 3 onwards). These neonates also
necrotizing enterocolitis (NEC) between enteral received slow incremental enteral feeding consisting
feeding started on Day 1 vs. Day 7 in sick VLBW in- of men on Day 1 (20 ml/kg/day) with successive
fants. Recent Cochrane review compared slow vs. feed increments of 20 ml/kg/day along with propor-
fast daily increments of feed volume with no increase tionate decrease in the IV fluid volume till the baby
in NEC, mortality or feed interruption [16]. Survey reached full feeds (150 ml/kg/day). Between July
of enteral feeding practices among 127 tertiary and December of 2010, 20 stable VLBW neonates
Neonatal intensive care units (NICU) revealed initi- were provided ETEF (total fluid requirement for the
ation within 24 h in 35% of preterm neonates when day provided as enteral feed with no IV fluid) on a
gestational age (GA) was <25 weeks, 43% for GA pilot basis and strictly monitored for any feeding-
25–27 weeks and 71% if GA was 28–31 weeks [17]. related adverse events like feed intolerance or NEC.
However, none of these studies initiated exclusive The response was encouraging, and unit policy to
enteral feeding on the first day of life. Recently, it start ETEF was implemented from January 2011
has been postulated that total enteral feeding may be based on literature evidence and pilot experience.
started safely on the first day in stable low-risk During Phase 2, eligible babies were prospectively
VLBW infants [18]. This physiological approach included from January to June 2011. In this period,
may be of greater significance in resource-limited set- there was a change in feeding policy to ETEF, where
tings, where cost and logistics of providing TPN are the day’s fluid requirement was provided as enteral
usually the limiting factors. Hence, this uncontrolled feeding with 80 ml/kg on Day 1 of life. Feeding vol-
before and after study was undertaken to assess the ume of this magnitude (80 ml/kg on Day 1 with
feasibility, efficacy and sustainability of early total en- daily increments of 20 ml/kg/day to reach 150 ml/
teral feeding (ETEF) at a tertiary neonatal centre kg on Day 5/Day 6 of life) was achieved with bio-
with high patient load. logical mother’s expressed breast milk predominantly
along with remaining deficit fulfilled by preterm for-
MATERIALS AND METHODS mula (80 kcal/100 ml). Bolus feeds were provided
This uncontrolled before and after study was con- every 2 h through orogastric tube with prefeed ab-
ducted at a tertiary care teaching institution. The dominal girth (AG) charting and abdominal assess-
ethics committee of the institute approved the study ment. With this regimen, the neonate did not receive
protocol. any IV fluid. During the second half of 2011 (July–
Study subjects: Stable preterm infants (GA 28– December), the unit continued with the practice of
34 weeks, birth weight 1000–1499 g), not requiring ETEF. Phase 3 consisted of evaluation of sustenance
resuscitation beyond initial steps, with normal hae- of ETEF practice from January to June 2012.
modynamic status (capillary refill time <3 s and Case management: Apart from hemodynamic
mean arterial pressure normal for GA as per and vital monitoring, prefeed blood sugar monitoring
Zubrow’s chart) and absence of significant respira- for all neonates was done by heel-prick method. The
tory compromise [(no requirement of respiratory frequency of monitoring was every 4 h in the initial
26  Early Total Enteral Feeding in VLBW Neonates

72 h of life and thereafter every 8 h till the end of first 4. Clinical sepsis: Clinical signs and symptoms
week. Sepsis screen and blood culture were sent at suggestive of sepsis with positive sepsis
admission and in case of any clinical deterioration. screen
Abdominal x-ray and ultrasound were done to look 5. Culture-proven sepsis: Blood culture-proven
for evidence of NEC in suspected cases. Standard sepsis in neonate with compatible signs and
management protocols as per the unit policy were symptoms
followed for other clinical problems. There was no
major change or revision in the unit protocols during Statistical Analysis

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019

the study period. All baseline and outcome data were recorded in a
predesigned pro forma, and known variables
Outcome measures were compared between Phase 1 (before) and Phase
The primary objective was to assess the feasibility 3 (after).
and sustainability of ETEF by assessing the day of The data were entered in Excel datasheet, and
achievement of full enteral feeds, incidence of feed coded and analysed statistically using software ver-
intolerance and incidence of NEC. Secondary object- sion 11.1 (StataCorp, College station, Texas, USA).
ives included all-cause mortality, incidence of clinical Descriptive data were analysed using means and SD.
and culture-proven sepsis and durations of antibiotic Continuous data with normal distribution were ana-
therapy, IV fluid therapy and hospital stay. lysed by ‘t’ test and non-normally distributed data by
Wilcoxon rank sum test (Mann–Whitney).
Definitions Categorical data were analysed using v2 test or
1. Full feed achievement: total enteral intake Fisher’s exact test. The groups were compared for
of 150 ml/kg/day sustained for 24 h. continuous variables by one-way analysis of variance
2. Feed intolerance: presence of one or more test. A ‘p’ value of <0.05 was taken as significant.
of the following:
a. Vomiting more than three times during RESULTS
any 24 h period A total of 208 infants were included in the study dur-
b. Any episode of bile- or blood-stained ing the 6 month periods of consecutive years 2010–
vomiting 2012. Phase-wise distribution was 73, 51 and 84 in
c. AG increase >2 cm (measured at umbil- Phases 1, 2 and 3, respectively (Fig. 1).
icus—prefeed) Demographic and baseline characteristics of study
d. Abdominal wall erythema or tenderness subjects during the three phases were comparable
e. Gross or occult blood in stools (Table 1).
Prefeed aspirates were ordered if the AG increased
by >2 cm; if the aspirate was milky and <50% of the Primary outcome
previous feed volume, then the feed was reintro- With ETEF, there was a statistically significant de-
duced and one feed was omitted, but if milky and crease in days required to achieve full feeds along
>50%, then the feeds were omitted for 24 h. with an earlier regain of birth weight (p ¼ 0.0001)
Subsequent feed decisions were based on abdominal (Table 2). Moreover, ETEF was not associated with
assessment and girth measurement. If the aspirates any increase in feed intolerance or NEC.
were haemorrhagic or bilious, then the feeds were
omitted and neonates evaluated for NEC, sepsis and Secondary outcome
surgical condition. ETEF was associated with a marked decrease in inci-
dence of both clinical and culture-proven sepsis
3. NEC was suspected in infants with abdom- (92 and 44% in Phase 1 to 23 and 3.5% in Phase 3;
inal or systemic symptoms and signs and p ¼ 0.0001). This was accompanied by a significant
staged as per modified Bell’s classification reduction in duration of antibiotic therapy and IV
[21]. fluid administration (p ¼ 0.0005). Days to regain
 Early Total Enteral Feeding in VLBW Neonates  27

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019

Fig. 1. Trial flow of study.

Table 1. Baseline characteristics of study subjects

Parameter Phase 1 (N ¼ 73) Phase 2 (N ¼ 51) Phase 3 (N ¼ 84) p value

Gestation (weeks)* 31.561.1 32.5 6 0.78 32.2 6 2.3 0.30

Birth weight (grams)* 12436192 1341 6 182 1252 6 220 0.078
Male sex# 41 (56%) 21 (42%) 46 (55%) 0.50
Registration status in antenatal clinic—booked# 59 (77%) 43 (84%) 73 (87%) 0.15
APGAR score§ 7 (6–8) 7 (7–7) 8 (7–8) 0.38
Small for gestational age# 27 (37%) 18 (35%) 33 (40%) 0.50
Antenatal steroids course# 48 (66%) 42 (72%) 66 (82%) 0.79

Note: *Mean (95% confidence interval), #Number (%), §Median (interquartile range).
28  Early Total Enteral Feeding in VLBW Neonates

Table 2. Outcome parameters

Parameter Phase 1 (N ¼ 73) Phase 2 (N ¼ 51) Phase 3 (N ¼ 84) p value

Day of full feed achievement* (days) 14.446 6.2 8.976 4.9 5.476 1.8 0.0001
Day of regaining birth weight* (days) 16.467.6 14.166.5 12.365.8 0.0006
Incidence of feed intolerance# 16 (22%) 7 (14%) 12 (14%) 0.28
Incidence of NEC# 10 (14.2%) 2 (4%) 0 0.028
Incidence of clinical sepsis# 67 (92%) 24 (47%) 19 (23%) 0.0001

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019

Incidence of culture-proven sepsis# 32 (44%) 6 (12%) 3 (3.5%) 0.0001
Duration of antibiotic therapy* (days) 11.266.8 4.366.1 2.164.2 0.0001
Duration of IV therapy*(days) 12.165.7 6.4763.2 1.560.4 0.0005
Duration of hospital stay*(days) 28.0466.76 19.4765.22 15.564.04 0.0005
Mortality# 3 (4%) 1 (2%) 1 (1.2%) 0.18

Note: *Mean 6 SD, #Number (%).

birth weight and duration of hospital stay were sig- by Sanghvi et al. [24], which initiated full enteral
nificantly lower post-intervention. No episodes of feeding on Day 1 of life. In Sanghvi’s study, the
hypoglycaemia were recorded in any of the three group started on full enteral feeds on Day 1 regained
phases. All-cause mortality remained comparable be- birth weight earlier (5.52 vs. 12.7 days) with a
tween the three study periods (Table 2). shorter duration of hospital stay and no increased
risk of NEC. However, this study included babies be-
DISCUSSION tween 1200 and 1500 g with a much smaller sample
This study strongly suggests the nutritional and size. Nonetheless, it suggests ETEF as being a re-
growth benefits of total enteral feeding introduced sourceful and relatively safe practice, especially in a
from Day 1 of life in stable preterm VLBW infants resource-poor set-up.
without significant gastrointestinal or infectious com- In the present study, ETEF was also associated
plications. The results of this study when taken to- with a significant reduction in both clinical and cul-
gether with other recent works suggest the potential ture-proven sepsis, thereby limiting the need for pro-
benefits of total enteral feeding outweighing the un- longed IV antibiotics and cannulations. The
proven risks of NEC in stable VLBW infants. observational study of Hartel et al. [25] found that
Feeding intolerance and increased length of time VLBW infants born in centres with slow advance-
to reach full enteral feedings are significantly associ- ment of feeds had a significantly higher rate of sepsis
ated with a poorer mental outcome in preterm neo- compared with centres with rapid feed advancement,
nates at 24 months corrected age [22]. In a which was particularly evident for late-onset sepsis
Cochrane meta-analysis of nine studies assessing the (14.0 vs. 20.4%; p ¼ 0.002). Furthermore, higher
role of trophic feeding on number of days to reach usage of central venous lines (48.6 vs. 31.1%,
full feeds, the weighted mean difference was lower p < 0.001) and antibiotics (92.4 vs. 77.7%,
by 2.55 days in the trophic feeding group [23]. The p < 0.001) was seen in centres with slow advance-
results of our study also show faster achievement of ment. Flidel-Rimon et al. [26] also concluded that
full feeds and faster regaining of birth weight. early enteral feeding was associated with a reduced
Despite introduction of full feeds, the incidence of risk of nosocomial sepsis. The possible mechanisms
NEC decreased significantly along with significantly involved include prevention of gastrointestinal atro-
shorter duration of hospital stay decreasing parental phy, prevention of alteration in gut flora and associ-
concern and economic burden in a resource-limited ated overgrowth of enteropathogenic species,
setting. These results are congruous with pilot study promotion of mucosal immunity by gut-associated
 Early Total Enteral Feeding in VLBW Neonates  29

lymphoid tissue; and decreased use of TPN and IV 7 McClure RJ, Newell SJ. Randomized controlled study of
catheters, thereby preserving the skin integrity [26]. clinical outcome following trophic feeding. Arch Dis Child
Being a before and after study, the current study Fetal Neonatal Ed 2000;82:F29–33.
8 Schanler RJ, Shulman RJ, Lau C, et al. Feeding strategies
suffers from the limitations of a non-experimental
for premature infants: randomized trial of gastrointestinal
study with the first phase being a retrospective chart priming and tube-feeding method. Pediatrics
review-based data, potentially overestimating the ef- 1999;103:434–9.
fectiveness of ETEF. Although efforts were made to 9 Berseth CL, Bisquera JA, Paje VU. Prolonging small feed-
ensure consistency in the measurements over similar ing volumes early in life decreases the incidence of necrotiz-

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019

time with no major environmental or policy change ing enterocolitis in VLBW infants. Pediatrics
(except for the feeding regimen) to prevent any 2003;111:529–34.
threats to internal validity of the study, sustained im- 10 Caple J, Armentrout D, Huseby V, et al. Randomized con-
trolled trial of slow versus rapid feeding volume advance-
provement in patient care processes over the years
ment in preterm infants. Pediatrics 2004;114:1597–600.
may have contributed in some ways to the better re- 11 Karagianni P, Briana DD, Mitsiakos G, et al. Early versus
sults in the last two phases. In view of the encourag- delayed minimal enteral feeding and risk for NEC in pre-
ing results of this study, we are tempted to term growth restricted infants with abnormal antenatal
recommend early initiation of total enteral feeding in Doppler results. Am J Perinatol 2010;27:367–73.
stable VLBW infants. However, it would be worth 12 Krishnamurthy S, Gupta P, Debnath S, et al. Slow versus
exploring the same in a well-designed randomized rapid enteral feeding advancement in preterm newborn
controlled trial with adequate numbers to substanti- infants1000–1499 g: a randomized controlled trial. Acta
Paediatr 2010;99:42–6.
ate the findings.
13 Rayyis SF, Ambalavanan N, Wright L, et al. Randomized
trial of "slow" versus "fast" feed advancements on the inci-
CONCLUSION dence of NEC in VLBW infants. J Pediatr 1999;134:293–7.
In stable preterm VLBW infants (1000–1499 g), 14 Salhotra A, Ramji S. Slow versus fast enteral feed advance-
total enteral feeding initiated on Day 1 of life appears ment in very low birth weight infants: a randomized control
safe without an increased risk of NEC and with trial. Indian Pediatr 2004;41:435–41.
15 Ostertag SG, LaGamma EF, Reisen CE, et al. Early feeding
benefits of optimizing nutrition, avoidance of paren-
does not affect the incidence of necrotizing enterocolitis.
teral fluid therapy and inadvertent antibiotic usage
Pediatrics 1986;77:275–80.
along with a significantly shorter duration of hospital 16 Morgan J, Young L, McGuire W. Slow advancement of enteral
stay. feed volumes to prevent necrotising enterocolitis in very low
birth weight infants. Cochrane Database Syst Rev
REFERENCES 2014;12:CD001241. doi:10.1002/14651858.CD001241.pub5.
1 Thureen JP. Early aggressive nutrition in the neonate. 17 Klingenberg C, Embleton ND, Jacobs SE, et al. Enteral
Pediatr Rev 1999;20:45–55. feeding practices in very preterm infants: an international
2 American Academy of Pediatrics, Committee on Nutrition. survey. Arch Dis Child Fetal Neonatal Ed
Nutritional needs of low-birth-weight infants. Pediatrics 2012;97:F56–61.
1985;75:976–86. 18 Embleton ND. When should enteral feeds be started in pre-
3 Dusick AM, Poindexter BB, Ehrenkrantz RA, et al. Growth term infants? Paediatr Child Health 2008;18:200–1.
failure in the preterm infant: can we catch up? Semin 19 Zubrow AB, Hulman S, Kushner S, et al. Determinants of
Perinatol 2003;27:302–10. blood pressure in infants admitted to neonatal intensive
4 Agostoni C, Buonocore G, Carnielli VP, et al. Enteral nutri- care units: a prospective multicenter study. Philadelphia
ent supply for preterm infants: commentary from the Neonatal Blood Pressure Study Group. J Perinatol
European Society of pediatric Gastroenterology, 1995;15:470–9.
Hepatology and Nutrition Committee on nutrition. 20 Silverman WA, Anderson DH. A controlled clinical trial of
J Pediatr Gastroenterol Nutr 2010;50:85–91. effects of water mist on obstructive respiratory signs, death
5 Neu J, Zhang L. Feeding intolerance in very-low-birth rate and necropsy findings among premature infants.
weight infants: what is it and what can we do about it? Acta Pediatrics 1956;17:1–10.
Paediatr Suppl 2005; 94:93–9. 21 Walsh MC, Kliegman RM. Necrotizing enterocolitis: treat-
6 Schanler RJ. Suitability of human milk for the low birth ment based on staging criteria. Pediatr Clin North Am
weight infant. Clin Perinatol 1995;22:207–22. 1986;33:179–201.
30  Early Total Enteral Feeding in VLBW Neonates

22 Patole S. Strategies for prevention of feed intolerance in 25 H€artel C, Haase B, Browning-Carmo K, et al. Does the en-
preterm neonates: a systematic review. J Maternal Fetal teral feeding advancement affect short-term outcomes in
Neonatal Med 2005;18:67–76. very low birth weight infants? J Pediatr Gastroenterol Nutr
23 Tyson JE, Kennedy KA. Trophic feedings for parenterally 2009;48:464–70.
fed infants. Cochrane Database Syst Rev 2009;3:CD000504. 26 Flidel-Rimon O, Friedman S, Lev E, et al. Early en-
24 Sanghvi KP, Joshi P, Nabi F, et al. Feasibility of exclusive teral feeding and nosocomial sepsis in very low birth-
enteral feeds from birth in VLBW infants >1200 g–an weight infants. Arch Dis Child Fetal Neonatal Ed
RCT. Acta Paediatr 2013;102:e299–304. 2004;89:289–92.

Downloaded from https://academic.oup.com/tropej/article-abstract/64/1/24/3739807 by guest on 26 February 2019