Association of Early Versus Late Caffeine Administration on Neonatal Outcomes in Very Preterm Neonates


Lodha A, Seshia M, McMillan DD, Barrington K, Yang J, Lee SK, Shah PS for the Canadian Neonatal Network. Association of Early Caffeine Administration and Neonatal Outcomes in Very Preterm Neonates. JAMA Pediatr. 2015;169(1):33-38. PMID: 25402629


Stephanie Adzikah, MD
Julia Maletzki, MD
Christoph Rüegger, MD
Dirk Bassler, MD MSc
Department of Neonatology
University Hospital of Zurich




In very preterm neonates born at less than 31 weeks´ gestation the administration of early (< 3 days after birth) versus late (≥ 3 days after birth) caffeine therapy was compared regarding the composite primary outcome of death or BPD at 36 weeks of gestational age.


  • Design: retrospective cohort study
  • Allocation: Observational study so not applicable
  • Blinding: observational, does not apply
  • Follow-up Period: until death or discharge from NICU
  • Setting: 29 tertiary-level NICUs participating in the Canadian Neonatal Network (CNN)
  • Patients:
    • Inclusion criteria: all neonates born at less than 31 weeks` gestation who were admitted to one of the 29 participating tertiary-level NICUs in the CNN between January 1st, 2010 and December 31st, 2012
    • Exclusion criteria: neonates born outside a tertiary-level NICU, moribund neonates, neonates with major congenital anomalies, neonates who died before day 3 after birth
  • Intervention: Patients who received caffeine were identified from the CNN database and divided into two groups: early caffeine administration (defined as neonates with receipt of caffeine within the first two days of life) and late caffeine administration (defined as patients receiving caffeine on or later than day 3 of life). The day of birth was defined as day of life 1. Differences in daily or cumulative doses were not taken into account.
  • Outcomes:
    • Primary outcome:
      • Composite of death and bronchopulmonary dysplasia defined as supplemental oxygen use at 36 weeks’ post- menstrual age or at discharge from the NICU.
    • Secondary outcomes:
      • Death, Bronchopulmonary dysplasia (BPD), severe retinopathy of prematurity, severe neurological injury, periventricular echogenicity persisting beyond 21 days of age or periventricular leukomalacia, persisting ductus arteriosus (PDA), necrotizing enterocolitis, ventilation days, oxygen days
  • Sample size and analysis:
    • No sample size calculation was performed; all infants in the CNN that met entry criteria were included
    • Unadjusted analysis was used to compare maternal and infant demographic characteristics, perinatal risk factors, and the incidence of mortality and morbidity among neonates in the early and late groups (Pearsons X2, t test, Wilcoxon rank test)
    • Multivariate logistic regression analysis was used to examine the effect of significant outcomes from the univariate analysis of the early versus late caffeine group, with adjustment for various clinically significant predictors of neonatal morbidity and mortality: gestational age, intubation on day 2 after birth, study site, SNAP-II severity score, and surfactant.
  • Patient follow-up:
    • Of 38,163 neonates admitted to participating NICUs, 7,274 were born at less than 31 weeks’ gestational age. 5,517 patients were eligible for the study of whom 5101 (92.4%) received caffeine.
    • 3,806 (74.6%) were included in the early- (<3d) and 1,295 (25.4%) in the late (≥3d) caffeine group.


Table 1. Characteristics of Preterm Infants Included in the Studya

Baseline Characteristics Caffeine Group, No. (%) P Value
Early (n = 3806) Late (n = 1295)
Antenatal steroids 3533 (94.4) 1173 (92.9) 0.05
Multiples 1228 (32.3) 366 (28.3) 0.01
Cesarean section 2321 (61.3) 778 (60.7) 0.72
Gestation age, median (IQR), week-      ≤24-      25-28-      29-30 28 (26-29)54 (1.4)2273 (59.7)1479 (38.9) 28 (26-30)32 (3.2)717 (55.4)535 (41.3) 0.07<0.01
Birth weight, median (IQR), g 1070 (850-1310) 1050 (790-1360) 0.66
Small for gestational age (<10th percentile) 340 (9.0) 139 (10.8) 0.06
Female 1749 (46.1) 565 (43.8) 0.15

Abbreviations: IQR, interquartile range; a Born at less than 31 weeks’ gestational age.

Table 2. Characteristics of Preterm Infants Included in the Studya

Postnatal Characteristics Caffeine Group, No. (%) P Value
Early (n = 3806) Late (n = 1295)
Apgar score at 5 min, median (IQR) 8 (6-9) 7 (6-8) <0.01
SNAP-II score, median (IQR) 9 (5-16) 12 (5-22) <0.01
Intubation at birth 1363 (38.3) 427 (34.8) 0.03
Surfactant 1557 (43.8) 536 (43.7) 0.98
Duration of caffeine, median (IQR), d 37 (21-55) 30 (14-51) <0.01

Abbreviations: IQR, interquartile range; SNAP II, Score for Neonatal Acute Physiology version II. a Born at less than 31 weeks’ gestational age.

Table 3. Non Pre-Specified Secondary Outcomes/Co-Interventions

Postnatal Characteristics Caffeine Group, No. (%) P Value
Early (n = 3806) Late (n = 1295)  
Postnatal steroids 590 (15.5) 368 (28.4) <0.01
Air leak syndrome 124 (3.3) 74 (5.9) <0.01
Conventional ventilation on d 2 1496 (39.3) 496 (38.3) 0.52
High-frequency ventilation on d 2 236 (6.2) 251 (19.4) <0.01
CPAP on d 2 1447 (38.0) 350 (27.0) <0.01

Abbreviation: CPAP, continuous positive airway pressure.

Table 4. Comparison of Outcomes Between Preterm Infants Administered Caffeine Either Within (Early Group) or After (Late Group) the First 2 Days of Life

Variable Caffeine Group, No. (%) Adjusteda Odds Ratio(95% CI)
Early (n = 3806) Late (n = 1295)
Death or BPD at 36 wks 1197 (31.5) 403 (31.1) 0.81 (0.67-0.98)
Death 217 (5.7) 75 (5.8) 0.98 (0.70-1.37)
BPDb–      36 wks-      28 d 999 (27.8)1535 (42.2) 340 (27.7)502 (40.2) 0.79 (0.64-0.96)0.90 (0.74-1.09)
PDA 1503 (40.5) 576 (46.2) 0.74 (0.62-0.89)
PDA, No./No.(%)-      Treated with Indomethacin and Ibuprofen-      Treated with surgical ligation 847/1503 (56.3)200/1503 (13.3) 305/576 (52.9)144/576 (25.0) 1.17 (0.92-1.49)0.58 (0.42-0.80)
ROP (≥stage 3) 229 (9.8) 95 (12.8) 0.78 (0.56-1.10)
NEC (≥stage 2) 240 (6.5) 78 (6.3) 0.88 (0.65-1.20)
Severe neurological injuryc 432 (12.0) 167 (14.3) 0.80 (0.63-1.01)

aAdjusted for gestational age, antenatal steroid exposure, small for gestation age, site, intubated on day 2, SNAP-II score, and surfactant administration. bDefined as oxygen needed. cSevere neurological injury included the following: presence of parenchymal echolucency, periventricular echogenicity, or periventricular leukomalacia.


The authors conclude that early initiation of caffeine is associated with a reduction in the rate of death or bronchopulmonary dysplasia and patent ductus arteriosus in neonates born at less than 31 weeks’ gestational age. In addition, prophylactic administration of caffeine had no adverse impact on any other outcome.


Caffeine is commonly used for preterm infants with apnea of prematurity. The caffeine for apnea of prematurity trial (CAP) was launched to determine whether survival without neurodevelopmental disability at a corrected age of 18 months is improved if apnea of prematurity is managed without methylxanthines in infants at a high risk of apneic attacks (1). Infants had to be within 10 days of birth. Forty percent of the patients assigned to caffeine died or survived with a neurodevelopmental disability compared with 46% assigned to placebo (OR 0.77, 95% CI 0.64-0.93; p = 0.008). Physicians were not asked to alter their own prescribing indications for caffeine. However only 22.5% of the neonatologists in the trial started caffeine to prevent apnea (2). Thus, the risk-benefit ratio of caffeine administered exclusively for prophylaxis and therefore likely earlier – remains unclear. This was emphasised by Schmidt in a recent editorial (3) commenting upon a large observational cohort study that showed that early administration decreased the incidence of BPD in their US group of NICUs (4). This is in line with another retrospective data analysis by Taha et al. (5). In addition, a subgroup analysis of the CAP trial demonstrated that early caffeine treatment correlated with a larger reduction in days of respiratory support compared to late caffeine (6).
In this third large-scale observational study of the Canadian Neonatal Network, Lodha et al. retrospectively investigated the effect of early versus late initiation of caffeine therapy on neonatal short-term outcomes of very preterm infants. The nation-wide participation resulted in an impressive number of included infants. Nevertheless, there are a few comments to make.
First, the diagnosis of a PDA was based on clinical signs, with or without echocardiography. However several studies have demonstrated that echocardiograms are required for the diagnosis of a PDA in preterm infants, as clinical signs are not reliable in the first few days of life (7, 8).

Second, there are imbalances between the infants of the early and those of the late treatment group. We also suggest that some baseline characteristics should rather be declared as outcomes and/or co-interventions (Table 3) due to there timely relation to the intervention. The late treatment group, for example, showed a more frequent use of postnatal steroids and an increased occurrence of air leak syndromes.
The differences in baseline characteristics may have resulted in an early administration of caffeine in those babies without and a late administration in those with major risk factors for adverse outcomes, such as BPD for example. Despite adjustments for known imbalances between groups, observational studies do not have the ability to completely eliminate selection bias and observational studies investigating the effect of drug treatments are prone to confounding by indication.
Thirdly, information concerning the indication to start caffeine is lacking, and no information on dosages is provided.
In conclusion, this is an important study but more data from randomised controlled trials on the optimal timing of caffeine administration are needed before introducing early prophylactic caffeine into routine clinical practice (9).


  1. Schmidt B, Roberts RS, Davis P, Doyle LW, Barrington KJ, Ohlsson A, et al. Long-term effects of caffeine therapy for apnea of prematurity. N Engl J Med 2007; 357:1893-902
  2. Schmidt B, Roberts RS, Davis P, Doyle LW, Barrington KJ, Ohlsson A, et al. Caffeine therapy for apnea of prematurity. N Engl J Med 2006; 354:2112-21
  3. Schmidt B, Davis PG, Roberts RS. Timing of caffeine therapy in very low birth weight infants. J Pediatr 2014; 164:957-8
  4. Dobson NR, Patel RM, Smith PB, Kuehn DR, Clark J, Vyas-Read S, et al. Trends in caffeine use and association between clinical outcomes and timing of therapy in very low birth weight infants. J Pediatr 2014; 164:992-8 e3
  5. Taha D, Kirkby S, Nawab U, Dysart KC, Genen L, Greenspan JS, et al. Early caffeine therapy for prevention of bronchopulmonary dysplasia in preterm infants. J Matern Fetal Neonatal Med 2014; 27:1698-702
  6. Davis PG, Schmidt B, Roberts RS, Doyle LW, Asztalos E, Haslam R, et al. Caffeine for Apnea of Prematurity trial: benefits may vary in subgroups. J Pediatr 2010; 156:382-7
  7. Davis P, Turner-Gomes S, Cunningham K, Way C, Roberts R, Schmidt B. Precision and accuracy of clinical and radiological signs in premature infants at risk of patent ductus arteriosus. Arch Pediatr Adolesc Med 1995; 149:1136-41
  8. Alagarsamy S, Chhabra M, Gudavalli M, Nadroo AM, Sutija VG, Yugrakh D. Comparison of clinical criteria with echocardiographic findings in diagnosing PDA in preterm infants. J Perinat Med 2005; 33:161-4
  9. Kreutzer K, Bassler D. Caffeine for apnea of prematurity: a neonatal success story. Neonatology 2014; 105:332-6