Evid Based Med 18:36-37 doi:10.1136/ebmed-2012-100622
  • Aetiology
  • Cohort study

Ranitidine use is associated with increased morbidity and mortality in very low birthweight infants

  1. William F Malcolm
  1. Department of Pediatrics-Neonatology, Duke University, Durham, North Carolina, USA
  1. Correspondence to: C Michael Cotten
    Pediatrics-Neonatology, Duke University, 2424 Erwin Road suite 504, Durham, NC 27705, USA; cotte010{at}

Commentary on: [Abstract/FREE Full text]


Gastroesophageal reflux (GER) is the retrograde movement of stomach contents into the oesophagus. Pharmacotherapy aimed at reducing problems attributed to GER includes histamine receptor blockers (H2 blockers), such as ranitidine, which reduce stomach acidity. Extremely preterm infants often exhibit clinical signs consistent with GER,1 and GER medications are among the most commonly prescribed medications in neonatal intensive care units (NICUs).2

Approximately one third of extremely premature (<28 weeks gestation), very low birthweight (VLBW; <1500 g) infants acquire infections during their NICU stay. Necrotising enterocolitis (NEC), a serious intestinal morbidity in preterm infants, occurs in approximately 10%. About half of NEC cases require surgery; among these, mortality approaches 50%.3 Clinical practices associated with these morbidities may be worthwhile targets for quality improvement efforts that could reduce risk and improve neonatal outcomes.


Terrin et al conducted a prospective observational cohort study to assess association of ranitidine with infection, NEC, and death in a cohort of VLBW infants between 24 and 32 weeks gestation from four NICUs. Ranitidine was started and dosed at the caregiver's discretion. The primary outcome was infection, defined as sepsis (blood culture positive and clinical signs), pneumonia or urinary tract infection. The secondary outcomes were NEC (stage II or greater), death and duration of hospitalisation. Binary logistic regression was used to detect differences in demographics and severity of illness between ranitidine recipients and non-recipients.


Ninety-one of 274 enrolled infants received ranitidine. Demographics and illness severity scores were not different between groups. Infection was more likely among ranitidine recipients (OR 5.5, 95% CI 2.9 to 10.4), as was each component of the composite outcome. NEC was more common among ranitidine recipients (9.8% vs 1.6%; OR 6.6, 95% CI 1.7 to 25.0). Nine ranitidine recipients died versus three non-recipients (p=0.003). Ranitidine recipients had longer length of stay (median 52 vs 36 days; p<0.001).


This prospective study reinforces prior studies cited by the authors identifying the association between ranitidine exposure and morbidities and mortality of extremely premature infants. The current study is notably the first prospective cohort study. Incidence of morbidities, like NEC and infection, vary widely among NICUs, as does use of H2 blockers.3 ,4 One potential pitfall of this study related to centre variations is the lack of information about variation in prevalence of rantidine use and outcomes among centres. If infection and NEC rates are highest where hand hygiene is poor, mother's milk feeding is less prevalent and ranitidine use is high, confidence for attributing hazard risk to ranitidine alone would be significantly reduced.

Speculation on the mechanism for ranitidine's potentially deleterious effects focuses on potential influence on gastric and intestinal colonisation. The authors cite two reports supporting this, but neither measured microbial colonisation. In a randomised trial, Carrion and Egan compared infants <1250 g birthweight fed slightly acidified milk versus infants fed non-acidified milk. One of 34 fed acidic feeds developed NEC versus 8 of 34 controls (p=0.02). Gastric pH was lower (3.0) in acid-exposed infants versus controls (4.0), and gastric enteric colonisation rate and quantitative bacterial counts were correlated with gastric pH>4 (p<0.001).5 In a study assessing gastric colonisation with pathogenic bacteria or yeast, 20 of 23 ranitidine-treated versus 34 of 63 untreated premature infants were colonised.6 Manipulation of gastric pH appears to influence gastric colonisation, but mechanisms explaining the biological links between gastric pH, colonisation and clinical outcomes are not yet known.

While we await molecular, microbial and physiologic evidence to characterise mechanisms linking antacid to complications, this study adds to the current epidemiologic evidence indicating lack of benefit and potential harms from ranitidine use in extremely preterm infants during their NICU stay. Clinicians should avoid its empirical use until better data, for example from randomised clinical trials, provides pharmacokinetic, pharmacodynamic, safety and efficacy data for this population.


  • Competing interests WFM has received funding from the Gerber Foundation for studies of gastroesophageal reflux. CMC has no conflicting interests.