Caffeine dosing regimens in preterm infants with or at risk for apnea of prematurity

Background: Very preterm infants often require respiratory support and are therefore exposed to an increased risk of bronchopulmonary dysplasia (chronic lung disease) and later neurodevelopmental disability. Caffeine is widely used to prevent and treat apnea (temporal cessation of breathing) associated with prematurity and facilitate extubation. Though widely recognized dosage regimes have been used for decades, higher doses have been suggested to further improve neonatal outcomes. However, observational studies suggest that higher doses may be associated with harm.

Objectives: To determine the effects of higher versus standard doses of caffeine on mortality and major neurodevelopmental disability in preterm infants with (or at risk of) apnea, or peri-extubation.

Search methods: We searched CENTRAL, MEDLINE, Embase, CINAHL, the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP), and clinicaltrials.gov in May 2022. The reference lists of relevant articles were also checked to identify additional studies.

Selection criteria: We included randomized (RCTs), quasi-RCTs and cluster-RCTs, comparing high-dose to standard-dose strategies in preterm infants. High-dose strategies were defined as a high-loading dose (more than 20 mg of caffeine citrate/kg) or a high-maintenance dose (more than 10 mg of caffeine citrate/kg/day). Standard-dose strategies were defined as a standard-loading dose (20 mg or less of caffeine citrate/kg) or a standard-maintenance dose (10 mg or less of caffeine citrate/kg/day). We specified three additional comparisons according to the indication for commencing caffeine: 1) prevention trials, i.e. preterm infants born at less than 34 weeks' gestation, who are at risk for apnea; 2) treatment trials, i.e. preterm infants born at less than 37 weeks' gestation, with signs of apnea; 3) extubation trials: preterm infants born at less than 34 weeks' gestation, prior to planned extubation.

Data collection and analysis: We used standard methodological procedures expected by Cochrane. We evaluated treatment effects using a fixed-effect model with risk ratio (RR) for categorical data and mean, standard deviation (SD), and mean difference (MD) for continuous data. MAIN RESULTS: We included seven trials enrolling 894 very preterm infants (reported in Comparison 1, i.e. any indication). Two studies included infants for apnea prevention (Comparison 2), four studies for apnea treatment (Comparison 3) and two for extubation management (Comparison 4); in one study, indication for caffeine administration was both apnea treatment and extubation management (reported in Comparison 1, Comparison 3 and Comparison 4). In the high-dose groups, loading and maintenance caffeine doses ranged from 30 mg/kg to 80 mg/kg, and 12 mg/kg to 30 mg/kg, respectively; in the standard-dose groups, loading and maintenance caffeine doses ranged from 6 mg/kg to 25 mg/kg, and 3 mg/kg to 10 mg/kg, respectively. Two studies had three study groups: infants were randomized in three different doses (two of them matched our definition of high dose and one matched our definition of standard dose); high-dose caffeine and standard-dose caffeine were compared to theophylline administration (the latter is included in a separate review). Six of the seven included studies compared high-loading and high-maintenance dose to standard-loading and standard-maintenance dose, whereas in one study standard-loading dose and high-maintenance dose was compared to standard-loading dose and standard-maintenance dose. High-dose caffeine strategies (administration for any indication) may have little or no effect on mortality prior to hospital discharge (risk ratio (RR) 0.86, 95% confidence of interval (CI) 0.53 to 1.38; risk difference (RD) -0.01, 95% CI -0.05 to 0.03; I² for RR and RD = 0%; 5 studies, 723 participants; low-certainty evidence). Only one study enrolling 74 infants reported major neurodevelopmental disability in children aged three to five years (RR 0.79, 95% CI 0.51 to 1.24; RD -0.15, 95% CI -0.42 to 0.13; 46 participants; very low-certainty evidence). No studies reported the outcome mortality or major neurodevelopmental disability in children aged 18 to 24 months and 3 to 5 years. Five studies reported bronchopulmonary dysplasia at 36 weeks' postmenstrual age (RR 0.75, 95% CI 0.60 to 0.94; RD -0.08, 95% CI -0.15 to -0.02; number needed to benefit (NNTB) = 13; I² for RR and RD = 0%; 723 participants; moderate-certainty evidence). High-dose caffeine strategies may have little or no effect on side effects (RR 1.66, 95% CI 0.86 to 3.23; RD 0.03, 95% CI -0.01 to 0.07; I² for RR and RD = 0%; 5 studies, 593 participants; low-certainty evidence). The evidence is very uncertain for duration of hospital stay (data reported in three studies could not be pooled in meta-analysis because outcomes were expressed as medians and interquartile ranges) and seizures (RR 1.42, 95% CI 0.79 to 2.53; RD 0.14, 95% CI -0.09 to 0.36; 1 study, 74 participants; very low-certainty evidence). We identified three ongoing trials conducted in China, Egypt, and New Zealand.

Authors' conclusions: High-dose caffeine strategies in preterm infants may have little or no effect on reducing mortality prior to hospital discharge or side effects. We are very uncertain whether high-dose caffeine strategies improves major neurodevelopmental disability, duration of hospital stay or seizures. No studies reported the outcome mortality or major neurodevelopmental disability in children aged 18 to 24 months and 3 to 5 years. High-dose caffeine strategies probably reduce the rate of bronchopulmonary dysplasia. Recently completed and future trials should report long-term neurodevelopmental outcome of children exposed to different caffeine dosing strategies in the neonatal period. Data from extremely preterm infants are needed, as this population is exposed to the highest risk for mortality and morbidity. However, caution is required when administering high doses in the first hours of life, when the risk for intracranial bleeding is highest. Observational studies might provide useful information regarding potential harms of the highest doses.