Predictors of Failure of Conventional Dose Caffeine Therapy in Neonates Below 32 Weeks' Gestation: An Ambispective Cohort Study

Over half of all preterm neonates started on conventional maintenance-dose caffeine therapy require dose escalation. This study aimed to determine the predictors of failure of conventional maintenance-dose caffeine therapy for apnea of prematurity, which may guide individualized dosing regimens.

In this ambispective cohort study, the authors enrolled neonates <32 wk gestation receiving caffeine therapy, retrospectively (Jan 2019-Dec 2020) and prospectively (Aug 2021-Dec 2022). Conventional dose was defined as caffeine citrate, 20 mg/kg loading followed by 5 mg/kg/d. 'Failure' was defined as clinician-driven escalation of maintenance-dose. Predictor variables differing between the failure and success groups were analyzed using multivariable logistic regression. Additional outcomes included prediction of plasma caffeine levels using regression analysis of clinical variables.

A total of 164 neonates (70 prospective, 94 retrospective) were enrolled (gestation 28.7 ± 1.8 wk). Failure occurred in 94 (57%; 95% CI 49-65) neonates. Only gestation independently predicted caffeine failure on multivariable analysis; each one-week increase in gestation nearly halved the odds of treatment failure (aOR: 0.55; 95% CI 0.41-0.72). Neonates with gestation <28 wk had significantly higher odds of having caffeine failure compared to those ≥28 wk gestation (92% vs. 46%; OR 14.5, 95% CI 4.2-49.5). Regression model showed a significant positive association between caffeine dose and plasma caffeine levels (β = 0.59, p <0.001); other variables were not statistically significant.

In neonates below 32 wk gestation, lower gestation predicts failure of conventional-dose caffeine therapy. Higher initial maintenance doses warrant further exploration in neonates below 28 wk.