Correlation between plasma and urinary caffeine levels in preterm infants

Background: We hypothesize that urine levels might be reliable to assess the therapeutic range of caffeine. Objectives: We correlated plasma and urinary levels of caffeine in preterm infants treated with this drug for apnea of prematurity. Methods: Infants (n=56) were given a loading dose of caffeine citrate (10 mg/kg, per os) and 24 h later a maintenance dose (2 mg/kg, per os, once a day). Plasma and urinary levels of caffeine were determined 24 h after the loading dose (before administration of the maintenance dose) and then weekly. Results: Plasma and urinary levels correlate at all examined ages: 29 weeks (r=0.92, P < 0.001), 30 weeks (r=0.97, P < 0.001), 31 weeks (r=0.82, P < 0.001), 32 weeks (r = 0.92, P < 0.001), 33 weeks (r = 0.87, P < 0.001), 34 weeks (r = 0.81, P < 0.001). Conclusion: Urinary levels of caffeine might be a useful means to assess therapeutic ranges.