Compounding a preservative-free caffeine citrate solution for a preterm neonate
A 1.8 kg preterm neonate (gestational age 28 weeks, now 5 days old) in the neonatal intensive care unit is to start caffeine for apnea of prematurity. The order reads: 'caffeine 20 mg/kg loading dose once now, then 5 mg/kg once daily starting in 24 hours.' The unit's standard reference converts these to caffeine citrate. The commercial caffeine citrate 20 mg/mL oral solution is on national backorder with no release date. Your hospital pharmacy holds caffeine citrate USP powder, purified water, and a Class A balance, and can prepare a non-sterile oral solution. The pharmacist on service proposes making 60 mL of a 20 mg/mL caffeine citrate solution, preservative-free, for this infant, and asks you to review the plan before preparation. The infant is currently on 24 mL/kg/day of enteral feed and is also receiving a benzyl alcohol–containing heparin flush.
1. Is the order clear enough to prepare from? Explain the salt/form issue and calculate the loading and maintenance doses in the units you would use.
Read the worked answer
The order is not safe to prepare from as written. 'Caffeine 20 mg/kg' does not specify whether it means caffeine citrate or caffeine base, and the two differ by a factor of two: 20 mg caffeine citrate = 10 mg caffeine base. The standard apnea-of-prematurity dosing (20 mg/kg load, 5–10 mg/kg/day maintenance) is conventionally expressed as caffeine citrate.
Loading dose (as caffeine citrate): 20 mg/kg × 1.8 kg = 36 mg caffeine citrate (= 18 mg caffeine base). From a 20 mg/mL caffeine citrate solution that is 1.8 mL.
Maintenance (as caffeine citrate): 5 mg/kg × 1.8 kg = 9 mg caffeine citrate (= 4.5 mg caffeine base) = 0.45 mL of the 20 mg/mL solution daily.
If the order were interpreted as caffeine base, the same numeric dose would deliver roughly double the intended citrate dose - a real risk of tachycardia, jitteriness, poor feeding and, at higher exposures, seizures.
Before preparing anything, have the order clarified and rewritten as 'caffeine citrate 20 mg/kg (1.8 mL of 20 mg/mL) loading, then 5 mg/kg (0.45 mL) daily' and label both the bag and the MAR with the salt and the concentration. The dose volume of 0.45 mL is small; an appropriate oral syringe must be supplied so the nursing measurement is accurate.
2. Critique the proposed 60 mL batch: quantity of powder, formulation issues in a neonate, beyond-use date and labelling.
Read the worked answer
Quantity: 60 mL × 20 mg/mL = 1200 mg (1.2 g) caffeine citrate. That is weighable on a Class A balance with acceptable accuracy (well above the minimum weighable quantity), but it should be weighed by two people or checked independently, and the purity/assay and water content of the USP powder referenced in the calculation.
Formulation issues specific to a 1.8 kg preterm infant:
- No preservative. Benzyl alcohol has caused gasping syndrome and metabolic acidosis in neonates, and parabens are also avoided; the infant is already receiving benzyl alcohol via heparin flushes, so total exposure matters. A preservative-free aqueous solution is the right choice.
- pH and stability. Caffeine citrate solutions are pH-sensitive; a validated stability reference should be consulted rather than relying on appearance.
- Osmolality and volume. The 0.45 mL maintenance dose is trivial relative to 24 mL/kg/day of feeds, but the load should be given undiluted or per unit protocol, not added to a full feed.
- Microbial risk. A preservative-free aqueous oral liquid is a growth medium; a 60 mL batch for one 1.8 kg infant is far more than will be used within the BUD. A smaller batch is preferable.
Beyond-use date: under the default non-sterile compounding limits, a non-preserved water-containing oral preparation is assigned about 14 days refrigerated unless a published stability study supports longer. If the pharmacy instead uses a commercially available preserved vehicle with documented stability, the BUD may be longer - but that reintroduces excipients inappropriate for this neonate, so do not solve the BUD problem that way.
Labelling: 'Caffeine citrate 20 mg/mL (= caffeine base 10 mg/mL). Preservative-free. Refrigerate. Shake. Discard after [date].' Include the batch number and the patient-specific dose in mL.
3. What monitoring and follow-up would you build into the plan, and what would you do if the pharmacy could not meet an appropriate quality standard?
Read the worked answer
Monitoring: heart rate and rhythm (tachycardia is the earliest sign of excess), apnea and bradycardia recurrence, feeding tolerance and gastric residuals, weight trajectory, jitteriness, and seizure activity. Caffeine has a long and variable half-life in preterm neonates (measured in days), so accumulation is real: the maintenance dose is the part to reassess, and therapeutic drug monitoring is not routine but is used when toxicity or non-response is suspected.
Follow-up: confirm the first maintenance dose is actually given 24 hours after the load, verify the nurse draws up 0.45 mL into an oral syringe rather than a teaspoon, and re-check the concentration on the label against the order whenever the batch is replaced - batch changes are a classic source of concentration error.
If quality cannot be assured: do not improvise. Options in order of preference are (1) locate remaining commercial stock from another site or a regional supply arrangement; (2) have the preparation made by a licensed compounding facility with appropriate quality control, release testing and documentation; (3) use an alternative agent with an available commercial formulation per the neonatologist, recognizing that alternatives have less evidence in this population. The decision should be documented and communicated to the prescriber and the NICU pharmacist, with the reason for the choice recorded - 'we made it ourselves' is not an adequate quality argument for a 1.8 kg infant.