Objective
Diuretics are often prescribed off-label to premature infants, particularly to prevent or treat bronchopulmonary dysplasia (BPD). We examined their use and safety in this group.
Study Design
Retrospective cohort study of infants <32 weeks gestation and <1500 g birth weight exposed to diuretics in 333 neonatal intensive care units from 1997–2011. We examined use of acetazolamide, amiloride, bumetanide, chlorothiazide, diazoxide, ethacrynic acid, furosemide, hydrochlorothiazide, mannitol, metolazone, or spironolactone combination. Respiratory support and FiO2 on the first day of each course of diuretic use were identified.
Results
Thirty-seven percent (39,357/107,542) of infants were exposed to at least 1 diuretic; furosemide was the most commonly used (93% with ≥1 recorded dose), followed by spironolactone, chlorothiazide, hydrochlorothiazide, bumetanide, and acetazolamide. Seventy-four percent were exposed to 1 diuretic at a time, 19% to 2 diuretics simultaneously, and 6% to 3 diuretics simultaneously. The most common combination was furosemide/spironolactone, followed by furosemide/chlorothiazide and chlorothiazide/spironolactone. Many infants were not receiving mechanical ventilation on the first day of each new course of furosemide (47%), spironolactone (69%), chlorothiazide (61%), and hydrochlorothiazide (68%). Any adverse event occurred on 42 per 1000 infant-days for any diuretic and 35 per 1000 infant-days for furosemide. Any serious adverse event occurred in 3.8 for any diuretic and 3.2 per 1000 infant-days for furosemide. The most common laboratory abnormality associated with diuretic exposure was thrombocytopenia.
Conclusion
Despite no FDA indication and little safety data, over one third of premature infants in our population were exposed to a diuretic, many with minimal respiratory support.