Three infections cannot settle fluconazole prophylaxis

After matching, late-onset invasive fungal infection occurred in 2 of 442 fluconazole recipients and 1 of 442 controls. Low baseline risk matters, but three events cannot prove no benefit.

The hot take

In NICUs with a very low incidence of late-onset IFI, routine fluconazole prophylaxis is not recommended for very preterm infants/VLBW infants.

The study behind that conclusion found late-onset invasive fungal infection in 2 of 442 prophylaxis recipients and 1 of 442 matched controls. Fisher's exact P was 1.000. The observed baseline was genuinely low, but three matched infections cannot establish equivalence, noninferiority, or proof that prophylaxis has no benefit.

What the cohort actually did

This retrospective cohort included 1,961 infants admitted before day 3 of life to three NICU wards at two Sichuan medical institutions from 2018 through 2023. Eligibility required birth weight under 1,500 g or gestational age under 32 weeks, plus a stay longer than 3 days. Fluconazole prophylaxis was selected empirically by clinicians for 599 infants; 1,362 did not receive it. Propensity-score matching retained 442 infants in each group.