Despite availability of rapid fungal potassium hydroxide (KOH) tests, many care providers rely on visual assessment to determine the diagnosis of monilial diaper dermatitis (MDD).
To determine whether a KOH test, when MDD is suspected, would result in more accurate diagnoses, with decreased antifungal medication prescription and exposure.
Quality improvement project from 2016 through 2017 with protocol implemented in 2017 for treatment of MDD after positive KOH testing. If monilial rash suspected, after 2 negative KOH tests, then antifungal ordered (considered false negative). χ2 testing and cost determination were performed.
Neonates in 2 level III neonatal intensive care units.
KOH test results, use of antifungal medication, and cost.
The patient census included 1051 and 1015 patients in the year before and after the protocol initiation. The medical orders for antifungal medication decreased from 143 to 36 (P < .001; 95% odds ratio confidence interval, 2.24-4.38). There was a 75% reduction in both use and cost, as charged, of antifungal agents. Overall charges, including KOH test costs, decreased by 12%. Three infants received multiple negative KOH tests, then a positive one. These met the definition of false-negative tests, per protocol. There were no cases of fungal sepsis.
Use of a quality improvement protocol, in which the use of KOH testing is required, before antifungal agents are prescribed, results in decreased exposure and costs.
To test the feasibility of bedside “point-of-care” KOH testing, and whether KOH testing and reduced antifungal medication use affects antimicrobial resistance or invasive fungal sepsis.
Department of Neonatology, Seattle Children's Hospital, Washington (Mss Campbell and Zuill and Drs Sacco and Bosque); and Neonatal Intensive Care Unit, Providence Regional Medical Center, Everett, Washington (Ms MacConnell).
Correspondence: Elena Bosque, PhD, ARNP, NNP-BC, Department of Neonatology, Seattle Children's Hospital, PO Box 5371/M1-12, Seattle, WA 98145 (firstname.lastname@example.org).
The authors declare no conflicts of interest.