The episode unfolded at the eleven-bed neonatal intensive care unit operated by the Department of Pediatrics and Adolescent Medicine of AMEOS Klinikum St. Elisabeth Neuburg within Ingolstadt Hospital, a unit that treats approximately 150 patients annually, including extremely preterm infants born from 23+0 weeks of gestation. Routine surveillance in the ward involves weekly swabs from the axillae, nose, ear, and rectum of all patients, with additional samples taken from wounds, suspected skin lesions, or the affected eye when clinically indicated. Although the screening program primarily targets multi-drug-resistant organisms under national German infection-control guidelines, the culture procedures also detect susceptible potentially pathogenic bacteria, including S. aureus, and these findings are routinely reported and reviewed by the pediatric and infection-control teams.

To define the temporal boundaries of the event, the investigators retrospectively analyzed microbiological records from January 2021 through December 2025, counting only the first detection of penicillin-susceptible S. aureus per patient. A cumulative sum control chart, a statistical tool that flags sustained deviations from a baseline rate, identified an unusual increase in detections and delimited the outbreak period as December 2023 through May 2024. Outside that window, the detection rate was 1.17 patients per month; during the outbreak it rose to 3.33 patients per month, a difference the authors report as statistically significant by Poisson regression with p < 0.001. Confirmed infections climbed in parallel, from 0.30 per month across the remaining 54 months of observation to 1.67 per month during the outbreak, also significant at p < 0.001.

The severity profile shifted as well as the volume. During the outbreak period, ten of the twenty detections, or 50 percent, were classified as confirmed infections, compared with 16 of 63 detections, or 25.4 percent, outside the outbreak period, yielding an odds ratio of 2.94 with p-values between 0.032 and 0.039 depending on how ambiguous cases were classified. Infection proportions before and after the outbreak were similar, at 26.2 percent and 23.8 percent respectively, suggesting the cluster was a discrete episode rather than a gradual drift toward more invasive disease. Case definitions were strict: colonization meant detection in a surveillance sample without clinical evidence of infection, possible infection meant antimicrobial treatment without an established causal link, and confirmed infection required isolation of the organism from a clinically infected site such as a wound, an inflamed eye, or blood culture with compatible findings.

Molecular typing told the decisive part of the story. Because penicillin-susceptible S. aureus isolates had not been routinely archived before the investigation, only four isolates from three patients sampled during the central outbreak period were available for detailed characterization. Two came from the umbilical wound of one patient, whose colonies showed different macroscopic appearances on the primary culture plate, and the others originated from conjunctival samples of two further patients. All four isolates shared identical spa type t328, sequence type ST15, and core-genome multi-locus sequence type 38152, and they formed a single cgMLST cluster with allelic distances ranging from zero to seven alleles, strong evidence of transmission among these cases. Every isolate carried the eta gene, which encodes exfoliative toxin A, and all lacked mecA and blaZ, confirming both methicillin susceptibility and the absence of the classic penicillinase gene.

Exfoliative toxins are serine proteases that cleave desmoglein-1, a protein that anchors skin cells together, and can cause localized bullous impetigo or, when toxin effects become systemic, staphylococcal scalded skin syndrome. Neonates, particularly preterm infants, are especially vulnerable to disease caused by toxin-producing strains because antibody production against these toxins increases with gestational age. Consistent with that pattern, the median gestational age in this outbreak was significantly lower among patients with confirmed infection than among those with colonization only: 214.5 days, equivalent to 30.6 weeks, versus 243.5 days, or 34.8 weeks, a difference the authors report as significant by the Mann-Whitney U test with p = 0.029. The two patients classified as possible infections had gestational ages of 181 and 193 days, but the group was too small for inferential comparison, and the presence of a term infant among the infected patients precludes firm conclusions about a causal role for prematurity.

Two isolates obtained later told a different story and sharpened the interpretation. An isolate from a patient sampled on May 27, 2024, at the end of the epidemiologically defined outbreak period, belonged to spa type t049, ST45, and cgMLST type 38914, while an isolate from June 3, 2024, fell into spa type t084, ST582, and cgMLST type 38917. Neither carried eta, and both differed from the ST15 cluster by more than 300 alleles, ruling them out as members of the characterized outbreak clone. This means the epidemiologically defined outbreak window may overstate the true duration of clonal transmission, since sporadic colonizations and infections by unrelated strains could have occurred alongside the ST15 spread. The authors note that this limitation highlights the value of prospective molecular surveillance, since routine storage or rapid genomic characterization of susceptible isolates could enable earlier recognition of transmission events when individual cases are separated by weeks or when clinical manifestations remain mild.

The infection-control response combined environmental sampling, enhanced hygiene measures, and an unusual approach to staff screening. Environmental investigations, including ten patient-environment samples, five reprocessed incubators, samples from touch surfaces, plaster strips from an affected patient, and the unit’s routine twice-yearly contact surface sampling plus annual testing of five crocheted toys reprocessed externally, failed to identify any environmental reservoir of S. aureus. Because data-protection requirements restricted systematic identifiable staff screening, all NICU staff were offered voluntary anonymous nasal screening between June 20 and July 28, 2024, with results accessible only to each staff member through an individual code. Of 57 screening samples analyzed, 43 were negative, seven yielded penicillin-susceptible S. aureus, and seven yielded penicillin-resistant S. aureus. Staff with positive results could anonymously obtain mupirocin-containing nasal ointment for decolonization, and repeat testing was possible. Five of the seven penicillin-susceptible detections occurred within the first six days of screening, one on day ten, and one on day twenty-one, with no further detections during the final two and a half weeks.

The investigators also checked whether changes in sampling intensity could explain the cluster. Monthly detections of penicillin-susceptible S. aureus showed little correlation with detections of penicillin-resistant S. aureus, with a Pearson correlation coefficient of 0.14, and only weak correlations with monthly counts of swabs, blood cultures, and catheter-tip cultures, at 0.30, 0.12, and 0.21 respectively. Screening frequency and sampling sites remained unchanged throughout the observation period, so the rise does not appear to reflect intensified testing, although the authors acknowledge that residual confounding cannot be excluded. A further constraint is that patient-days were unavailable for the complete observation period because of a hospital cyberattack, meaning incidence rates could not be standardized to patient exposure time.

Several features of the outbreak point toward nosocomial transmission rather than introduction from outside. All of the children suffering from conjunctivitis were born by caesarean section, which makes mother-to-child transmission during vaginal delivery unlikely. Staphylococci are recognized causes of neonatal conjunctivitis, and in this cohort the infections remained localized and responded well to topical antibiotic treatment, though S. aureus is also an important cause of severe and systemic infections in preterm infants. The outbreak clone belonged to ST15, a globally distributed lineage that occurs as both methicillin-susceptible and methicillin-resistant variants and shows considerable diversity in resistance and virulence profiles; the eta gene has previously been described in individual ST15 isolates. Previous NICU outbreaks caused by typed MSSA strains have involved sequence types ST21, ST84, and ST121, and recent genomic surveillance studies suggest that MSSA transmission clusters in neonatal settings may occur more frequently than previously recognized. Because staff screening was anonymous and staff isolates were never molecularly compared with patient isolates, the contribution of healthcare workers to transmission could not be determined, and the authors caution that the observational design cannot separate the effects of individual interventions. Nevertheless, no further outbreak-associated detections occurred after the bundle of enhanced measures was implemented, and the temporal association suggests the combined response likely contributed to containment. The central lesson, the team concludes, is that antibiotic-susceptible pathogens can drive clinically relevant outbreaks in neonatal intensive care, and surveillance that also includes susceptible organisms may detect suspicious clusters that resistance-focused screening would miss.

Subject of Research: An outbreak of penicillin-susceptible, eta-positive Staphylococcus aureus ST15 in a German neonatal intensive care unit

Article Title: Outbreak of penicillin-susceptible, eta-positive Staphylococcus aureus in a neonatal intensive care unit

Article References: Meintrup, D., Vilser, D., Layer-Nicolaou, F., Müller, F., Strommenger, B., Rieß, B., Hücherig, B., & Borgmann, S. (2026). Outbreak of penicillin-susceptible, eta-positive Staphylococcus aureus in a neonatal intensive care unit. Frontiers in Microbiology, 17, 1946607. https://doi.org/10.3389/fmicb.2026.1946607

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1946607

Keywords: Staphylococcus aureus, neonatal intensive care unit, outbreak, penicillin-susceptible, exfoliative toxin A, ST15, cgMLST, purulent conjunctivitis, infection control, preterm infants, surveillance, mupirocin decolonization