Methicillin-resistant Staphylococcus aureus (MRSA) outbreaks are frequent in neonatal intensive care units (ICUs).Reference Nelson and Gallagher ¹ Toxic-shock-syndrome toxin 1 (TSST-1)–producing MRSA Geraldine clone represented 6.3% of invasive MRSA isolates in France in 2006 and 2007,Reference Dauwalder, Lina and Durand ² and has been implicated in one outbreak among newborns.Reference Leroyer, Lehours and Tristan ³ We describe here a neonatal MRSA Geraldine clone outbreak.

On March 31, 2014, a TSST-1–positive MRSA was isolated in bronchial aspirates from 2 ICU neonates (case patients 3 and 4) (Figure 1). Our subsequent investigation identified 2 prior cases of TSST-1 MRSA carriage, the index case 1 by umbilical swab in November 2013 and case 2 by bronchial aspirate in December 2013. A case was defined as a positive culture for an MRSA strain expressing TSST-1 and/or specific antibiotic susceptibility in a patient hospitalized in the neonatal ICU or general neonatal ward. In total, we identified 8 cases (7 cases of carriage and 1 skin infection) over a 9-month period (Figure 1). All case patients were premature (26–30 weeks gestation; mean birth weight, 975.2 g) and were hospitalized in the neonatal ICU. Among them, the mean interval of MRSA carriage detection was 25.1 hospitalization days, and mean length of ICU stay was 33.1 days. During the outbreak, case surveillance consisted of weekly nasal S. aureus carriage screening of the neonates of both wards; this procedure remained in place for 5 months after the last case was discovered. All MRSA isolates expressed resistance to penicillin G, methicillin, kanamycin, tobramycin, and fucidic acid according to guidelines of the French Antibiogram Committee. All of the isolates were typed by the National Reference Center for staphylococci (S. aureus Genotyping Identibac, Alere, Waltham, MA) and were identified as the Geraldine clone, which is characterized by the following criteria: (1) sequence type ST5, agr2, (2) positivity for TSST-1, enterotoxins SEC, SED, SEJ, SEL, and SER as well as the egc locus, and (3) negativity for Panton-Valentine leukocidin.Reference Monecke, Slickers and Ehricht ⁴ All case isolates underwent molecular analysis except strains from cases 1 and 2, as these strains had not been stored.

Figure 1

Synoptic table of the cases and wards. Each square represents a week: red for the neonatal ICU, dark blue for the general neonatal ward, and light blue for the mother–neonate unit to which neonates are transferred when their condition allows, usually close to hospital discharge. The numbers in the squares represent the room number, and stars represent isolations of the outbreak clone.

Immediately after the alert, we implemented contact precautions (ie, glove and gown usage) for HCP in contact with infected and colonized neonates. We also held information meetings for healthcare personnel (HCP) and audited HCP practices. The audit revealed a lack of consistency in standard precaution application and hygiene practices. The control measures implemented consisted of team support for multidrug-resistant bacteria management, standard precautions, and hand-hygiene reinforcement. We focused on the use of hydroalcoholic solutions, lack of hand jewelry verifications, and daily changes of work outfits. We assessed the effectiveness of these measures using indicators such as bedsore prevalence, cleaning activities records, environmental samples, and compliance with hand hygiene procedures, which was assessed by hydroalcoholic solutions consumption according to French guidelines. ⁵ Compliance to the minimum hydroalcoholic consumption, calculated according to clinical activity, increased from 57.4% 6 months before the outbreak to 84.4% during the outbreak to 102.9% 6 months after the outbreak.

We sought environmental links between cases. In total, 60 environmental swabs and 20 surface samples from patient rooms, drug preparation area, transfrontanellar ultrasound apparati, and x-ray devices were tested between May 3 and June 25, 2014. No medical devices or environmental sources were found to be involved in transmission.

Despite the control measures, transmission continued. Some carrier neonates were hospitalized in neighboring rooms (Figure 1), suggesting possible cross transmission via HCP hands, especially because HCP compliance to the measures was not consistent at the beginning of the outbreak. In addition, S. aureus may have been spread by airborne transmission by HCP.Reference Halcomb, Griffiths and Fernandez ⁶ The long interval between the first 2 and subsequent 6 cases also pointed to HCP carriage. HCP are often involved in horizontal MRSA transmission to neonates,Reference Halcomb, Griffiths and Fernandez ^{6 ,} Reference Scheithauer, Trepels-Kottek and Häfner ⁸ and HCP decolonization is a proven outbreak control measure.Reference Leroyer, Lehours and Tristan ^{3 ,} Reference Heinrich, Mueller, Bartmann, Simon, Bierbaum and Engelhart ^{7 ,} Reference Scheithauer, Trepels-Kottek and Häfner ⁸ We opted for universal decolonization of all HCP, both permanent and rotating staff (including students, radiology technicians, radiologist physiotherapist, psychologist, milk-bank technicians, cleaning staff, social workers, laboratory couriers, and secretaries), regardless of their screening results, in order to cover the risk of false negatives due to intermittent carriage. We sampled both the noses and throats of the HCWs to improve sensitivity.Reference Sollid, Furberg, Hanssen and Johannessen ⁹ Decolonization consisted of a 5-day course based on twice-daily mupirocin nasal ointment and daily showering with chlorhexidine soap,Reference Abad, Pulia and Safdar ¹⁰ which were dispensed to each HCP during screening interviews to promote adherence. HCP voluntarily participated in decolonization; no one refused. No adverse event related to decolonization was recorded. Our strategy was aggressive because the same bundle of standard control measures plus non-exhaustive HCP decolonization failed to control an MRSA outbreak in another neonatal ICU.Reference Lepelletier, Corvec, Caillon, Reynaud, Rozé and Gras-Leguen ¹¹ The first round of screening involved all 155 HCP working with the neonates during July and August 2014. Overall, 61 HCP were S. aureus carriers (39.3%); among them, 2 (1.3%) carried the outbreak strain. Both were nurses, 1 nurse worked in neonatal ICU and the other one in general neonatal ward. They had been in contact with the case patients since the outbreak started. We were not able to determine whether the transmission originated with an HCP or a patient. These results confirmed our working hypothesis and justified the decolonization strategy. We checked screening efficacy with a second round of screening 1 month later on 30 HCP including the 2 identified carriers, both of whom tested negative. In the subsequent screening round, global S. aureus carriage decreased from 53.3% to 20%. The overall cost of the outbreak was US$18,821 (17,600 euros), which included consumption of protections for contact precautions for HCP and neonates, follow-up screening, and decolonization treatments.

In conclusion, we have described an outbreak of the MRSA Geraldine clone in a neonatal department, which was finally controlled by screening and decolonizing all HCP. The screening results suggested circulation of the outbreak clone between HCP and patients, which explains the persistence of new cases despite classical control measures. Our findings support universal HCP decolonization during neonatal outbreaks of MRSA TSST-1.