Introduction
Group B Streptococcus (GBS, Streptococcus agalactiae) is the leading cause of sepsis during the neonatal period. 1 Neonates and preterm infants in neonatal intensive care units (NICUs) face an increased risk of healthcare-associated infections due to their immature immune systems, the use of indwelling medical devices for monitoring and treatment, and prolonged hospital stay.Reference Glaser, Hughes and Jnah 2
GBS can cause early-onset GBS (EOGBS) infection whose onset within the first six days of life is presumptively a result of GBS acquisition during delivery. Late-onset GBS (LOGBS), in contrast, occurs between 7and 90 days of life. 1 Transmission routes for LOGBS are poorly understood, and those cases are assumed to be largely sporadic.Reference Mukhopadhyay and Puopolo 3 It is presumed that LOGBS is acquired postnatally from the mother, non-maternal caregivers in the community or hospital, or environmental sources. Reference Collin, Lamb and Jauneikaite4,Reference Puopolo, Lynfield and Cummings5
Neonatal GBS is a reportable disease in Canada. The incidence of LOGBS in Alberta, Canada in neonates, increased slightly from 0.38 to 0.41 cases per 1,000 live births between 2014 and 2020. This rise may reflect advances in neonatal care, which have improved survival of increasingly preterm and medically fragile infants who remain at risk for late-onset disease. In contrast, the incidence of EOGBS decreased significantly from 0.2 to 0.07 cases per 1,000 live births, presumed to be related to the success of targeted intrapartum chemoprophylaxis programs in reducing EOGBS which was introduced in Canada in 1994.Reference Ma, Thompson and Corsiatto 6 A comparable pattern, characterized by a decline in EOGBS with stable LOGBS rates, has also been observed in Ontario, Canada. Reference Fakhraei, Fell and El-Chaâr7
Few healthcare-associated GBS outbreaks have been reported, mostly in NICU settings as LOGBS or a mix of EOGBS and LOGBS. Reference Collin, Lamb and Jauneikaite4,Reference Delettre, Billion and Guyonnet8 Since GBS is typically considered maternally acquired, identifying hospital-related transmission factors and confirming case-relatedness can be challenging. Genomic methods such as whole-genome sequencing (WGS) have proven effective in investigating and confirming transmission in GBS outbreaks. Reference Delettre, Billion and Guyonnet8–Reference Sansom, Logan and Green10
Here, we describe a series of clusters of healthcare-associated GBS infections in six infants in a NICU over seven months, involving three distinct GBS serotypes confirmed by WGS.
Methods
Patients and setting
This is a retrospective description of the investigation and management of a GBS outbreak in the NICU of Sinai Health in Toronto, Canada, between July 27, 2023, and February 28, 2024. Neonates and infants were cared for in a level 3 NICU with 57 private rooms, subdivided into three zones. This unit primarily admits newborns; cases reported here occurred during prolonged hospitalization rather than following readmission. Invasive GBS infection was defined as isolation of GBS from a normally sterile site in a neonate or infant during their NICU stay. Non-invasive GBS infection was defined as isolation of GBS from a non-sterile site in association with clinical disease. Colonization was defined as isolation of GBS from a non-sterile site in the absence of infection. A nosocomial case was defined as infection or colonization caused by an isolate that was epidemiologically and microbiologically linked to a NICU cluster identified during the outbreak and surveillance period.
Surveillance and clinical data
Enhanced surveillance included three consecutive weekly point-prevalence surveys (PPS), during which rectal and throat swabs were obtained from all neonates in the NICU. The aim was to assess the extent of GBS colonization in the NICU, understand possible transmission mechanisms, identify potential common factors among affected neonates, and detect colonized neonates potentially at risk for clinical infection. Clinical information (gestational age, chronological age, infectious syndrome, maternal GBS status, antimicrobial treatment, and outcomes) were extracted from medical records and outbreak documents.
Laboratory methods
Clinical samples from sterile and non-sterile sites were obtained as part of the investigation of clinical infections. Rectal and throat swabs were collected as part of the PPS during the outbreak investigation. Rectal and throat samples were inoculated into Carrot Broth for GBS and incubated aerobically at 35°C for 24 hours. If the broth turned orange, it was subcultured onto colistin-nalidixic acid agar and incubated for 24 hours at 35°C under aerobic conditions. GBS-like colonies were confirmed as GBS using streptococcal latex agglutination testing. Antibiotic susceptibility testing of GBS isolates from clinical sites was performed and interpreted according to the Clinical and Laboratory Standards Institute M100 guidelines. 11
Short-read Illumina WGS was conducted at Canada’s National Microbiology Laboratory (NML). Sequence quality was assessed using the FastQC tool (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and genomes were assembled using Shovill (https://github.com/tseemann/shovill#readme). Serotyping was performed using the WADE pipeline (https://github.com/phac-nml/wade). Core SNV phylogenetic analysis was performed using SNVPhyl with default parameters. Reference Petkau, Mabon and Sieffert12 A total of 2,395 sites were used in the phylogeny, representing 82.7% of the core genome. A recent serotype Ia isolate from Ontario was used as a mapping reference. The genomic data reported in this study have been deposited in the NCBI Sequence Read Archive as part of BioProject PRJNA1372154. WGS turnaround time was up to 14 days during the investigation, compared with 30 days for routine surveillance.
Outbreak management and IPAC protocols
Following the identification of the initial cluster, an outbreak management team was convened to oversee the investigation and review IPAC protocols. To contain the spread, corrective actions were implemented to address identified lapses in disinfection workflows and hygiene practices. Furthermore, IPAC personnel were integrated into NICU leadership rounds to ensure sustained oversight, accountability, and ongoing monitoring of corrective measures.
Ethics
The study was approved by the Research Ethics Board of Sinai Health.
Results
Initial outbreak and cluster identification
In August 2023, two cases of LOGBS bacteremia occurred (patients 2 and 3) within two weeks prompting an investigation (Table 1 and Figure 1). Patient 2 presented with bacteremia and meningitis; Patient 3 experienced two episodes of bacteremia three weeks apart. WGS of the two isolates was performed and a review of the prior six months identified five cases of invasive GBS disease in neonates. One of these occurred in July 2023 (Patient 1) had EOGBS bacteremia on day 1 of life, with serotype Ia genetically identical (0 SNVs) to the isolates from Patients 2 and 3 and was considered the primary case. The remaining earlier cases were genetically unrelated to Patient 1 and to each other. Two subsequent cases followed: Patient 4 (serotype V, unrelated) one week later and Patient 5 three weeks later, matching Cluster 1 (serotype Ia, 0 SNVs). Given the relatedness of the four LOGBS cases, a GBS outbreak was declared. In November 2023, Patient 6 developed GBS conjunctivitis (serotype V), matching Patient 4 with a 1-SNV difference, forming Cluster 2 (Figures 1 and 2). Details regarding the clinical data of unrelated cases are provided in the Supplementary Table.
Timeline of neonatal intensive care unit stay and group B streptococcus infection onset for patients 1–8.
Core genome single nucleotide variant analysis of group B streptococcus isolates from a multi-cluster outbreak in neonatal intensive care unit. A total of 2,395 sites were used in the phylogeny, including 82.7% of the core genome. A recent serotype Ia isolate from Ontario was used as a mapping reference.
Clinical manifestations of clustered group B streptococcal infections in NICU (July 2023 to February 2024)
Table 1. Long description
A table with 8 rows and 8 columns summarizing clinical data of patients with group B streptococcal infections in a NICU. The columns are Cluster number, Serotype, Gestational age at birth (weeks), Chronologic age, Maternal GBS status, Clinical presentation, Antimicrobial treatment, and Mortality. Row 1: Cluster number, 1; Serotype, Ia; Gestational age at birth (weeks), 24; Chronologic age, 1 days; Maternal GBS status, Positive; Clinical presentation, Bacteremia; Antimicrobial treatment, Penicillin G; Mortality, Survived. Row 2: Cluster number, 1; Serotype, Ia; Gestational age at birth (weeks), 29; Chronologic age, 8 days; Maternal GBS status, Negative; Clinical presentation, Bacteremia and meningitis; Antimicrobial treatment, Penicillin G; Mortality, Deceased due to fulminant infection. Row 3: Cluster number, 1; Serotype, Ia; Gestational age at birth (weeks), 23; Chronologic age, 1st infection: 17 days Relapse: 39 days; Maternal GBS status, Negative; Clinical presentation, 1st bacteremia, 2nd bacteremic pneumonia; Antimicrobial treatment, Penicillin G + Tobramycin; Mortality, Survived. Row 4: Cluster number, 2; Serotype, V; Gestational age at birth (weeks), 23; Chronologic age, 51 days; Maternal GBS status, Unknown; Clinical presentation, Bacteremia; Antimicrobial treatment, Penicillin G + Vancomycin; Mortality, Survived. Row 5: Cluster number, 1; Serotype, Ia; Gestational age at birth (weeks), 27; Chronologic age, 19 days; Maternal GBS status, Unknown; Clinical presentation, Bacteremia; Antimicrobial treatment, Ampicillin; Mortality, Survived. Row 6: Cluster number, 2; Serotype, V; Gestational age at birth (weeks), 30; Chronologic age, 123 days; Maternal GBS status, Negative; Clinical presentation, Conjunctivitis; Antimicrobial treatment, Topical treatment - bacitracin zinc and polymyxin B sulfate; Mortality, Survived. Row 7: Cluster number, 3; Serotype, IV; Gestational age at birth (weeks), 25; Chronologic age, 62 days; Maternal GBS status, Positive; Clinical presentation, Bacteremia; Antimicrobial treatment, Penicillin G; Mortality, Survived. Row 8: Cluster number, 3; Serotype, IV; Gestational age at birth (weeks), 25; Chronologic age, 54 days; Maternal GBS status, Unknown; Clinical presentation, Bacteriuria; Antimicrobial treatment, Ampicillin + Tobramycin; Mortality, Survived.
* One of twins, the other twin was negative for GBS.
§ Primary cases.
Point-prevalence survey results
Weekly PPS identified five additional colonized cases. However, WGS confirmed that none were related to the outbreak clusters (Figure 2).
IPAC investigation findings
IPAC audits identified several lapses: suboptimal hand hygiene compliance, non-adherence to the “bare below the elbows” policy and artificial nail restrictions, gaps in the cleaning and disinfection workflow and documentation of shared medical equipment requiring low-level disnfection (e.g., glucometers, ultrasound devices, and electrocardiogram machines), inconsistent discarding of patient-dedicated disposable supplies at discharge and deficiencies in environmental cleaning. Additionally, variable adherence to aseptic technique was observed, including preparation of oral medication at the bedside. No deficiencies were found in breast milk or formula preparation; however, patient-dedicated pump cleaning protocols were reinforced as a precaution. No single healthcare worker was linked to any of the clusters; however, shared nursing assignments occurred within NICU zones (Figure 3).
Spatiotemporal mapping of NICU bed allocation and transfers among group B streptococcus-affected infants.
A structured, multi-layered response was implemented. Infection prevention measures were reinforced through unit-based multidisciplinary education focusing on hand hygiene, “bare below the elbows” adherence, artificial nail restrictions, aseptic medication preparation, and standardized shared equipment disinfection workflows. Compliance monitoring was intensified through observational audits with real-time feedback and weekly reporting to NICU leadership. Environmental cleaning and post-discharge supply management processes were reviewed in collaboration with environmental services, with targeted re-education where needed. IPAC personnel joined NICU leadership rounds to enhance ongoing oversight, accountability, and barrier mitigation. Additional Precautions were deemed unnecessary due to the unit’s private room structure.
Late cluster and resolution
Eight weeks after the last affected infant was discharged from the NICU, no further cases of related GBS infection were detected. IPAC practices were implemented and audited, and the outbreak was declared over on February 5, 2024. Two weeks later, four additional cases of LOGBS were identified. None of these isolates were related to previous cases, but two of these were related to each other. Patient 7 (bacteremia) matched their mother’s antepartum and postdelivery isolates. Horizontal transmission was confirmed from Patient 7 to Patient 8 (asymptomatic bacteriuria), forming Cluster 3 (0–1 SNVs). An outbreak was not declared, as the investigation identified only a single transmission event resulting in a non-invasive GBS infection, with no evidence of sustained transmission or additional invasive disease beyond Patient 7. Weekly PPS were not introduced, as they did not provide additional epidemiologic or operational value in guiding decision-making in previous clusters. Surveillance for clinical cases of GBS infection continued. No additional cases of LOGBS infections were detected as of October 2025.
All infants with invasive GBS infection received intravenous beta-lactam therapy (e.g., penicillin or ampicillin ± gentamicin) in accordance with institutional guidelines. Duration of therapy ranged from 10 to 21 days depending on site of infection. Patient 3 developed recurrent infection 3 weeks later. Patient 2 developed meningitis and subsequently died due to complications of invasive GBS disease. The other infants in the cohort survived (Table 1).
Foundational IPAC revision
Following audits that identified informal transfer of pump kits between mothers, contrary to established infection prevention protocols, IPAC practices were strictly reinforced. This unauthorized transfer of kits from discharged to newly admitted mothers was immediately discontinued, and education regarding single-user designations was provided. Additionally, protocols for hand hygiene, terminal cleaning, and disposal of supplies were reviewed to ensure long-term compliance.
Discussion
We describe a complex, extended series of GBS transmission events in a NICU setting, involving 5 definite nosocomial cases of LOGBS infection from 3 distinct clusters (serotypes Ia, V, and IV) confirmed by WGS over seven months. Additional cases identified through enhanced surveillance and weekly point prevalence surveys were genetically unrelated to the outbreak clusters.
This investigation highlights the complexity of managing a GBS outbreak in a level III NICU. Our findings underscore three critical themes in modern neonatal IPAC: recognition of horizontal NICU transmission in LOGBS, the value of WGS in understanding LOGBS sources, and the added vulnerability stemming from increased informal caregiver involvement.
WGS revealed a series of concurrent, distinct clusters rather than a single continuous outbreak. It identifies Cluster 1 primary case and distinguished between horizontal transmission and independent introductions, a feat impossible with serotyping alone. WGS also prevented the conflation of Clusters 1 and 2 and the misclassification of five unrelated colonized cases identified during PPS. This genomic clarity directly informed our response, avoiding the unnecessary escalation of infection control precautions for infants with sporadic colonization. Despite ongoing challenges with accessibility and real-time WGS results, Canada’s routine WGS of invasive GBS isolates provided essential data for this investigation. WGS is becoming a standard component of outbreak analysis, and its expanding availability will further strengthen future surveillance efforts. Reference Delettre, Billion and Guyonnet8,Reference Sansom, Logan and Green10
Reflecting broader trends, the clusters comprised commonly circulating GBS serotypes Ia, V, and IV. Across North America, serotypes V and Ia, followed by III, drive most adult invasive cases, while III and Ia are the most frequent in infant EOGBS and LOGBS. Reference Navarro-Torné, Curcio, Moïsi and Jodar13,Reference Bianchi-Jassir, Paul and To14 These findings align with Canadian data: serotypes III and V historically predominated in the Greater Toronto Area among infants and adults,Reference Teatero, McGeer and Low 15 while Ia and V are highly prevalent in Alberta. Reference Davies, Raj and Adair16,Reference Alhhazmi, Hurteau and Tyrrell17
Lapses in foundational IPAC practices triggered a comprehensive response focused on multidisciplinary education and real-time compliance monitoring. Key interventions targeted hand hygiene, aseptic techniques, and standardized disinfection for shared medical and breast milk equipment. By integrating IPAC personnel into NICU leadership rounds, we established a framework for sustained accountability. In a systematic review of healthcare-associated GBS clusters, published in 2019, unsatisfactory infection prevention practices were reported in 8 out of 17 studies, with inadequate disinfection of equipment and surfaces, such as shared breast milk equipment and formula preparation, reported in 6 studies and crowding and high patient-to-nurse ratio in 5 studies.Reference Collin, Lamb and Jauneikaite 4
This outbreak was the second-longest genomically confirmed GBS outbreak reported in the literature.Reference Jauneikaite, Kapatai and Davies 9 GBS outbreaks with horizontal transmission may exhibit intervals of several weeks between consecutive cases. Reference Delettre, Billion and Guyonnet8,Reference Jauneikaite, Kapatai and Davies9 These prolonged intervals likely reflect either GBS environmental contamination Reference Åberg, Ottosson and Granlund18 or extended periods of colonization before the onset of invasive disease. Because the cornerstone for identifying most nosocomial outbreaks in the NICU setting has been the occurrence of at least two invasive GBS infections within a short time frame, these prolonged intervals may have reduced the detection of transmission in the NICU. Reference Jauneikaite, Kapatai and Davies9,Reference Sansom, Logan and Green10,Reference Petkau, Mabon and Sieffert12,Reference Åberg, Ottosson and Granlund18,Reference Berardi, Guidotti and Creti19 The high rectovaginal GBS colonization rate in mothers of infants with LOGBS may also interfere with transmission assessment if WGS is not available. Reference Verani, McGee and Schrag20,Reference Boyer, Vogel and Gotoff21
The anticipated high GBS carriage rate among staff,Reference Boyer, Vogel and Gotoff 21 coupled with the complex multi-cluster nature of the outbreak and unclear transmission directionality, led to the decision to forgo staff screening. Furthermore, as GBS can be harbored within the NICU environment,Reference Åberg, Ottosson and Granlund 18 we prioritized reinforcing clinical workflows and IPAC practices compliance over environmental sampling.
Additional Precautions were not initiated during the outbreak, as accommodation of our NICU is entirely private rooms. For units with an open layout, the need for Contact Precautions should be considered based on factors such as bed proximity, the acuity level of the unit, and workload. Proximity of cots was reported in 7 out of 17 GBS hospital outbreaks. Reference Collin, Lamb and Jauneikaite4 Cohorting of infected or colonized patients should be considered in open-space NICUs.Reference Verani, McGee and Schrag 20 Nonetheless, horizontal transmission occurred in our private room-only configuration, likely mediated by shared equipment or healthcare worker contact, highlighting that physical barriers cannot completely substitute for adherence to IPAC practices. Preventing GBS spread requires a continuous focus on foundational IPAC measures, including hand hygiene and disinfection of shared clinical and feeding equipment, to mitigate the risks posed by healthcare worker- and equipment-mediated cross-contamination.
Enhanced screening of infants in the NICU during a GBS outbreak remains a topic of debate. In our outbreak, enhanced weekly screening detected five cases unrelated to the outbreak. The detection of GBS did not alter our outbreak management and could potentially increase parental anxiety. Screening for GBS in open-space NICUs may facilitate cohorting of colonized or infected infants, thereby reducing exposure of non-colonized patients. However, cohorting may also pose risks if infants colonized with less-virulent strains are placed alongside those carrying outbreak strains, potentially enabling transmission of more virulent GBS to otherwise stable infants.
Knowledge of infants’ GBS status is unlikely to affect thresholds for sepsis evaluation or empiric antibiotic choice, as GBS remains susceptible to standard first-line empiric therapy. However, this may change if high-level gentamicin resistance becomes more prevalent. Reference Creti, Imperi and Gherardi22
In conclusion, identifying GBS transmission in NICUs is challenging and requires a high index of suspicion. The interval between identified cases can be long, and outbreaks may span months. WGS is essential for confirming outbreaks, excluding maternal sources, and identifying epidemiological links between cases. Screening of staff and patients should be considered on a case-by-case basis. Foundational IPAC practices are often the most effective means of containing outbreaks.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/ice.2026.10503.
Financial support
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests
The authors declare that they have no competing interests or personal relationships that could have appeared to influence the work reported in this paper.
