Current NICU practice and interventions adapted from adult ICUs

Current practices for preventing invasive disease due to MRSA in the NICU include a “seek and destroy” infection control program, whereby periodic surveillance cultures are obtained and colonized infants are decolonized and/or isolated. ^{Reference Popoola, Colantuoni and Suwantarat8} Decolonization procedures are unit based, gestational, and chronological age specific. A survey of UK NICUs demonstrated that 73.8% of MRSA-colonized neonates underwent routine decolonization. ^{Reference Francis, Khan and Kennea9} Furthermore, the choice of decolonization regimen varied, with more than 15 different regimens in use. Subsequent source isolation after successful decolonization also varied among the units. ^{Reference Francis, Khan and Kennea9}

Universal decolonization in adult intensive care units (ICUs) has been found to be more effective than targeted decolonization or screening and isolation in reducing rates of MRSA clinical isolates and bloodstream infection from any pathogen. Huang et al used a pragmatic, cluster-randomized trial to compare 3 strategies for reducing MRSA infections in 74 adult ICUs, across 43 American hospitals. ^{Reference Huang, Septimus and Kleinman10} The strategies included MRSA screening and isolation and targeted decolonization (screening, isolation, and decolonization of MRSA carriers) and universal decolonization (no screening, but decolonization of all patients). The trial demonstrated universal decolonization reduced MRSA-positive clinical cultures by 37% and bloodstream infections from any pathogen by 44%, and adverse events were mild and related to chlorhexidine. ^{Reference Huang, Septimus and Kleinman10} It remains to be seen whether universal decolonization can obviate the need for all contact precautions for carriers of MRSA or other multidrug-resistant organisms. ^{Reference Huang, Septimus and Kleinman10}

Challenges with interventions in neonates

To reduce the morbidity and mortality associated with invasive MRSA infection, screening of infants in neonatal care facilities has become common practice. However, body site and frequency of screening are not universally agreed upon, and as previously mentioned, the prevalence of MRSA in each locality may necessitate different practices. ^{Reference Gregory, Eichenwald and Puopolo11–Reference Akinboyo, Zangwill, Berg, Cantey, Huizinga and Milstone13} The US Centers for Disease Control and Prevention (CDC) S. aureus NICU guideline, published in 2021, stipulates as a minimum, the anterior nares should be sampled in NICU patients, based on moderate evidence from two diagnostic studies. ^{Reference Milstone, Elward, Brady and Cox14} Recently updated National Institute of Clinical Excellence (NICE) guidelines on the management of MRSA have removed a recommendation specifically listing umbilicus and throat as preferential screening sites in neonates, citing a lack of evidence. ^{Reference Coia, Wilson and Bak15}

Additionally, targeted MRSA decolonization strategies have limitations. First, despite weekly surveillance cultures, 42% of infections had no preceding positive MRSA screening swab, providing no opportunity to attempt decolonization. ^{Reference Popoola, Colantuoni and Suwantarat8} Second, the interval between colonization and infection in many neonates was short (median 5 days), suggesting a narrow window of opportunity for decolonization. Third, 38% of neonates who received decolonization treatment became recolonized during their NICU stay, and 16% developed an MRSA infection, so the efficacy of decolonization to eradicate MRSA colonization and prevent MRSA infections may be limited. ^{Reference Andersen, Lindemann and Bergh16,Reference Back, Linnemann, Staneck and Kotagal17} Some authors have described the universal treatment of all neonates with mupirocin to successfully reduce MRSA infections in neonates. ^{Reference Ericson, Popoola and Smith18} However, a universal approach in some settings has led to the emergence of mupirocin resistance, and the long-term impact of indiscriminately treating all neonates, including those not colonized with S. aureus, is unknown ^{Reference Carey, Duchon, Della-Latta and Saiman19} and may negatively affect the neonatal microbiome. ^{Reference Carey20}

The optimal decolonization agent or combination of agents remains an unresolved issue. As reported in the product specifications, the safety and efficacy of intranasal mupirocin are not established in patients aged less than 12 years. Additionally, in neonates and premature infants, systemic absorption occurs following intranasal administration, but it remains uncertain whether this absorption causes adverse health consequences in neonates. The application of a nasal ointment can be technically challenging in very low birth weight infants. Furthermore, certain agents, such as topical chlorhexidine, can cause adverse skin reactions, including chemical burns, especially in premature infants. ^{Reference Paternoster, Niola and Graziano21} A prospective randomized trial is needed to formally evaluate the efficacy and safety of mupirocin with or without chlorhexidine for eradicating MRSA colonization and preventing MRSA infections in neonates. This could be expanded to include primary caregivers and family.

Duffy et al examined 21,736 surveillance specimens from 3,784 admissions to a tertiary neonatal unit over an 8-year period. ^{Reference Duffy, Garbash, Sharland and Kennea22} They demonstrated infants colonized with MRSA were smaller and of lower gestational age. Furthermore, infants initially positive on groin swabs tended to have greater gestational age and weight at birth compared with those initially positive on nose swabs or both nose and groin swabs. ^{Reference Duffy, Garbash, Sharland and Kennea22} Duffy et al concluded weekly nasal swabs and swabbing of suspected infected sites could detect 85% of colonized infants, while groin swabs detected only a small number of larger, mature babies who were discharged before 2 weeks of age. ^{Reference Duffy, Garbash, Sharland and Kennea22}

Description of the Leeds neonatal service

The Leeds Neonatal Service admits approximately 1,600 neonates annually. The service operates over 2 sites: the first functions as a Level 1 special care baby unit (20 cots) and the second as a Level 3 neonatal intensive care unit with surgery (34 cots). The latter provides care for both inborn and outborn extremely preterm infants and neonates requiring specialized surgical and cardiac care. Given this profile, these neonates are at an increased risk of developing S. aureus infections. Admission swabs from both outborn and inborn neonates have similar MRSA rates, as did the 2 neonatal units, and therefore for this study, the data from both units were analyzed together.

Nationally, S. aureus bacteremia exhibited a remarkable surge in MRSA strains during the 1990s, followed by a subsequent stabilization and decline between 2000 and 2007. ^{Reference Johnson, Pearson and Duckworth23} Although MRSA infections are predominantly observed in older patients, data from 1990 to 2001 highlighted an increase in MRSA bacteremia among children, with the number of cases rising from less than 5 (constituting <1% of all S. aureus cases) in 1990 to over 70 in 2000 and 2001 (representing >10% of all S. aureus cases). ^{Reference Khairulddin, Bishop, Lamagni, Sharland and Duckworth24} It is suspected that neonates accounted for most of these cases, although the granularity of data is not captured.

In 2008–2009, the Leeds neonatal service recorded the highest number of MRSA bacteremia cases in the United Kingdom, with 8 instances, surpassing the total of 7 cases reported by 15 other neonatal units combined. The baseline rate for all hospital-onset MRSA for 2008–2009 in England was reported as 3.3 cases per 100,000 acute trust days [Mandatory surveillance data]. Leeds Teaching Hospitals Trust (LTHT) had a rate of 9.8 for the same time frame. However, it is important to highlight that LTHT is one of the largest acute trusts in England with an estimated 653,228 bed days compared with the average of 325,950, for the same period. Due to the associated morbidity of MRSA bacteremia, the neonatal service undertook a 4-week assessment period and systematic review of cases. The assessment included environmental audits, a revision of the junior medical staff induction procedures, ward housekeeping/cleaning inspections, daily hand hygiene audits, and an evaluation of antimicrobial prescription practices.

Interventions undertaken in the Leeds neonatal service

Figure 1 demonstrates the rising MRSA isolation from patients within the Leeds neonatal service. A root cause analysis (RCA) of bacteremia cases demonstrated that patients with Broviac or peripherally inserted central catheter lines, current MRSA colonization, previous abdominal surgery, and discharging skin lesions were at higher risk of MRSA bacteremia. Following RCA analysis, 3 domains were targeted for focused intervention including healthcare staff, the environment, and the patient. The interventions are detailed in Figure 2 and summarized in Table 1.

Figure 1. Methicillin-resistant Staphylococcus aureus (MRSA) isolation rates per 1,000 admissions in the Leeds neonatal service, from 2008 until 2023. Infection rate refers to skin and soft tissue infections only; bacteremia rate refers to the isolation of MRSA from blood.

Figure 2. The Leeds neonatal service methicillin-resistant Staphylococcus aureus (MRSA) time line. Actions undertaken associated with each case are described. Each circle represents a month and year.

Table 1. A summary of key interventions to target methicillin-resistant Staphylococcus aureus (MRSA) bacteremia. The table summarizes the interventions undertaken, divided broadly into three categories

Note. PPE, personal protective equipment.

Addressing understaffing was thought to improve adherence to infection control standards; therefore, recruitment was conducted to align the nurse-to-patient ratio with the British Association of Perinatal Medicine standards (BAPM). ^{Reference Howard25} A system was implemented to achieve universal compliance with aseptic training and e-learning. The IPC strategy also included staff engagement activities and nursing empowerment, focusing on education to enable staff to challenge poor practices. Voluntary staff MRSA screening and decolonization was undertaken, offering onetime screening and decolonization for positives, with no known failures in staff decolonization.

Additionally, a weekly interdisciplinary infection control meeting was initiated to streamline communication and improve practices. Considering the environmental aspect, ward modifications included a reassessment of cot spacing, de-cluttering, and organization of the unit for enhanced functionality, improved access to personal protective equipment (PPE), and a comprehensive audit of ward cleaning practices.

A series of patient-focused infection control interventions were introduced, including weekly MRSA screening (axilla, groin, umbilicus, and nose), daily antiseptic skin washes (Octenisan) of neonates above 28 weeks’ gestation, the use of 2% chlorhexidine for skin asepsis before invasive procedures, and a change of dressings used for Broviac lines (see Figure 2). This was a significant change in neonatal practice; therefore walk-arounds on the unit and staff education sessions were arranged to understand the perceived barriers before the introduction of these IPC changes.

Taken together, these measures resulted in a significant reduction in MRSA infections (see Figure 1). The implementation of weekly screening initially resulted in a spike in colonization rates. However, through the combination of the above interventions, MRSA bacteremia have been eradicated within the Leeds neonatal service, with no cases of MRSA bacteremia since 2009. While we propose that our IPC interventions led to the reduction in neonatal MRSA infection and colonization rates, we recognize the potential impact of extrinsic factors, such as pathogen variation, wider antibiotic use, community transmission, environmental conditions, and improved neonatal nutrition may affect these outcomes.

MSSA bacteremia in the Leeds neonatal service

Despite MRSA control, MSSA remained a challenge in the Leeds neonatal service. Analysis of MSSA bacteremia data demonstrated a significant portion of these cases occurred in babies born at less than 28 weeks’ gestation. Interestingly, Octenisan washes for MRSA prevention targeted babies of 28 weeks’ gestation or older, as younger babies were excluded due to concerns regarding skin integrity and thermostability.

In 2014, a screening program for MSSA colonization in babies less than 28 weeks’ gestation was implemented. This included admission and weekly screening swabs until 28 weeks postconceptual age. At this time, the standard daily decolonization with Octenisan began. Upon a positive MSSA screening swab, the baby received decolonization. This intervention significantly reduced MSSA in this population as shown in Figure 3.

Figure 3. Number of methicillin-sensitive Staphylococcus aureus (MSSA) bacteremia and MSSA/1,000 central line days. Arrows signify the year in which interventions were implemented and extended.

Despite this initial screening intervention, MSSA bacteremia persisted in infants born before 28 weeks but occurred after they reached 28 weeks postconceptual age. To address this, in 2019, weekly screening was expanded to all such infants for their entire admission, leading to a further reduction of MSSA bacteremia rates (Figure 3). Additional enhancements were also introduced in 2019, including increased scrutiny of aseptic practices, the implementation of Matching Michigan checklists for line insertions, an individual patient infection prevention booklet mandating daily line dressing assessments, with changes only necessitated for soiled or lifting dressings, and revised humidity and skin care protocol.

Staffing levels and their impact on MSSA rates

Understaffing results in decreased compliance with IPC measures. ^{Reference Clements, Halton and Graves26} For neonatal care, nursing staff levels are based on care requirements and described by BAPM; ^{Reference Howard25} high dependency care requires at least a 1:3 ratio, and intensive care necessitates a minimum 1:1 ratio. A retrospective analysis of the Leeds neonatal service from March 2008 to December 2011 showed a notable negative correlation between staffing levels and monthly S. aureus rates, with diminished staffing associated with higher S. aureus culture rates (Spearman’s correlation coefficient = −0.316; P = .03, unpublished data). Those months with lower staffing compliance (<80%) were associated with higher rates of S. aureus isolation. This implies that reduced staffing might increase the prevalence of S. aureus in the NICU, even with other IPC measures in place. The exact reasons behind this association may be complex and influenced by various factors such as length of stay and high-risk interventions. Nonetheless, the findings underscore the potential impact of staffing levels on infection rates in the NICU setting.