MRSA colonisation rates on a neonatal and paediatric intensive care unit

Abstract

Methicillin-resistant Staphylococcus aureus (MRSA) colonisation is a challenge in healthcare institutions worldwide. In this retrospective nation-wide study, the rates of MRSA colonisation and infection from 2012 to 2015, on the only neonatal and paediatric intensive care unit (NPICU) in the country, were determined. Mean local rates were compared to rates of MRSA colonisation reported in units in North America, Asia and Europe between 2001 and 2010. The average rate of MRSA colonisation on admission to NPICU from 2012 to 2015 was 3.71% (95% confidence interval [CI] 2.17–5.25), while the mean rate of acquired colonisation was 14.60% (95% CI 6.16–23.04). Both were significantly higher than in units abroad: 1.9% and 4.1%, respectively (P = 0.04 and P < 0.001). There were no cases of invasive MRSA infection, while the mean rate of non-invasive infection was 0.77% (95% CI 0.54–1.01). Improved adherence to infection control measures and newer molecular diagnostic techniques are needed to further decrease the acquisition of MRSA.

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) colonisation has become a constant challenge in healthcare institutions worldwide. Colonisation may progress to infection (Huang et al., 2006), varying from skin infections to sepsis, leading to significant morbidity and increasing the cost burden of healthcare systems (Song et al., 2010). Children needing intensive care have a number of risk factors that predispose them to MRSA infection, such as an immature immune system, immature natural barriers, the presence of central lines, and treatment with broad spectrum antibiotics (Manzoni et al., 2013; Nelson and Gallagher, 2012).

In this study, we determined the rates of MRSA colonisation and infection on a neonatal and paediatric intensive care unit (NPICU), providing up to Level 3 care, from 2012 to 2015, with this unit being the only unit having intensive care services for neonates and children aged up to three years in the country.

Methods

According to the hospital’s infection control policies, all patients admitted to the NPICU had nasal and umbilical cord surface swabs taken for MRSA detection. Thereafter, surface swabs were taken routinely once a week during the infant’s stay in intensive care to detect acquired colonisation. MRSA isolates from these surface swabs were recorded. Any patients not screened on admission but screened later during their hospitalisation were excluded.

Eye swabs and wound swabs in infants with signs of infection at these sites and from bronchoalveolar lavage performed in children with suspected ventilator associated pneumonia were included if positive for growth of MRSA. These were taken as indicative of non-invasive infections. MRSA-positive blood cultures were taken to indicate invasive MRSA infection.

The rates of MRSA colonisation on admission to the NPICU, MRSA colonisation acquired during hospitalisation, and MRSA non-invasive and invasive infections were calculated. The Chi-squared test was used to assess differences between the study years and to compare mean local rates to published rates of MRSA colonisation from neonatal and paediatric intensive care units in North America, Asia and Europe (Zervou et al., 2014).

Results

There was a mean of 357 admissions (95% confidence interval [CI] = 337.23–377.27) to the NPICU per year over the four-year study period. Of these, a mean of 92% of patients (329/357; 95% CI = 291.28–365.72) were screened at some point during their admission, but only a mean of 78% of patients admitted (279/357; 95% CI = 242.29–315.21) were actually screened on admission as per protocol. The male:female ratio was 1.35, the median age at admission was 0 days and median age at time of screening was one day. The mean length of stay when a positive swab was first detected for acquired colonisation was 22.95 days (95% CI = 17.78–28.13), while that for infection was 39 days (95% CI = -15.23–93.23).

Table 1 is a summary of the data over the four-year study period, including the number of colonised and infected cases per year and the respective rates. The mean rate of colonisation on admission to the NPICU was 3.71% (95% CI = 2.17–5.25), while the mean rate of acquired colonisation was 14.60% (95% CI = 6.16–23.04). There were no cases of invasive MRSA infection between 2012 and 2015, but the mean rate of non-invasive infections was 0.77% (95% CI = 0.54–1.01). The relative risk for MRSA infection among patients who were colonised (when compared to patients who were not colonised) was 14.81 (95% CI = 4.39–50.01), compared to the 24.20 (95% CI = 8.90–66.00) calculated in a meta-analysis by Zervou et al. (2014).

Table 1.

Summary data for 2012–2015.

	2012	2013	2014	2015	Mean
Admissions (n)	361	370	340	358	357
Total patients screened at any point in time (n)	340	348	295	331	329
Patients screened on admission (n)	280	296	246	293	278.75
Patients rescreened (n)	147	156	139	161	150.75
Initially MRSA-negative patients rescreened (n)	142	153	136	157	147
Patients first screened beyond 48 h after admission (n)	60	52	49	38	49.75
Patients colonised with MRSA on admission (n)	13	7	10	11	10.25
Rate of MRSA colonisation on admission (%)	4.64	2.36	4.07	3.75	3.71
Patients who acquired MRSA during hospitalisation (n)	14	22	30	19	21.25
Rate of acquired MRSA colonisation (%)	9.86	14.38	22.06	12.10	14.60
Patients with non-invasive MRSA infection (n)	2	3	3	3	2.75
Rate of non-invasive MRSA infection (%)	0.55	0.81	0.88	0.84	0.77

The most frequently colonised site was the nose, at 87.30% of the colonised cases. The most common MRSA infection site was the conjunctiva, at 63.64% of all infected cases.

There was no significant difference between the years for the rate of MRSA colonisation on admission (P > 0.05). The 3.71% average rate of MRSA colonisation on admission over these four years was comparable to the average MRSA colonisation rate on admission obtained from the studies done in North America (3.05%, P > 0.05), but higher than rates in Europe (1.59%, P = 0.01) and Asia (1.66%, P = 0.02) (Zervou et al., 2014). The pooled prevalence of MRSA colonisation on admission as determined by Zervou et al. (2014) was 1.9%, with the local rate being significantly higher (P = 0.04).

There was a significant difference between the rates of acquired MRSA colonisation between the years 2012 and 2014 (9.86% and 22.06%, respectively; P = 0.005), and between 2014 and 2015 (22.06% and 12.10%, respectively; P = 0.02). The 14.60% average rate of acquired MRSA colonisation was significantly higher than the pooled acquisition rate of MRSA colonisation of 4.1% (P < 0.001) in units abroad (Zervou et al., 2014).

There were no significant differences between the rates of MRSA infection at the NPICU across the four years (P > 0.05).

Discussion and conclusion

Colonisation of neonates and children on the NPICU with MRSA occurs mainly by horizontal transmission via contact with healthcare workers, the hospital environment or other contacts (Nelson and Gallagher, 2012). However, vertical transmission from the mother, such as during passage through the birth canal and through breast milk, has also been reported (Nelson and Gallagher, 2012). Infants may thus already be colonised with MRSA on admission because of transmission from the mother or other contacts or the environment outside the unit.

MRSA colonisation acquired during the patient’s stay in the NPICU may be of a different subtype: hospital-associated MRSA, which has a different genotype, characteristics and sensitivities when compared to community-associated MRSA (Giuffrè et al., 2013). Overcrowding and understaffing in a neonatal intensive care unit have been shown to increase spread of MRSA (Nelson and Gallagher, 2012). The NPICU in question is the only unit providing intensive care for neonates and young children in a country with a population of around 500,000 which has otherwise limited access to other units in view of it being geographically isolated. This could therefore partly explain the significantly higher average acquired MRSA colonisation rate obtained in this study.

Similar to other intensive care units, the NPICU in this study implements policies to prevent MRSA transmission, with a policy providing guidance on how to manage MRSA available on the hospital Intranet. Despite this, the acquisition rate of MRSA is still relatively significantly high and this is a phenomenon that has also been described in units elsewhere (Popoola et al., 2014). A possible explanation could be the wait between time of sampling and the laboratory issuing a result, leading to a delay in implementing infection control measures (Zervou et al., 2014). Polymerase chain reaction (PCR)-based screening methods give quicker results and have been shown to lead to a reduction in the rate of MRSA transmission (Cunningham et al., 2007). Another possible explanation could be that there is suboptimal implementation of infection control policies (Zervou et al., 2014).

The significant decrease between 2014 and 2015 in the acquired MRSA colonisation rate could be explained by a reported outbreak in October 2014 that then led to the application of rigorous infection control procedures.

Hand hygiene and screening for MRSA on admission have been shown to be the most successful infection control policies to reduce transmission of MRSA (Raboud et al., 2005). The nares and umbilicus being the most common sites for MRSA colonisation (Huang et al., 2006), these sites are sampled for screening purposes on admission for all patients in the NPICU. Active surveillance is done since asymptomatic carriers are the main source of transmission and screening on admission allows detection of carriers and thus implementation of infection control strategies early on (Giuffrè et al., 2013). Once MRSA colonisation or infection is identified, according to the local policy guideline, the infant/child should be placed in a single room or at least cohorted with other similarly colonised or infected patients. Apart from the standard hand hygiene and ‘bare below the elbow’ policies, further contact precautions with use of aprons and gloves should be implemented, and medical equipment coming into contact with the patient’s skin should be dedicated for that sole patient. Recommendations for cleaning of the surrounding environment are also given in the hospital’s policy. The standard decolonisation protocol involves use of mupirocin.

Such measures are all part of a consensus statement by Gerber et al. (2006) that brought forward a number of recommendations for control of MRSA in neonatal intensive care units. Other recommended measures included cohorting of nurses for care of MRSA-positive patients, periodic screening of patients in the neonatal unit (weekly to monthly depending on local MRSA transmission rates) and use of molecular epidemiological tools to investigate an outbreak (Gerber et al., 2006).

A possible limitation of this study is that local rates for 2012 till 2015 were compared with rates from studies carried out in 2010 or before. Also, there are differences in methodology between the local study and the studies from which pooled rates were obtained, and local numbers were smaller. Since not all patients were screened on admission and a number of these patients were later identified as colonised with MRSA, it was not possible to determine whether these were colonised on admission or whether colonisation was acquired during their stay on the unit; these patients were thus excluded from calculations of colonisation rates.

More efforts need to be taken to adhere to infection control measures already enacted in order to further decrease the acquisition of MRSA colonisation on NPICU. Adoption of molecular diagnostic techniques to detect MRSA colonisation more rapidly for immediate isolation and containment of spread are needed (Zervou et al., 2014).