Abstract
INTRODUCTION:
Methicillin-resistant Staphylococcus aureus (MRSA) is a common etiological agent of a life-threatening infection in neonatal intensive care units (NICUs). Neonates with very low birth weight and patients with serious diseases are more likely to be exposed to invasive procedures which make them at a high risk of MRSA colonization and infection. Since MRSA colonization is a risk factor for MRSA infection, prevention of MRSA transmission is an important issue in NICUs. NICUs in Japan practice standard contact precautions and active surveillance cultures (ASC) to prevent MRSA transmission. In this report, we analyzed the clinical characteristics of MRSA colonization and infection between January 2010 and December 2015 in our perinatal care center.
METHODS:
We conducted retrospective analysis of 1716 neonates hospitalized in our perinatal care center.
RESULTS:
120 cases had MRSA colonization (6.99%) and among them 33 neonates were infected. The duration of stay (P≤0.001) and the birth weight (P≤0.001) showed statistically significant differences between MRSA-colonized neonates and non-MRSA-colonized neonates. The number of central venous catheterization showed statistically significant differences (P = 0.001) and the number of digestive system diseases showed marginally significant differences (P = 0.072) between MRSA-colonized non-infected neonates and MRSA-infected neonates.
CONCLUSIONS:
As previous reports have shown, we present that the neonates with central venous catheterization were more likely to be infected with MRSA. We also need to pay attention to neonates with digestive system diseases, showing signs of infection, because they may be potentially infected with MRSA.
Keywords
Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) is a frequent cause of infection in premature and critically ill neonates in neonatal intensive care units (NICUs). They are particularly vulnerable to MRSA colonization and infection, therefore many studies have been carried out to clarify the risk factors for MRSA colonization and infection. MRSA colonization and infection in neonates are significantly associated with high morbidity and cost of medical treatment. MRSA infection is associated with serious sequelae in neonates, therefore much effort has been made to control epidemics in NICUs. In our institution, we perform various practices including collecting active surveillance cultures (ASCs), hand hygiene, frequent hand washing, and the usages of disposable rubber gloves and gowns for patient care. Neonates with MRSA colonization and infection are isolated with standard contact precautions. ASCs have been taken as strategies of basic infection control targeting both the individual and population level risk of MRSA colonization, and it also enable us to identify MRSA-colonized neonates and place them on contact precautions [1]. The prevalence rates of MRSA colonization and infection in NICUs depend on the facility and many different factors including the surrounding local communities and country, therefore several studies have shown its wide differences between NICUs surveyed [2–5]. Many studies showed various risk factors that predispose neonates to MRSA colonization and infection such as low birth weight, young gestational age, long duration of hospitalization, and the use of various devices including endotracheal intubations, mechanical ventilations, percutaneous central venous catheterizations, and gavage feedings [3, 6]. Several studies have shown that neonatal MRSA colonization is a risk factor for subsequent development of MRSA infection [5]. Nakao et al. reported that the feces of neonates contain MRSA strains at considerable amount and could serve as a potential source of MRSA dissemination in NICUs [7], so we reviewed about the cultures of feces taken from neonates in our center. The purpose of our study is to identify the risk factors for MRSA colonization and infection in our perinatal care center.
Materials and methods
This study was approved by the Toho University Omori Hospital Ethical Committee. We conducted a retrospective study of neonates who were hospitalized in perinatal care center of Toho University Medical Center Omori Hospital between 1 January 2010 and 31 December 2015. Our perinatal care center admits about 300 neonates annually and has 15 beds in NICU and 18 beds in GCU (Growing care unit) for premature neonates and neonates requiring medical and surgical treatments. We perform ASCs in every neonate by collecting samples from nares, ear canals, navel, gastric drainage, and tracheal aspirate on admission. Subsequently, we take the samples for the cultures from nares and navels every two weeks, and from tracheal aspirate every week. Moreover, we carry out infection control measures taking samples from feces of patients who have abdominal symptoms or devices. Samples were collected with SEEDSWAB No. 2 (Eiken chemical Co. Ltd, Tokyo, Japan) and inoculated onto the MRSA screening agar chrom ID MRSA (Bio Merieux, France).
Comparisons between MRSA-colonized neonates and non-MRSA-colonized neonates
Demographic characteristics to compare MRSA-colonized neonates with non-MRSA colonized neonates were as follows; birth weight (g), gestational age (weeks), Apgar score at 1 minute and 5 minutes, and the duration of stay from admission to discharge from our perinatal care center.
Comparisons between MRSA-colonized non-infected neonates and MRSA-infected neonates
MRSA infections were defined as positive results of MRSA surveillance cultures in the setting of clinical signs of infections and subsequent treatments with appropriate antimicrobial therapies. To compare the demographic characteristics between MRSA-colonized non-infected neonates and MRSA-infected neonates, the following factors were evaluated: birth weight (g), gestational age (weeks), Apgar score at 1 minute and 5 minutes, duration from admission to MRSA colonization, and the presence of risk factors within one week before being colonized; intubation, central venous catheterization, gavage feeding, formula feeding, small-for-date neonate (SFD), cardiovascular disease, digestive system disease, respiratory system disease, brain and nervous system disease, urologic disease, and chromosomal aberration. Furthermore, we reviewed the presence of MRSA in fecal samples of the cases with abdominal symptoms or devices such as diarrhea, vomiting, and stomas for cultures. Additionally, we reviewed the number of cases diagnosed with sepsis, the duration of empiric antimicrobial administration, and the presence of MRSA in the cultures of the vaginal secretions of the mothers whose neonates had MRSA colonization.
Statistical analysis
To compare MRSA-colonized neonates with non-MRSA-colonized neonates, demographic characteristics were evaluated using backward stepwise multivariable logistic regression analysis. Odds ratios (OR) with 95% confidence intervals (CI) were calculated. P < 0.05 indicated a statistically significant difference. As for comparisons between MRSA-colonized non-infected neonates and MRSA-infected neonates, demographic characteristics were evaluated using the univariate analysis, and characteristics associated with P value <0.2 in the univariate analysis were included in the backward stepwise multivariable logistic regression analysis. OR with 95% CI were calculated. P < 0.05 indicated a statistically significant difference. The model with the lowest Akaike information criterion was considered to have the best fit, and the Hosmer-Lemeshow goodness of fit was used to determine how well the final model fit the data. Adjusted odds ratios and 95% confidence intervals were calculated for the variables retained in the best fitting models. (SPSS, Inc., Chicago, IL, USA).
Result
Comparisons between MRSA-colonized neonates and non-MRSA-colonized neonates
During the study period, there were 1716 neonates hospitalized in our perinatal care center. The number of enrolled neonates increased gradually during the 6-years study period. The characteristics of the patients on admission to our perinatal care center were as follows: median length of stay was 28.9 days, median birth weight was 2459.0 g, and median gestational age was 36.3 wk. Among the 1716 hospitalized neonates 120 had MRSA colonization (6.99%).
The median of annual prevalence rates of MRSA colonization were 6.39% during study period. There was no MRSA infection outbreak in our perinatal care center.
The clinical characteristics of MRSA-colonized neonates and non-MRSA-colonized neonates are summarized in Table 1. The duration of stay (P≤0.001) and the birth weight (P≤0.001) showed statistically significant differences between MRSA-colonized neonates and non-MRSA-colonized neonates.
The clinical characteristics of MRSA-colonized and non-MRSA-colonized neonates.
The clinical characteristics of MRSA-colonized and non-MRSA-colonized neonates.
*For statistical analyses, backward stepwise multivariable logistic regression analysis was conducted. (SPSS, Inc., Chicago, IL, USA).
Other demographic characteristics such as gestational age (P = 0.612), Apgar score at 1 minute (P = 0.908) and 5 minutes (P = 0.741) did not show statistically significant differences between the two groups.
Among 120 cases of colonized neonates 33 neonates were infected. The clinical characteristics of MRSA-colonized non-infected neonates and MRSA-infected neonates are summarized in Table 2A. Demographic characteristics were evaluated using the univariate analysis (2B), and characteristics associated with P value <0.2 in the univariate analysis were included in the backward stepwise multivariable logistic regression analysis (2C). The frequency of central venous catheterization showed statistically significant differences (P = 0.001) and digestive system diseases showed marginally significant differences (P = 0.072) between MRSA-colonized non-infected neonates and MRSA-infected neonates. The digestive system diseases are composed of meconium aspiration syndrome, gastroesophageal reflux, Hirschsprung’s disease, and gastrointestinal allergy. They had diarrhea, vomiting, and stomas.
The clinical characteristics of MRSA-colonized non-infected neonates and MRSA-infected neonates (2A). *For statistical analyses, demographic characteristics were evaluated using the univariate analysis (2B), and characteristics associated with P value <0.2 in the univariate analysis were included in the backward stepwise multivariable logistic regression analysis (2C). Abbreviation; SFD (small-for-date neonate)
The clinical characteristics of MRSA-colonized non-infected neonates and MRSA-infected neonates (2A). *For statistical analyses, demographic characteristics were evaluated using the univariate analysis (2B), and characteristics associated with P value <0.2 in the univariate analysis were included in the backward stepwise multivariable logistic regression analysis (2C). Abbreviation; SFD (small-for-date neonate)
In our perinatal care center, among 87 MRSA-colonized non-infected neonates, 12 had their fecal samples collected for culture, 5 (41.7%) of which were MRSA-positive. Among 33 MRSA-infected neonates 13 had their fecal samples collected for culture, 8 (61.5%) of which were MRSA-positive. The number of the neonates with sepsis due to MRSA were 9 cases during the study period and all of them had central venous catheterizations within one week before being colonized. Among them 7 cases had late-onset sepsis and 3 were died of sepsis. Empiric antimicrobial administration is discontinued within 7 days in all cases.
Eighty-eight mothers of MRSA-colonized neonates had results of the cultures of vaginal secretions, among them only 2 mothers had MRSA colonization. None of the MRSA-infected neonates were born to mothers with MRSA colonization of vaginal secretions which could make mothers to supply MRSA to their offspring during delivery.
The Centers for Disease Control (CDC) and the Healthcare Infection Control Practices Advisory Committee (HICPAC) guideline for Infection Prevention in the NICU recommended the minimization of the duration of central venous catheterization, intubation, and antimicrobial agents use for neonates in NICU, particularly those who with low birth weight, young gestational age, multiple gestation, and MRSA colonization. CDC also recommended to educate all healthcare providers and families to prevent MRSA transmission [8]. In most of the NICUs in Japan including our center, working clinical staffs practice hand hygiene, wear gloves, and isolate cohorts of MRSA-colonized neonates to prevent MRSA transmission. Hand hygiene protocols dramatically decreased MRSA colonization in Japan [9], since the publication of recommendations from the CDC.
In this study, duration of stay and birth weight showed statistically significant differences between non-MRSA-colonized neonates and MRSA-colonized neonates, as previous reports also showed the same results [10]. Other characteristics, such as gestational age, Apgar score at 1 minute and 5 minutes did not show statistically significant differences.
Various procedures and devices including endotracheal intubations, central venous catheterizations, and surgeries that neonates in NICUs often require during their hospital stay are also associated with increased risk of MRSA colonization. Several studies have shown that neonatal colonization with MRSA is a risk factor for subsequent development of MRSA infection [4, 10].
Our study also showed that central venous catheterizations is the risk factor with statistically significant differences for MRSA infection and digestive system diseases showed marginally significant differences between MRSA-colonized non-infected neonates and MRSA-infected neonates. We suggest that close attention should be given to the cases with digestive system diseases as contact with feces increase inevitably. The neonates who had digestive system diseases needed to be cared with handling with vomitus and feces. Previous reports have emphasized the role of MRSA in feces, Nakao et al. reported that the feces from MRSA carriers contained the same clones as the nasal isolates in considerable amounts [7], suggesting that more careful attention should be paid in the handling of feces in neonates than in adults [11, 12]. The intestines of adults are generally occupied by established microflora which prevents the colonization of newly incorporated bacteria. In contrast, the intestinal microflora of premature neonates admitted to the NICU is not yet established [13], and the bacterial profiles in feces of neonates have also been shown to change during the course of the lactation period [14]. These systematic review highlights that MRSA positivity in feces is a potential risk factor for subsequent infections. Furthermore, gastrointestinal symptoms may not occur in all patients with fecal MRSA colonization, but we have not collect fecal samples who did not have abdominal symptom. When changing diapers or cloths and bathing neonates and infants in the NICU, healthcare workers may come in contact with MRSA. In the cases of inappropriate precautions, hand washing or hand hygiene, the transmission of MRSA is likely to occur. In our center, we do not routinely collect fecal samples for culture previously and we should be aware of importance of taking fecal samples for culture henceforth. During this study period, there were not MRSA outbreak in our center but the increase in the number of MRSA colonized neonates coincided with detection of MRSA in feces.
In this study, the number of the neonates with sepsis due to MRSA were 9 cases during the study period and among them 7 cases had late-onset sepsis and 3 cases were died of sepsis. All of them had central venous catheterizations within one week before being colonized. Ana et al., suggested that central line-associated bloodstream infection (CLABSI) rates increase in preterm infants because they are managed at young gestational ages with naive immune systems, incomplete bacterial microbiome development. Other important risk factors of CLABSI are poor compliance with hand hygiene and use of sterile precautions during insertion and manipulation of central lines [15]. Dimitorios et al., reported that the significance of education and adherence to guidelines of health care providers to reduce CLABSI [16].
Takeda et al., reported that increase of the number of cases of late-onset sepsis resulted in the increase of the prevalence rate of MRSA [17]. Afjeh et al., reported that prolonged empiric antimicrobial administrations (>14 days) for neonates result in adverse consequences including late-onset sepsis and necrotizing enterocolitis [18], and they also noted the necessity of early empiric therapy in very low birth weight neonates because of their immature immune system and subjected to various invasive procedures. In our perinatal care center empiric antimicrobial administration is discontinued within 7 days in all cases. Based on above, antimicrobial administration should be discontinued once symptoms disappear and culture results are negative.
In this study, we presented that ASCs contributed to obtain the reliable information of the MRSA dissemination in our center. Risk factors for MRSA colonization and infection deserve further studies about ASCs taking fecal samples into consideration.
Conclusions
Several reports have shown the clinical epidemiology of MRSA in NICUs, especially regarding the neonates with central venous catheterization. We also need to pay attention to neonates with digestive system diseases, showing signs of infection because they may be potentially infected with MRSA and we should collect fecal samples for culture routinely.
Disclosure statement
We have no conflict of interest about this study that should be declared. Prof. Yoda H. received research grant from Meiji Seika Pharma Corporation and Food Science Institute Foundation. Dr. Arai H. received research grant from Meiji Seika Pharma Corporation. Prof. Tateda K. received research grants from Eiken Chemical Corporation, Meiji Seika Pharma Corporation, Becton, Dickinson and Company, Asahi Kasei Pharma Corporation and Eidia Corporation.