MRSA indicates methicillin-resistant S aureus.
Ericson JE, Popoola VO, Smith PB, Benjamin DK, Fowler VG, Benjamin DK, Clark RH, Milstone AM. Burden of Invasive Staphylococcus aureus Infections in Hospitalized Infants. JAMA Pediatr. 2015;169(12):1105-1111. doi:10.1001/jamapediatrics.2015.2380
Staphylococcus aureus is a frequent cause of infection in hospitalized infants. These infections are associated with increased mortality and morbidity and longer hospital stays, but data on the burden of S aureus disease in hospitalized infants are limited.
To compare demographics and mortality of infants with invasive methicillin-resistant S aureus (MRSA) and methicillin-susceptible S aureus (MSSA), to determine the annual proportion of S aureus infections that were MRSA, and to contrast the risk of death after an invasive MRSA infection with the risk after an invasive MSSA infection.
Design, Setting, and Participants
Multicenter retrospective study of a large, nationally representative cohort at 348 neonatal intensive care units managed by the Pediatrix Medical Group. Participants were 3888 infants with an invasive S aureus infection who were discharged from calendar year 1997 through calendar year 2012.
Invasive S aureus infection.
Main Outcomes and Measures
The incidence of invasive S aureus infections, as well as infant characteristics and mortality after MRSA or MSSA infection.
The 3888 infants had 3978 invasive S aureus infections (2868 MSSA and 1110 MRSA). The incidence of invasive S aureus infection was 44.8 infections per 10 000 infants. The yearly proportion of invasive infections caused by MRSA increased from calendar year 1997 through calendar year 2006 and has moderately decreased since then. Infants with invasive MRSA or MSSA infections had similar gestational ages and birth weights. Invasive MRSA infections occurred more often at a younger postnatal age. For infants with available mortality data, more infants with invasive MSSA infections (n = 237) died before hospital discharge than infants with invasive MRSA infections (n = 110). The proportions of infants who died after invasive MSSA and MRSA infections were similar at 237 of 2474 (9.6%) and 110 of 926 (11.9%), respectively (P = .05). The adjusted risk of death before hospital discharge was similar after invasive MSSA and MRSA infections (risk ratio, 1.19; 95% CI, 0.96-1.49). The risks of death at 7 and 30 days after invasive infection were similar between infants with invasive MSSA infection and infants with invasive MRSA infection.
Conclusions and Relevance
Infant mortality after invasive MRSA and MSSA infections is similar, but MSSA causes more infections and more deaths in infants than MRSA. Measures to prevent S aureus infection should include MSSA in addition to MRSA.
Staphylococcus aureus is a frequent cause of infection in hospitalized infants.1 Infections due to S aureus are associated with increased mortality,2 morbidity,3 and length of hospital stay.4Staphylococcus aureus is the second most frequent cause of late-onset sepsis in very low-birth-weight (VLBW) infants less than 1500 g1 and is resistant to methicillin sodium in 8% to 28% of cases.5- 7
Antibiotic-resistant S aureus strains, specifically methicillin-resistant S aureus (MRSA), have emerged and become prevalent in neonatal intensive care units (NICUs). Because MRSA-colonized infants often serve as a reservoir for spread to other infants, many NICUs have developed procedures to detect and isolate colonized infants.8,9 Identification of colonization is often followed by efforts to eradicate MRSA.8,10 Although infection due to methicillin-susceptible S aureus (MSSA) also occurs, few centers have screening protocols for MSSA.11
The proportion of nosocomial S aureus infections caused by MRSA has decreased in some settings, including NICUs.12,13 However, most studies5,6,14 evaluating S aureus and MRSA in NICU patients have included small numbers of patients or culture sites and have been conducted in single institutions, at tertiary care centers, or within outbreak settings. Our objective was to describe the epidemiological characteristics of S aureus invasive infections using a large, nationally representative cohort of hospitalized infants.
Staphylococcus aureus is a frequent cause of infection in hospitalized infants. These infections are associated with increased mortality and morbidity and longer hospital stays.
This retrospective study of 348 neonatal intensive care units in the United States found that invasive S aureus infections occurred in 44.8 of 10 000 hospitalized infants.
Invasive methicillin-susceptible S aureus (MSSA) infections were more common than invasive methicillin-resistant S aureus (MRSA) infections, 2868 vs 1110, respectively.
Mortality after invasive MSSA and MRSA infections was similar (risk ratio, 1.19; 95% CI, 0.96-1.49).
Measures to prevent S aureus infection should include MSSA in addition to MRSA.
We identified all infants with an invasive S aureus infection who were discharged from calendar year 1997 through calendar year 2012 from 348 NICUs managed by the Pediatrix Medical Group. This group cares for more than 20% of infants admitted to NICUs in the United States and is composed of academic and community sites in 34 states. Data were obtained from the electronic medical record generated prospectively by clinicians on all infants. Data were extracted, deidentified, and stored in the Pediatrix Clinical Data Warehouse.13 Information on infants includes maternal history, birth information, and demographics. Medications, laboratory values, microbiology results, diagnoses, and procedures are recorded on a daily basis. Microbiology results are reported as documented by the treating physician in the medical record. This study was approved by the Duke University Institutional Review Board.
Positive S aureus cultures obtained within 21 days of each other were considered to be a single infection.15 Infections in which any positive culture was obtained from cerebrospinal fluid, blood, sterile fluid, or an abscess were considered to be invasive. Infections in which all positive cultures were obtained from the trachea, urine, conjunctiva, or a wound were considered to be noninvasive. Abscess cultures included all cultures labeled “abscess” and may have included both cutaneous and deep tissue abscesses. Because the method of collection was missing for most urine samples, positive urine cultures for MSSA and MRSA were considered to be noninvasive. Cultures obtained from the skin surface, umbilicus, rectum, nasopharynx, or gastric aspirate were considered to represent surveillance cultures. Noninvasive and surveillance cultures were excluded from the analysis. Cultures for which the specimen type was “unknown” or “other” were also excluded. Infections that had growth of both MSSA and MRSA were considered to be MRSA infections.
Inotropic support was defined as exposure to dopamine hydrochloride, dobutamine, epinephrine, norepinephrine bitartrate, or milrinone. Mechanical ventilation was defined as exposure to any invasive mechanical ventilation. Oxygen supplementation was defined as the administration of any fraction of inspired oxygen exceeding 21%. Antibiotic exposure was defined as exposure to any antibiotic. Anti-MRSA antibiotic exposure was defined as exposure to vancomycin, rifampin, linezolid, daptomycin, clindamycin, trimethoprim-sulfamethoxazole, or mupirocin. Small-for-gestational-age status was defined as previously described.16
We determined the proportion of S aureus infections that were methicillin resistant by year of discharge and by NICU site. We calculated the incidence of positive MSSA and MRSA infections per 10 000 admitted infants for each year.
Infant demographics were assigned to S aureus category (MRSA vs MSSA) by the type of the first invasive infection and were compared using Wilcoxon rank sum tests for continuous variables and χ2 tests for categorical variables. Using Wilcoxon rank sum tests, we compared the median numbers of days of exposure to inotropic support, oxygen support, ventilator support, antibiotic use, and MRSA-active antibiotic use before the first positive invasive culture for infants with MRSA vs those parameters for infants with MSSA.
We determined the number of deaths that occurred before hospital discharge and within 7 and 30 days after the first positive culture of an invasive infection, both overall and separately for infants born at less than 1500 g and for infants born at 1500 g or higher. For infants with more than 1 invasive infection, we considered the risk of death after the last infection. We compared the risk of death after an invasive MRSA infection with the risk of death after an invasive MSSA infection using modified Poisson regression adjusted for gestational age, small-for-gestational-age status, male sex, and race/ethnicity, with random effects for site.
Statistical analyses were performed using a software package (Stata, version 14; StataCorp LP). Statistical significance was set at P < .05.
We identified 3888 of 887 910 infants (0.4%) with 3978 invasive S aureus infections. Infections were caused more commonly by MSSA (2868 of 3978 [72.1%]) than MRSA (1110 of 3978 [27.9%]).
Overall, invasive S aureus infections occurred at an incidence of 44.8 infections per 10 000 infants. The annual incidence of invasive S aureus infection increased from calendar year 1997 through calendar year 2006 and then declined from calendar year 2007 through calendar year 2012 (Figure 1). Invasive S aureus infections were more common in infants born at less than 1500 g (3061 of 136 797 [223.8 per 10 000 infants]) than in infants born at 1500 g or higher (915 of 748 715 [12.2 per 10 000 infants]) (P < .01) (Table 1). For all birth weight categories, MSSA was more common than MRSA.
Most characteristics of infants with invasive MSSA and MRSA infections were similar, including gestational age, birth weight, and the proportions who were male, small for gestational age, or born by cesarean section (Table 2). For all races/ethnicities, MSSA was more common than MRSA. A higher proportion of infants with MRSA were African American than those with MSSA (330 of 1035 [31.9%] and 681 of 2725 [25.0%], respectively). The median postnatal age at occurrence was younger for MRSA infections (15 days [9-26 days for the 25th-75th percentiles]) than for MSSA infections (18 days [10-32 days for the 25th-75th percentiles]) (P < .001). When considering exposure to potential risk factors before the first invasive S aureus infection, the median duration of oxygen support before the infection was shorter in infants whose first invasive S aureus infection was MRSA (5 days [1-15 days for the 25th-75th percentiles]) than in infants whose first invasive S aureus infection was MSSA (8 days [1-20 days for the 25th-75th percentiles]) (P < .001). Previous exposure to a surgical procedure and the number of days of antibiotic use, MRSA-active antibiotic use, ventilator support, and inotropic support before the first invasive S aureus infection were similar for infants with MSSA and infants with MRSA.
Among the 2474 infants with invasive MSSA infections for whom mortality information was available, 237 (9.6%) died before hospital discharge. Mortality was similar among the 926 infants with invasive MRSA infections, with 110 (11.9%) dying before hospital discharge (P = .05).
Death before hospital discharge occurred more often in infants with invasive S aureus infection born at less than 1500 g (302 of 2596 [11.6%]) vs born at 1500 g or higher (45 of 802 [5.6%]) (P < .01). When the risk of death was adjusted for gestational age group, small-for-gestational age status, sex, and race/ethnicity using random effects for site, there was no statistically significant difference in the risk of death after an invasive MSSA infection compared with after an invasive MRSA infection overall or separately for infants born at less than 1500 g vs born at 1500 g or higher (Table 3). The risk of death was also similar at 7 and 30 days after the start day of the invasive infection.
The proportion of S aureus infections resistant to methicillin increased from calendar year 1999 through calendar year 2006 (Figure 2). The proportion of S aureus cultures that were MRSA declined after 2007, but the annual proportion of MRSA cultures remains higher than at the beginning of our sample. For the 100 sites with at least 10 episodes of invasive S aureus infection, the proportion of invasive S aureus infections caused by MRSA varied widely by site (median, 26.3% [16.7%-36.8% for the 25th-75th percentiles]). The proportions of invasive infections caused by MRSA ranged from 0% (0 of 35) at 6 sites to 86.7% (13 of 15) at 1 site.
In this large representative cohort of hospitalized infants, invasive S aureus infections occurred in 0.4% (3888 of 887 910) of all infants and in 2.2% (3061 of 136 797) of VLBW infants. Overall, MSSA caused 2.6 times more invasive S aureus infections than MRSA. The adjusted risk of death was similar for MSSA and MRSA invasive infections overall and for VLBW infants.
Our findings confirm results of earlier studies demonstrating that MSSA infections are more common than MRSA infections. A single-center retrospective study6 of 172 S aureus infections found that 72% of S aureus infections were caused by MSSA, with most of the infections occurring in infants born at less than 1000 g. A study5 of VLBW infants in the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network similarly found that 72% of bacteremia and meningitis cases due to S aureus were caused by MSSA. Infection prevention strategies should broadly target both MSSA and MRSA in infants.
In our cohort, invasive S aureus infections occurred at an incidence of 2.2% (3061 of 136 797) in VLBW infants. Previous studies5,6 found that the incidence of S aureus infections in this population was approximately 3%. Our observed disease incidence of 2.2% may better reflect the national burden of disease but may not be applicable to all high-acuity centers. If current birth trends continue, invasive S aureus disease incidences of 2.2% in VLBW infants and 0.1% (915 of 748 715) in non-VLBW infants will result in more than 5000 invasive S aureus infections each year.17
Studies18- 20 of older patients have found that prolonged hospital stay, presence of a central catheter, recent surgical procedures, and recent antibiotic therapy may predispose patients to infection with MRSA. In infants, some studies6,21,22 have noted that low birth weight and younger gestational age, as well as a recent surgical procedure and prior treatment with systemic antibiotics, are associated with a higher incidence of MRSA colonization and subsequent infection. We found that infants who were of African American race/ethnicity, born at another hospital and transferred, and younger postnatal age had a greater proportion of S aureus infections that were MRSA than were MSSA. Other studies6,23,24 have also noted that infants with MRSA infection have a younger postnatal age than infants with MSSA infection. This finding is intuitively unexpected because MRSA is often a hospital-associated infection, so we would expect that more days in the hospital and more days of antibiotic exposure would increase the risk of MRSA infections. Other studies25- 27 have found MRSA infections to be more common in African American patients as well. The reasons for this difference are not clear but may include variations in immune function, maternal colonization rates, or other unknown factors.28
Worse outcomes after MRSA infections compared with after MSSA infections have been noted in older patients with bloodstream infections,29,30 endocarditis,31 and osteomyelitis.32 However, differences in outcomes after MRSA infection vs after MSSA infection in infants have been less clear. We found no difference in adjusted mortality after MRSA compared with after MSSA invasive infection at 7 days, 30 days, or hospital discharge. A single-center study6 of 172 S aureus infections in hospitalized infants found that mortality was similar between infants with MRSA and MSSA (16% vs 17%, respectively; P > .99), and length of hospital stay was also similar (64 days for MRSA and MSSA, P = .80). Another single-center study24 found no differences in length of hospital stay (P = .70), death (P = .20), or the incidence of neurodevelopmental impairment at 12 to 36 months of age (P = .28) for 53 infants with MRSA bacteremia compared with that caused by MSSA. A multicenter retrospective study5 of 316 infants found no difference in survival for infants born at 1500 g or less with MRSA vs MSSA bacteremia or meningitis.
Given similar outcomes in infants with MRSA and MSSA disease, strategies should focus broadly on S aureus prevention and not solely on prevention of MRSA. Findings of single-center studies have suggested that MRSA screening and decolonization may reduce MRSA infection.14 However, a lack of controlled clinical trials has led to a large variation in approaches in MRSA identification and treatment in the NICU.8 Although MSSA causes more infections and more deaths per year in infants, most centers only consider MRSA in their screening and decolonization protocols.8 As previously described, a broader approach to S aureus prevention should go beyond MRSA to also include MSSA.10 When MSSA has been included in screening and decolonization efforts, subsequent MSSA infections have decreased.33,34 One NICU was able to dramatically reduce both MRSA and MSSA infections with a universal S aureus screening and decolonization program.33 Screening and decolonization protocols have been demonstrated to reduce MRSA and MSSA infections in adult orthopedic and cardiac surgery patients.35,36
The annual proportion of MRSA cultures in our study population increased from calendar year 1996 through calendar year 2006, when a decline occurred. The early increase has been seen in other studies. A study37 that included data from 40 freestanding children’s hospitals found that the incidence of MRSA increased from 2002 to 2007, with the MSSA incidence remaining stable. A 10-year single-center study23 that included 156 infants with S aureus bloodstream infections found a peak in the incidence of infections due to MRSA in 2005, with a steadily decreasing incidence from 2006 to 2009. Further studies are needed to identify differences between centers having high and low proportions of isolates with methicillin resistance.
The decline in MRSA incidence that we observed starting in 2007 is also consistent with findings by other investigators. A study12 that included 1751 infants found that surveillance cultures positive for MRSA peaked in 2006 at 5.8% and then plateaued, with approximately 3% of cultures being positive. Using national surveillance data, the US Centers for Disease Control and Prevention found that the incidence of MRSA decreased almost 10% per year from 2005 to 2010 in infants younger than 90 days’ postnatal age.28 A large study13 that included infants, children, and adults found that S aureus bacteremia decreased by 41% from 2006 to 2010. This stabilization in MRSA rates may be owing to improved infection control practices, leading to fewer hospital-acquired infections.28
We opted to exclude surveillance and noninvasive cultures from the analysis because of differing surveillance practices (eg, centers with no surveillance, surveillance for MRSA only, and varying surveillance frequency and methods). Specimens from invasive infections may have been classified as surveillance cultures or noninvasive infections, but this number was likely small and should have been similar for MRSA and MSSA. Our analysis evaluated all-cause mortality and not S aureus–attributable mortality. However, the relative risk of death after invasive MRSA compared with after invasive MSSA is not likely to be influenced by this distinction.
The absolute numbers of infections and deaths due to MSSA exceed those due to MRSA. Consideration should be given to expanding hospital infection control efforts targeting MRSA to include MSSA as well.10,38,39 Future studies to better define the relationship between MSSA colonization and subsequent infection will help to clarify the importance of such interventions for preventing MSSA disease.
Accepted for Publication: July 11, 2015.
Corresponding Author: P. Brian Smith, MD, MPH, MHS, Duke Clinical Research Institute, Duke University School of Medicine, PO Box 17969, Durham, NC 27715 (firstname.lastname@example.org).
Published Online: October 19, 2015. doi:10.1001/jamapediatrics.2015.2380.
Author Contributions: Drs Ericson and Daniel K. Benjamin had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Ericson, Popoola, Smith, Daniel K. Benjamin Jr, Clark, Milstone.
Acquisition, analysis, or interpretation of data: Ericson, Popoola, Daniel K. Benjamin, Fowler, Clark, Milstone.
Drafting of the manuscript: Ericson, Daniel K. Benjamin Jr, Clark, Milstone.
Critical revision of the manuscript for important intellectual content: Ericson, Popoola, Smith, Daniel K. Benjamin, Fowler, Clark, Milstone.
Statistical analysis: Ericson, Popoola, Daniel K. Benjamin.
Obtained funding: Daniel K. Benjamin Jr.
Administrative, technical, or material support: Clark.
Study supervision: Smith, Clark, Milstone.
Conflict of Interest Disclosures: Drs Smith and Daniel K. Benjamin Jr reported receiving research support from Cempra Pharmaceuticals (subaward to HHS0100201300009C) and industry for neonatal and pediatric drug development (http://www.dcri.duke.edu/research/coi.jsp). Dr Fowler reported receiving grant or research support from Advanced Liquid Logic, Cubist, Cerexa, MedImmune, Merck, National Institutes of Health (NIH), Novartis, Pfizer, and Theravance; reported being a paid consultant to Affinium, Basilea, Baxter, Cerexa, Cubist, Debiopharm, Durata, Merck, Novartis, NovaDigm, The Medicines Company, MedImmune, Pfizer, Theravance, and Trius; reported receiving honoraria from Arpida, Astellas, Cubist, Inhibitex, Merck, Pfizer, Targanta, Theravance, Wyeth, Ortho-McNeil, Novartis, and Vertex Pharmaceuticals; and reported being a member of the Merck V710 Vaccine Scientific Advisory Board. Dr Milstone reported receiving grant support from Sage Products LLC. No other disclosures were reported.
Funding/Support: Dr Ericson is supported by grant 5T32HD060558 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the NIH. Dr Smith is supported by grant UL1TR001117 from the National Center for Advancing Translational Sciences of the NIH, by grants HHSN275201000003I and 1R01-HD081044-01 from the NICHD, and by grant 1R18-FD005292-01 from the US Food and Drug Administration. Dr Fowler is supported by grants K24-AI093969 and R01-AI068804 from the NIH and by grant UM1AI104681 from the NIH Antibiotic Resistant Leadership Group. Dr Daniel K. Benjamin Jr is supported by grant 2K24HD058735-06 from the NIH, by grant UL1TR001117 from the National Center for Advancing Translational Sciences, by contract HHSN275201000003I from the NICHD, and by contract HHSN272201500006I from the National Institute of Allergy and Infectious Diseases (NIAID). Dr Milstone is supported by grant R03AI117169 from the NIAID and by grant R01HS022872 from the Agency for Healthcare and Research Quality.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.