Staphylococcus aureus is responsible for a considerable portion of community-acquired (CA) bacterial infections that lead to hospitalization in children and causes several clinical syndromes, including healthcare-associated (HA) infections. Determining the bacterium's susceptibility to oxacillin is the first step in deciding upon a course of treatment; strains susceptible to this antimicrobial are called methicillin/oxacillin-susceptible S. aureus (MSSA) or methicillin/oxacillin-resistant S. aureus (MRSA).

Although MRSA clones have been historically associated with HA-MRSA, the proportion of clones responsible for community-acquired infections (CA-MRSA) has been increasing significantly worldwide. This epidemiological change has relevant clinical implications, since empirical antibiotic therapy for community-acquired staphylococcal infections often involves beta-lactams, a class that, except for ceftaroline, is not active against MRSA. Although CA-MRSA strains tend to be susceptible to non-beta-lactam antibiotics, the incidence and antimicrobial susceptibility profile can vary greatly and, in Brazil, the subject is little investigated, especially in pediatric populations (1–6).

The present study aims to define the proportion of MRSA isolates in a pediatric hospital in southern Brazil, comparing the characteristics of patients with MRSA and MSSA infections, as well as describing the susceptibility of community-acquired isolates to other antimicrobials.

Of the 654 cultures evaluated, 91 were excluded due to contamination (N = 44, 48.3%), polymicrobial growth (N = 41, 45.1%), and incomplete medical records (N = 6, 6.6%). A total of 563 patients were included and, among patients with a CA infection, the MRSA prevalence was 46.1% (129/280), whereas only 8.1% of HA infections involved MRSA isolates (23/283). The sample was 56% male and its median age was 2 years. There was no significant age difference between patients with CA-MSSA and CA-MRSA infections (p = 0.928). The majority of collected materials were abscess secretions (50.8%) and blood cultures (41.6%). The most frequent primary infection site was the skin/soft tissues (38.9%), followed by primary bloodstream (23.6%) and surgical site (20.8%) (Table 1).

Regarding the strains' oxacillin susceptibility (MSSA or MRSA), 46.1% of CA infections were MRSA and 8.1% of HA infections were MRSA. The MRSA prevalence did not change significantly during the study period for CA or HA infections (p = 0.515 and p = 0.137, respectively) (Table 1 and Figure 1).

There was a significant association between primary infection site and phenotype. Among CA infections, MSSA was significantly more frequent in osteoarticular and primary bloodstream infections (p = 0.001). On the other hand, MRSA was significantly more frequent in respiratory and intra-abdominal infections (p = 0.001). In HA infections, MSSA was significantly more frequent in primary bloodstream infections, while MRSA was significantly more frequent in skin/soft tissue and respiratory infections (p = 0.001) (Table 2).

Regarding the susceptibility of S. aureus to the other antimicrobials, CA-MSSA had a high rate of susceptibility to TMP-SMX, doxycycline, and ciprofloxacin, and a low rate of susceptibility to clindamycin and erythromycin (64.9% and 58.3%, respectively), with a downward but non-significant (p = 0.056) trend toward clindamycin susceptibility during the study period. CA-MRSA had a low rate of susceptibility to ciprofloxacin and erythromycin (78.3% and 55.8%, respectively) and these rates significantly declined over the study period (p < 0.001 and p = 0.005, respectively). The susceptibility of CA-MRSA to TMP-SMX, clindamycin, and doxycycline remained high, being 96.1%, 88.4% and 99%, respectively (Figure 2).

Regarding HA infections, MSSA had a low susceptibility rate to clindamycin and erythromycin (70.8% and 66.2%, respectively). MRSA had a low rate of susceptibility to ciprofloxacin, clindamycin, and erythromycin (52.2%, 60.9%, and 34.8%, respectively). HA-MRSA remained highly susceptible to TMP-SMX and doxycycline (91.3% and 95%, respectively).

Vancomycin resistance was not observed in any of the samples, and only 4 isolates showed resistance to rifampicin: 3 from CA infections (2 MSSA and 1 MRSA) and 1 from a HA infection (MSSA). All isolates were susceptible to some other antimicrobial in addition to vancomycin.

In Brazil, few studies with adequate sample sizes have attempted to determine the prevalence of MRSA in CA infections, which increases the importance of the present study. Our results include alarming data on high resistance to oxacillin, the main antibiotic used to treat clinical community-acquired staphylococcal syndromes, reaching a mean of 46.1% for community-acquired isolates. The chance of therapeutic failure with oxacillin or another beta-lactam in these infections is high, which reinforces the importance of submitting clinical specimens to culture. The pediatric literature recommends an association of beta-lactam and vancomycin for invasive Staphylococcus infections, which is to be adjusted after determining susceptibility, but this practice does not appear to be widespread in pediatric services (9–11).

In contrast to high resistance in community strains, we observed a low prevalence of MRSA (8.1%) in HA infections, indicating that MRSA has been losing ground among nosocomial infections and gaining importance in CA infections. Thus, the choice of antibiotic must be adapted to local epidemiology, with anti-MRSA agents included in the initial regimen for CA infections.

Clindamycin is the main recommendation for CA staphylococcal infections when resistance is suspected, especially in regions with a high prevalence of CA-MRSA (10, 11). However, our study shows a high prevalence (35.1%) of clindamycin resistance among CA-MSSA, with the resistance rate falling to 11.6% in CA-MRSA. These data should raise concern about the empirical use of clindamycin, mainly in cases of MSSA. In comparison, the resistance rates for trimethoprim-sulfamethoxazole and doxycycline were very low: 6.6% (MSSA) and 3.9% (CA-MRSA) for TMP-SMX, and 1.6% (MSSA) and 1% (CA-MRSA) for doxycycline, making them interesting therapeutic options for non-invasive infections. Doxycycline's use has been previously limited to children 8 years and older due to dental toxicity concerns but more recent data support the recommendation by the American Academy of Pediatrics that doxycycline can be safely used for short durations (21 days or less) regardless of the patient's age (9).

The changes in MRSA rates in 2020, which did not follow the pattern of previous years, coincide with the onset of the non-pharmaceutical measures implemented to reduce Sars-CoV-2 circulation (i.e., social isolation and school and daycare centre closures), which could have impacted the dynamics of pediatric CA-MRSA transmission and hospitalization, making a more accurate interpretation difficult.

Our study has several limitations. First, it was a retrospective, single-center study. Samples of non-sterile materials, such as respiratory secretions, were excluded due to the difficulty in differentiating between infection, colonization, and contamination. This, for example, could have led to an underestimation of respiratory infections, both for CA staphylococcal pneumonia and ventilator-associated pneumonia. Another limitation was that we analyzed cases based on positive culture for S. aureus. Excluding staphylococcal infections without a positive culture may have resulted in selection of more serious cases. This is especially important, considering that CA-MRSA strains can result in suppurative and more severe infections due to the production of toxins such as Panton-Valentine leukocidin. We believe that this, in association with the majority of CA infections being skin/soft tissue infections (71.8%), may have led to overestimation of the MRSA rate in CA infections. Nevertheless, the rate we found was significantly high, especially compared to similar studies in Brazil (1, 12).

In conclusion, our study showed high rates of MRSA in CA staphylococcal infections, highlighting the need to review, according to local epidemiology, initial treatment protocols for severe staphylococcal infections that require hospitalization. Further epidemiological studies are also needed to determine susceptibility differences in each region of the world.