Methicillin-resistant Staphylococcus aureus (MRSA) is one of the most common pathogenic bacteria in intensive care units (ICU) with the majority of isolates demonstrating multidrug resistance (MDR) which impacts clinical therapy [1, 2]. The resistance mechanism of MRSA is mainly secondary to the bacteria acquiring a genetic determinant (methicillin-resistant determinant A and C, abbreviated as mecA or mecC), which encode penicillin-binding protein PBP2A or PBP2A${}^{\prime }$respectively [3]. These mecA and mecC genes exist in the staphylococcal cassette chromosome mec, called SCCmec. SCCmec can carry other drug resistance genes other than mecA and mecC gene resulting in multiple drug resistance [4].

Emerging MRSA and multiple drug resistance are a major public health problem worldwide [5, 6]. They are the most common cause of healthcare-associated infections (HAI) in patients that are admitted to the ICU [7]. HAI occurring in the ICU from MRSA have become particularly problematic since they arise from the treatment received by critically-ill patients [8]. Although evidence suggests a significant increase in the proportion of MRSA hospital infections worldwide, ICU at obstetrics & gynaecology departments have reported only a limited number of MRSA isolates in China [9]. Effective and safe antimicrobial treatment is essential for treating infections in the ICU. Organizational-wide surveillance of infection-derived bacterial isolates and analysis of their susceptibility to different antimicrobial agents provides crucial information for the most effective antimicrobial therapy [10]. Furthermore, a comprehensive analysis of MRSA SCCmec typing of ICU infections and predicting the development trend of drug-resistant strains is critical in evaluating disease prognosis and essential for reducing infection mortality and morbidity in the ICU.

Molecular typing of MRSA is an important assay for the epidemiologic investigation and strains of origin in addition to antimicrobial agent selection and therapy. Recent data demonstrate that mecA gene expresses itself in coagulase-negative staphylococcus (CNS) while femB is considered commonly in S. aureus with restricted expression but not present in CNS. Therefore, only when mecA and femB genes are both found to be positive can MRSA be present[14]. This study amplified mecA gene (310 bp) and femB gene (651 bp) from the 120 MRSA strains isolated from clinical samples obtained that were determined to be MRSA.

In the form of a gene complex, mecA gene of MRSA exists in SCCmec that is composed of two gene complexes: mec gene complex and cassette chromosome recombinases (ccr). According to the structures of mec and ccr, SCCmec can be divided to five types [15]. Evidence exists that a majority of early nosocomial infections were generally considered as SCCmec type I while community-acquired MRSA infections were mostly SCCmec type IV and SCCmec type V [16, 17].

In the present study, multiplex PCR amplification was conducted to analyze characteristic genes of SCCmec genotypes. The findings showed that the predominant genotype was SCCmec type III (80%) followed by SCCmec type II (12.5%) while no SCCmec type IV or SCCmec type V was found among MRSA strains in the ICU. As SCCmec type IV and SCCmec type V are mainly community-acquired MRSA infections, it can be verified that community-acquired MRSA infections are not present in our ICU. This finding was in agreement with other reports that 250 clinically isolated MRSA strains were mainly SCCmec type III followed by SCCmec type II in 18 hospitals nationwide in China [18]. An explanation may be the fact that the regional difference provided disparate MRSA with a diversity of SCCmec genotypes, thus leading to distinct drug-resistant patterns. In addition, 9 (7.5%) of 120 isolated MRSA strains of this study were undefined SCCmec strains, probably caused by the selected primers and amplification conditions or that they belong to a novel SCCmec gene, which needs further studies.

In the study, all MRSA strains were recognized as multidrug resistant. They showed a high incidence of resistance to $\beta$-lactamase, quinolone antibiotics, aminoglycosides, tetracycline and macrolides. One of the benefits from SCCmec genotyping in MRAS isolates is differentiation of antimicrobial agent susceptibility patterns. According to our findings, SCCmec type II demonstrated resistance to clindamycin (93.3%) and tetracycline (26.7%). SCCmec type III was strongly resistant to gentamicin (91.7%) while undefined SCCmec types had high drug-resistance (77.8%) to sulfamethoxazole. Due to frequent immunosuppression, patients in ICU receive more antibacterial agents than ordinary patients. Therefore, the frequencies of drug-resistant strains significantly increase in ICU compared to general wards. All isolates were sensitive to quinupristin, vancomycin and nitrofurantoin. Similar to findings reported by others [12, 19].

Interestingly, a high prevalence of resistance to rifampicin was noted in the hospital. In fact, rifampicin was seldom administered for the treatment of MRSA. However, the percentage of rifampicin-resistant MRSA rapidly increased from 15.5% in 2004 to 50.2% in 2008 in China. Several reported that rifampicin resistance in S. aureus isolates including MRSA was associated with mutations of rpoB gene (encoding $\beta$ subunites of RNA polymerase), which conferred homogeneous methicilln resistance. However, no definitive mechanism has been elucidated to date [19, 20, 21].

The treatment of MRSA infection is a very difficult clinical problem. Risk factors for development of an MRSA infection A in the ICU include the following: widespread abuse, misuse, or overuse of antibiotics; the frequent renewal of antibiotics; increasing quantity and variety of pathogens present in the ICU; the diagnosis and progression of critical illnesses; postsurgical recovery; the use of invasive medical devices; and prolonged stay in the ICU [22, 23]. Patients should be monitored for drug resistance within S. aureus so as to provide a basis for determining appropriate therapy. Meanwhile, hospitals are expected to allocate dedicated ventilators, oxygen, transfusion system, sphygmomanometer and thermometer to each patient and to disinfest the equipment after being used by each person. It is necessary to train the medical staff to infection prevention since researches verify that the hands of the medical staff are an important medium to spread MRSA, which means that hand washing is a vital way to block the spread of MRSA [22, 23]. Also, attention should be paid to ventilation and air purification. Moreover, antibiotics should be used carefully in clinical applications to prevent drug-resistant strains along with disinfection and proper isolation when indicated in order to reduce cross infection of S. aureus [24].

This study has a few limitations. First, genotypic or molecular data including Panton-Valentine leukocidin genes among all unidentified strains were not determined. Second, the antibiotic sensitivity of MRSA isolates to Daptomycin was not included. Future research may consider focusing on genetic types and the mechanisms for transmission.

The findings of this study indicate that the MRSA isolates are circulating in the ICU at obstetrics & gynaecology departments and constitute a major source of infection at a large hospital in China. This study also found that vancomycin may be a reasonable choice in the treatment of MRSA isolates. There is a strong need for increased hospital-wide surveillance and the development of adequate infection prevention strategies.

ICU, intensive care unit; CLSI, the Clinical and Laboratory Standards Institute; HAIs, healthcare-associated infections; MRSA, methicillin-resistant Staphylococcus aureus.

ZHH conceived, designed the experiments and wrote a draft manuscript. BXL and MCL analyzed, interpreted the results of the experiments and revised the manuscript. PYL performed the experiments. XM collected the clinical data. All authors read and approved the final manuscript.

Ethical approval for collecting clinical samples was received by the institutional ethics committees of the participating hospital. Informed consent forms were reviewed and signed by all participants before samples collection (Ethical approval number: Protocol Number 2019-01-02).

We thank three anonymous reviewers for excellent criticism of the article.

The present project was a part of National Natural Science Foundation of China project (81402979), the Jilin Science and Technology Development Program (20190304101YY and 20190301014NY); the Health and Family Planning Commission of Jilin Province (2018J098) also supported the project.

The authors declare no conflict of interest.