The goal of surveillance in public health, to provide information to decrease morbidity and mortality, and to improve health, could be achieved through ongoing systematic collection, analysis, interpretation and dissemination of data regarding public health-related events. The surveillance system for MRSA in the Caribbean appears simple, most data are collected from a single or multiple, regional hospitals, primary health centers in the rural or urban communities with limited complex electronic system that receives and integrates data from the multiple sources. Out of 47 records identified during the literature screening process, an aggregate of 24 peer-reviewed publications met the search criteria, and were included in this review (Table 1), [11] [12] [19] - [38]. The publications, contributed by investigators from 10 countries (37%) of the 27 screened, ranged in dates from 1999 to 2020. The included studies were conducted in Trinidad and Tobago (n = 7), Jamaica (n = 5), Barbados (n = 2), Dominican Republic (n = 2), Martinique (n = 2), Haiti (n = 2), Cuba (n = 2), St. Kits & Nevis (n = 1), Guadeloupe (n = 1) and Guyana (n = 1).

The Caribbean researchers used cohort studies, choosing equally between prospective and retrospective study designs [39] as dictated by their budgets, availability of resources and relative access to patients’ samples. For the prospective study designs, hospital and/or community clinical specimens included high vaginal swabs, urine, skin and soft tissue swabs, surgical and burn wounds, pus/abscess, respiratory tract, blood, bone, nasal cavity swabs and catheters submitted by patients or subjects being investigated for staphylococcal colonization or infections [11] [20] [29] [32] [33] [34]. For the retrospective study designs, investigators conducted cohort analysis of patients admitted to the major hospitals over a given period (years) as part of surveillance programs established to screen surgical and burn wounds, nasal, groin, and axilla for colonization with Staphylococcus aureus and prevalence of MRSA [21] [26] [37]. The main outcomes of interest in the studies were the rate of MRSA in S. aureus in hospital and/or community settings, country or regional capabilities including phenotypic characterization through antimicrobial susceptibility testing, distribution of major MRSA genotypes through various molecular genotyping methods, detection of toxin and antibiotic resistance genes, and infection control practices.

An overview of the peer-reviewed articles that reported on the prevalence of MRSA in the Caribbean is shown in Table 2. The number of S. aureus or MRSA isolated from clinical samples and characterized in the included studies ranged from 16 to 1997. The results show that majority (75%) of the investigations were conducted on cultures of hospitalized patients’ samples [12] [16] [21] [23] [26] [28] [31] [32] [33] [34] [35] [37]. A few studies investigated only cultures obtained from clinical samples obtained from outpatients [24] [29], or from cultures of samples obtained from hospitalized patients and outpatients [11] [21] [27].

The results show that between 12.8% and 60% of Caribbean S. aureus isolates from samples collected from infected, hospitalized patients are MRSA [12] [16] [21] [23] [26] [28] [31] [32] [33] [34] [35] [37]. Interestingly. the prevalence of MRSA differed across the Caribbean countries and territories (Table 2). Even though the mean MRSA prevalence reported in hospitalized patients was characteristically < 21% in Barbados [11], Dominican Republic [25] and Trinidad and Tobago [27], the percentage of MRSA isolates ranged from 39% in Martinique [25], 45% in Jamaica [21] and St Kitts and Nevis [33], to 51% in Guyana [39], and 59% in Cuba [16]. Similarly, while the MRSA prevalence in the samples cultured from outpatients was 20% in the Dominican Republic [25], percentage of MRSA isolates in this patient group were, 39% and 45%, respectively, in Martinique [25], and Trinidad & Tobago [28].

Most of the MRSA isolates from hospitalized patients were associated with surgical wounds, and infections of skin and soft tissue, respiratory tract and urinary tract, age (peak range of 60 - 69 years), gender, ethnicity, duration of hospital stay, co-morbidities such as diabetes mellitus and hypertension, previous penicillin use or previous surgery [11] [12] [16] [21] [22] [25] [27] [31] [34] [35] [39]. However, prevalence of MRSA isolates in clinical samples collected from outpatients, though, associated with gender, diabetes, hypertension or asthma, some considerable proportion of the subjects tended to be healthy, students or those who participated in physical contact sports [11] [30] [33].

Treatment of MRSA infections has remained problematic in the Caribbean region because of the organism’s resistance to many antimicrobial agents. Routine characterization of the clinical MRSA isolates through antimicrobial susceptibility testing and analyses, in the most instances, using the Kirby-Bauer disc diffusion method on Müller-Hinton agar [40] has been reported in approximately 50% of the studies to guide empirical treatment of mild or moderate infections [12] [21] [25] [27] [28] [30] [31] [32] [33] [35] [36]. In a study reported from Trinidad [27], involving MRSA isolates (n = 451) from clinical samples collected from hospitalized patients between 1999 and 2004, all the organisms were fully sensitive to vancomycin, while the greatest resistance was against erythromycin (86.7%) clindamycin (75.3%), tetracycline (78.7%) and ciprofloxacin (59.1%). However, the MRSA strains were less resistant to gentamicin (44.7%), chloramphenicol (17.3%) and trimethoprim-sulfamethoxazole (13%) [27].

In a study reported from Jamaica [21], all MRSA isolates (n = 80) collected from samples of hospitalized and community patients in 2002 were susceptible to vancomycin. Overall, 77.5% of the isolates were resistant to at least one antibiotic, and 10% of isolates were resistant to gentamicin, ciprofloxacin, tetracycline, chloramphenicol, and erythromycin. Sixty percent of the isolates were resistant to penicillin G, 22.5% each to trimethoprim-sulfamethoxazole and oxacillin, and 13.8% to tetracycline. More isolates from hospital sources were resistant to the antimicrobials evaluated (except for gentamicin). Notably, 82% of hospital isolates were resistant to penicillin, compared to 39% of community isolates. Further, all isolates resistant to oxacillin were from hospital sources [21]. In another study of hospitalized patients reported from Jamaica in 2010 [34], MRSA isolates (n = 33), again, showed sensitivity to vancomycin, but variable resistance to erythromycin (94%), clindamycin (52%), gentamicin (33%) and tetracycline (27%). However, resistance to trimethoprim-sulfamethoxazole and minocycline was 12% and 6%, respectively [34].

In a study reported from Barbados [11] antimicrobial susceptibility testing revealed that all hospital-associated MRSA isolates (n = 100) were resistant to ceftriaxone and ciprofloxacin, and 90% of isolates were resistant to erythromycin. All isolates were sensitive to vancomycin, rifampin, gentamicin, linezolid and trimethoprim-sulfamethoxazole; 82% were sensitive to clindamycin, with 2% inducible clindamycin resistance [11]. In the same study from Barbados, of the community-associated MRSA isolates (n = 193) evaluated, nine, or 4.7%, gave D-zones for clindamycin induction. A total of 94.3% susceptibility was recorded to trimethoprim-sulfamethoxazole. A susceptibility to vancomycin of 97.4% was observed; clindamycin susceptibility averaged 89.6%. All isolates were resistant to the β-lactam antibiotics and macrolides. However, for 4.2% of 193 MRSA isolates, four were resistant to vancomycin, whilst one isolate was resistant to cotrimoxazole, ciprofloxacin and vancomycin, and six isolates were resistant to clindamycin [11].

In a study conducted on hospitalized patients and community subjects from St Kitts & Nevis in 2019 [32], the prevalence of MRSA accounted for 46% (70/152) of the isolates. The highest rates of resistance to non-β-lactam agents were observed for daptomycin (97.1%), erythromycin (91.3%), levofloxacin (75.4%), moxifloxacin (73.9%), whereas lower proportions of resistant isolates were seen for tetracycline (10.1%), tobramycin (10.1%), gentamicin (4.3%), fusidic acid (4.3%), clindamycin (2.9), mupirocin (2.9%) and rifampicin (1.4%). All the MRSA isolates were susceptible to ceftaroline, linezolid, teicoplanin, telavancin, trimethoprim/sulfamethoxazole and vancomycin [32].

In the last decade, six molecular typing methods have been utilized in the molecular characterization of MRSA to monitor geographic spread of one or several clones among countries in the Caribbean region [11] [12] [19] [24] [27] - [32]. These molecular techniques included pulsed field gel electrophoresis (PFGE) after SmaI digestion [37], multi-locus sequence typing (MLST) [19] [25] [28], whole genome sequencing-MLST [27], multiplex polymerase chain reaction (PCR) to detect 16SrRNA, mecA, the staphylococcal chromosomal cassette (SCC) mec types, spa types, presence of exotoxins, Panton-Valentine Leukocidin (PVL) and LukAB, and the arginine catabolic mobile element (ACME) [11] [24] [27] [29] [30] [38], microarray hybridization [12] [28], and Multiple Locus Variable-number Tandem Repeat Analysis (MLVA) [27]. Each of these genotyping methods varies regarding the equipment, cost, and expertise required, and their ability to discriminate among related isolates is not the same [10] [41]. Notwithstanding, the sequence-based techniques are specific and sensitive enough to distinguish MRSA strains based on the genes encoding the staphylococcus protein A, the SCCmec types, the PVL, and ACME [41].

Molecular typing of clinical MRSA isolates was established in approximately 60% of the studies reviewed [11] [12] [19] [24] [27] - [32]. Only 15% of the studies reported the genotypes of Methicillin sensitive S. aureus (MSSA) isolates from either hospitalized or community patients [24] [28]. More than a third of the studies did not report on the MRSA genotypes [20] [22] [25] [26] [33] [34] [36]. The clonal results of the MRSA isolates from clinical samples collected from patients in hospital and community settings in the Caribbean region (Table 2), [11] [12] [19] [24] [27] - [32] are described below.

CA-MRSA clones

There was a wide-spread occurrence of the North American epidemic or endemic ST8 MRSA SCCmec IV, PVL positive, (CA-MRSA, USA300) clone in the clinical samples of hospitalized patients studied in several countries including Barbados [11], St Kitts & Nevis [32], Trinidad & Tobago [28] [29], Jamaica [27], Cuba [30], Dominican Republic and Martinique [24]. A similar ST8 MRSA SCCmec IV, PVL positive, prevalence frequency was shown in the clinical samples of outpatients studied in Barbados [11], Trinidad & Tobago [29], Dominican Republic and Martinique [24]. However, there was occurrence of geographical differences in the prevalence of most other MRSA clones in the region. Of particular interest, was the prevalence of CC8-MRSA-IV, “Lyone” clone, CC8-MRSA-IV “UK-EMRSA” clone, and European-CA-MRSA clone in a study of clinical samples of hospitalized patients from Martinique [12]. Another CA-MRSA clone, ST72 SCCmec V was moderately prevalent [25%] in the clinical samples of outpatients reported from Dominican Republic [24], Trinidad & Tobago [29], but rarely observed in outpatients from Barbados [11] and Martinique [24]. ST72 SCCmec V was moderately prevalent [14%] in hospitalized patients studied in Cuba [30] and Haiti [31], of extremely low frequency (1.5% to 2.22%) in Trinidad & Tobago [28], and St Kitts & Nevis [32], but rare in Jamaica [27]. ST30 MRSA SCCmec IV was moderately prevalent (27%) in the samples of outpatients studied in Dominican Republic [24)] but not reported in community patients studied in Martinique [24], Trinidad & Tobago [29], and Barbados [11]. ST30 MRSA SCCmec IV was rare (1.5%) in hospitalized patients reported from St Kitts & Nevis but it was not observed in hospitalized patients’ studies from Jamaica [27], Cuba [30], Haiti [31], and Barbados [11]. Similarly, spa-CC0044-ST80 SCCmec IV was moderately prevalent (13%) in the clinical samples of outpatients studied from Martinique [24] but rare in both hospital and community settings in the other Caribbean countries with published surveillance data (Table 2).

HA-MRSA clones

According to the results shown in Table 2, ST239-MRSA-III, a common cause of hospital-acquired MRSA, was highly prevalent (60%) in the clinical samples of hospitalized patients studied in Trinidad & Tobago [28], but was of considerably low occurrence (1.45% - 3%) in the clinical samples of hospitalized patients reported from Jamaica [27] and Martinique [12]. ST5 MRSA SCCmec II (New York-Japan) clone had low prevalence (3% - 12.5%) in the clinical samples of hospitalized patients in the reports from St Kitts & Nevis [32] and Jamaica [27], and moderately prevalent (18%) in the samples of outpatients from Martinique and Dominican Republic [24]. CC5-ST5-MRSA-SCCmec 1 “Geraldine” clone was of low prevalence (3% - 13%) in the samples of hospitalized patients studied in Jamaica [27], Guadeloupe and Martinique [12].

In a population structure study on several MRSA clones (n = 45) from Trinidad and Tobago in 2014 [28], the common antimicrobial resistance markers reported were the beta-lactamase operon (blaZ/I/R; in 86.39% of isolates), erm (A) (in 9.86%, mostly ST239-MRSA-III), msr (A)/mph (C) (in 8.16% and 7.14%, respectively; mostly associated with “USA300”) and aphA3/sat (in 15.31%, largely associated with ST239-MRSA-III and “USA300”). The gentamicin/tobramycin resistance gene aacA-aphD occurred in 9.52% of isolates that all belonged to ST239-MRSA-III or “USA300”. A gene associated with mupirocin resistance, mupA, was detected in 17.78% of MRSA isolates. Other resistance markers; vanA (vancomycin resistance) and cfr (linezolid resistance) were not found [28].

Similarly, a separate study conducted on outpatients in the rural communities from the same country (Trinidad & Tobago) in 2018 showed the presence of the ermA gene in 31% (5/16) MRSA isolates tested, but none of them tested positive for the ermC and vanA genes, respectively [29]. Majority (62.5%, 10/16) of the MRSA isolates from the outpatients possessed the pvl gene, whereas 25% (4/16) possessed the alpha hemolysin (hla) gene. None of the MRSA isolates possessed the tst1 gene, 18.8% (3/16) possessed both virulence genes, pvl and hla [29]. Nearly all the MRSA (ST8, 88%) isolates from hospitalized patients and community subjects from a study reported in 2019 from St Kitts & Nevis [32] carried genes encoding resistance to streptomycin [ant (6)-Ia], amikacin and other aminoglycosides [aph (3’)-III], fosfomycin (fosD), macrolides, lincosamides and streptogramins [mph (C) and mrs (A)], and penicillin (blaZ). Additionally, genes encoding resistance to trimethoprim (dfrG) and phenicols (cat) were found in four and one isolate, respectively [32].

MRSA infection prevention and control practices are well known, and they include effective personal hygiene, proper wound care, optimum laundry and cleaning or disinfection of high-touch or soiled surfaces [26]. However, the successful application of these practices is dependent on the knowledge, attitudes and practices of health care workers in the local or regional, major hospitals in the Caribbean. In Trinidad and Tobago, lack of effective infection control programs has been associated with poor level of knowledge, attitudes and practices among healthcare workers [42], limited resources, competing priorities, and other barriers [33]. Stringent MRSA infection control measures, on the other hand, appeared to have been set up in Barbados, which accounted for the rare prevalence of HA-associated MRSA infections in the community [11].