This was a cross-sectional study in the town of Mariental, located southeast of Namibia’s capital city of Windhoek on the B1 national highway. With written consent from their parents/guardians, the population that was screened for nasal staphylococci consisted of randomly chosen healthy learners attending five schools in the Mariental District. The children were divided into two age-groups: 6 - 10 years and 11 - 14 years, and consisted of 126 boys and 146 girls. Sample collection was done during March, September and October 2016. One nasal specimen was obtained from each child by gently rotating a sterile Amies transport medium swab (Labocare^TM, Johannesburg, South Africa) thoroughly around the perimeter of both nostrils. Specimens were kept frozen at −20˚C until transporting them to the University of Namibia’s Biomedical Research Laboratory for processing.

Two commercially obtained reference strains, S. aureus ATCC 25923 (an antibiotic susceptible strain) and S. aureus ATCC 33591 (a multi-drug resistant MRSA strain} (Microbiologics®, St. Cloud, US), as well as 433 staphylococcal nasal isolates from 272 Mariental school children, were used in antibiotic assays. Of these isolates, 352 were S. aureus and 81 were CoNS.

Antibiotic resistance testing on isolates obtained from the nasal specimens was carried out in 2017. The Kirby-Bauer disk diffusion assay was used to determine antibiotic susceptibility/resistance in isolates [4] [5] [6] [7]. Table 1 shows the antibiotics that were used and interpretation of inhibition zones. Staphylococcus aureus ATCC 25923, S. aureus ATCC 33591 (MRSA), 352 nasal S. aureus isolates and 81 nasal CoNS were tested. Resistance towards cefoxitin (30 µg), in other words, an inhibition zone diameter < 22 mm, indicated MRSA or methicillin-resistant CoNS (MRCoNS), while resistance towards three or more different classes of antibiotics was an indication of multi-drug resistance in bacteria. By direct colony suspension, three to five well-isolated colonies from overnight tryptone soy agar (Scharlau Microbiology, Spain) plate cultures were inoculated into 10 ml sterile phosphate-buffered saline pH 6.8 - 7.4 (Skylabs, Johannesburg, SA) and adjusted to 0.5 McFarland standard (1.5 × 10⁸ CFU/ml, absorbance reading 0.08 - 0.13 at 625 nm). Adjusted cultures were then swabbed onto Mueller-Hinton agar (Mast Diagnostics, Merseyside, UK) and left to dry for 5 - 10 minutes at room temperature before dispensing the antibiotic disks (Mast Diagnostics, Merseyside, UK) onto the plates. Plates were incubated at 35˚C for 18 - 20 hours and diameters were measured using a ruler. Isolates were classified as susceptible, resistant or intermediately resistant towards each antibiotic, according to the diameter (in millimetres) of their zones of inhibition. The reference strains S. aureus ATCC 25923 (susceptible) and S. aureus ATCC 33591

(multi-drug resistant MRSA) served as quality control strains. Clinical and Laboratory Standards Institute (CLSI) guidelines [6] and the European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoint tables [7] were used to interpret results.

A chi-square test for comparison of proportions with MedCalc statistical software (MedCalc statistical software version 16.4.3 {MedCalc software bvba, Ostend, Belgium; https://www.medcalc.org; 2016}) [8] was used to compare percentage differences in antibiotic resistance between S. aureus and CoNS. Statistical significant differences were indicated by a P-value of ≤ 0.05.

Altogether 433 staphylococcal isolates, as well as two reference strains, underwent antibiotic susceptibility testing against seven antibiotics (Table 1). Excluding the reference strains, 352 of these were S. aureus and 81 were CoNS. Only 4.8% of isolates were susceptible to all antibiotics tested. As expected, S. aureus ATCC 25923 was susceptible to all antibiotics, while S. aureus ATCC 33591 (MRSA) was resistant to cefoxitin and also multi-drug resistant.

Table 2 summarizes percentage susceptibility, intermediacy or resistance of the 352 S. aureus isolates to the antibiotics. Most isolates (96.0%) were resistant to ampicillin, rendering this antibiotic mostly ineffective against S. aureus. For CoNS, 66.7% of isolates were also resistant to ampicillin (Table 3). However, resistance towards ampicillin was significantly higher in S. aureus than in CoNS (P < 0.0001). Resistance towards ampicillin has become very common in staphylococci and can be ascribed to the action of the enzyme beta-lactamase which is under plasmid control. Plasmids containing resistance genes can be transferred from one bacterium to another [9]. Other studies also indicated high ampicillin resistance in S. aureus; in northeastern Brazil, for example, de Carvalho et al., 2017 [10] observed 80.0% ampicillin resistance in S. aureus from nasal secretions of children attending public daycare. In our study, ampicillin was still effective against 33.3% of coagulase-negative isolates. Ampicillin should not be the drug of choice for treating infections caused by S. aureus and CoNS, unless used in combination with other drugs.

As indicated in Table 2 and Table 3, respectively, a total of 83 (23.6%) S. aureus isolates and nine (11.1%) CoNS were resistant to erythromycin (P = 0.0064). These resistance rates are not that high and erythromycin is therefore expected to be effective against staphylococci in most instances. In comparison to our study, a study by Mengistu et al., 2013 [11] using Namibia Institute of Pathology (NIP) data showed higher resistance (32.3%) to erythromycin in Staphylococcus isolated from cerebrospinal fluid (CSF) for the period 2009-2012. Another Namibian study undertaken by Iileka et al., 2016 [12] reported a lower

erythromycin percentage resistance (10.2%) than ours for S. aureus isolates obtained from various clinical samples over the time period 2012-2014. According to PathCare Namibia data [13] 11.0% S. aureus bacteria were resistant to erythromycin from 2014 to 2015, which is very close to the 10.2% indicated by Iileka et al., 2016 [12]. In Brazil de Carvalho et al., 2017 [10] observed a higher percentage (32.8%) erythromycin resistance for S. aureus compared to our results. Their percentage resistance is however almost the same as the 32.3% from Mengistu et al., 2013 [11].

In the current study, only 17.0% S. aureus and 8.6% CoNS (P = 0.0595) were resistant to tetracycline, indicating its effectiveness against Staphylococcus. According to Mengistu et al., 2013 [11] 29.6% of staphylococci from CSF displayed resistance towards tetracycline. In agreement with our results, Iileka et al., 2016 [12] reported 17.4% tetracycline resistance in clinical S. aureus strains across the period 2012-2014. The resistance for S. aureus in our study towards this antibiotic is higher than the 4.3% found by de Carvalho et al., 2017 [10] in Brazil.

Rifampicin resistance in our study was relatively low at 18.2% and 17.3% (P = 0.8495) for S. aureus and CoNS isolates, respectively. However, this is somewhat higher than the 7.0% indicated in other Namibian resistance data [13]. Rifampicin resistance in other countries may be higher. In India for instance, Bharathi et al., 2014 [14] observed 64.7% resistance in nasal MRSA from school children. Rifampicin is among the antibiotics commonly used to treat MRSA infections in India [14]. Staphylococci can quickly develop resistance to rifampicin [9].

According to our results, ciprofloxacin and gentamicin were most effective against S. aureus, with 99.7% and 93.2% of isolates that were susceptible to these drugs, respectively. Ciprofloxacin was also the most effective drug against CoNS, with 100.0% susceptibility. Resistance between S. aureus and CoNS for the two antibiotics was not statistically significant (P = 0.6220; P = 0.1886). In line with our results, de Carvalho et al., 2017 [10] observed 92.9% S. aureus susceptibility towards ciprofloxacin in Brazil. In previous Namibian studies Mengistu et al., 2013 [11] reported 19.0% ciprofloxacin resistance and 52.9% gentamicin resistance for Staphylococcus from CSF samples, whereas Iileka et al., 2016 [12] reported a low 4.4% ciprofloxacin resistance for S. aureus clinical isolates. Namibian susceptibility data for 2011-2012 [13] showed S. aureus to be 96.0% susceptible to ciprofloxacin, while data for 2014-2015 indicated 95.0% susceptibility to gentamicin. Based on our findings, ciprofloxacin and gentamicin could be used to treat certain staphylococcal infections in our study population. However, taking into consideration the 52.9% resistance towards gentamicin in CSF Staphylococcus observed by Iileka et al., 2016 [12], care should be taken by medical doctors when prescribing gentamicin for meningitis in Namibia.

In this study, cefoxitin was used to detect methicillin-resistant bacteria. These bacteria are also resistant to all beta-lactam antibiotics. A total of 51/352 (14.5%) S. aureus isolates were resistant to cefoxitin and therefore identified as MRSA, whereas 7/81 (8.6%) of CoNS isolates were MRCoNS. Cefoxitin resistance between S. aureus and CoNS were statistically insignificant (P = 0.1603). Our MRSA prevalence is close to the 13.8% cefoxitin-resistant S. aureus reported by Reta et al., 2015 [15] in a community based cross-sectional study that involved children aged 6 - 12 years from nine primary schools in Bahir Dar Town, Ethiopia. It is also in line with other Namibian data that showed 13.6% MRSA (2010-2014) and 13.5% MRSA (2012-2014) obtained from various clinical specimens by Festus et al., 2016 [16] and Iileka et al., 2016 [12]. Considering these relatively low percentages, MRSA does not seem to be a major problem in Namibia yet. In general, there are not many studies available on methicillin resistance in CoNS. According to [17], MRSA and MRCoNS can be found together in the human nose and have similar antibiotics resistance genes that can be transferred between bacteria. We could not find studies involving nasal antibiotic resistant staphylococci from healthy school children in our neighboring countries (Angola, Zambia, Botswana and South Africa) to compare our results with.

Altogether 31 antibiotic resistance patterns were observed in this study (Table 4). Isolates with similar resistance patterns could be considered the same strain, unless they acquired these resistance genes from other strains, therefore sharing the same resistance pattern. For S. aureus, 27 different resistance patterns were obtained. For CoNS, there were 14 different patterns. Ten patterns (AP, RP,

Key: Multi-drug resistant ; Multi-drug resistant MRSA ; AP—ampicillin (10 µg); CIP—ciprofloxacin (5 µg); (E—erythromycin (15 µg); FOX—cefoxitin (30 µg); GM—gentamicin (10 µg); RP—rifampicin (5 µg); T—tetracycline (30 µg).

AP-E, AP-GM, AP-RP, AP-T, AP-GM-RP, AP-RP-FOX, AP-T-E and AP-T-RP-E) were shared by S. aureus and CoNS. AP-RP-FOX was the only pattern shared between MRSA and MRCoNS. The three most frequently encountered patterns were AP, AP-E and AP-T, with 194, 52 and 23 isolates having these, respectively.

Fifty-four out of 433 isolates (12.5%) were resistant against three or more classes of antibiotics and classified as multi-drug resistant. Of these, 50 isolates were S. aureus and four were CoNS. The most common multi-drug resistant pattern for methicillin-susceptible isolates was AP-T-E, displayed by eight S. aureus isolates, and one coagulase-negative isolate. Of the 51 MRSA isolates, 22 (43.1%) were multi-drug resistant, with AP-T-RP-FOX as the most encountered resistance pattern among them (See Figure 1). This is of concern, but is 14.5% less than the 57.6% multi-drug resistant MRSA isolated from Iranian children by Erami et al., 2014 [18], and 20.5% less than the 63.6% from Brazilian children as reported by de Carvalho et al., 2017 [10]. One of the multi-drug resistant MRSA isolates in our study showed resistance towards 6/7 antibiotics tested with only ciprofloxacin that was effective against it. Methicillin-resistant CoNS did not show multi-drug resistance, with the most common resistance pattern being AP-RP-FOX. Overall, our findings support the conclusion by Arali et al., 2016 [1] that healthy school children under the age of 16 years are potential carriers of MRSA and multi-drug resistant strains.