At 20.5%, the rectovaginal prevalence of GBS among pregnant women receiving antenatal care at KNH was slightly higher than that reported by prior studies in this region or similar settings. In a study conducted in Mombasa by Cools and Jespers (15), they reported a GBS prevalence of 20.2% while the prevalence has been reported to vary between 10 and 35% in the USA (4) and between 1 and 30% in other developing countries (4, 15–17). Even so, this prevalence is slightly lower than that reported by Salat et al, in a study conducted in the same population (9). Difference of the prevalence in the two studies could be explained by the difference in gestational ages of women who formed study participants of the two studies. Salat restricted his participants to women of gestational age 35 to 37 weeks while this study included participants from gestational ages 12 to 40 weeks. Previous studies (18) have reported gestational age of women as a risk factor to GBS with the prevalence being higher among women of gestational age 35 to 37 weeks (5, 18). This study agrees with this observation as grouping of participants based on gestational age, found those of gestational age 35 to 37 weeks to have higher prevalence (9.6%) of GBS in comparison to other gestational age groups.
Similarly, the prevalence reported in the current study is higher than that reported by Lu, Li (19) in Beijing, China (7%), Woldu, Teklehaimanot (20) in Ethiopia (7.2%) and Seale and Koech (8) in Kilifi (12%) (7, 8, 19). The variations in the prevalence reported in the current study, and that reported by Cools and Jespers (15) and Seale, Koech (8) agrees with findings by Stoll and Schuchat (16) of occurrence of regional variation in GBS prevalence even among people sharing geographical boundaries and with similar socio-economic conditions (8, 15, 16). Similarly, the difference in the prevalence could be explained by the culture method used in identification of GBS. This study used Granada agar in isolating GBS which has higher ability to isolate GBS compared to blood agar plates used by Lu and Li (19). It has been noted that variation in GBS prevalence could be as a result of culture methods and type of medium used to isolate GBS (20).
The antimicrobial susceptibility pattern of isolated GBS showed high GBS resistance to penicillin G and ampicillin of 72.4% and 54.2% respectively. There was however low resistance to clindamycin and vancomycin at 30.4% and 24.1% respectively. Previous studies have reported GBS to be evenly susceptible to penicillin, ampicillin and cephalosporins (17, 21) even though bacteria with increasing minimum inhibitory concentration to penicillin and ampicillin have been reported (17, 21). The high resistance of GBS to penicillin G reported in our study mirrors that reported by Mengist, Zewdie (22) of 77.3% (22). Findings of the current study partly agrees with those of Lu, Li (19), Yoon, Jo (23) and Mengist, Zewdie (22) who reported a resistance to clindamycin of 55.7%, 55.4% and 50% respectively (19, 22, 23). However, the resistance in the three studies is higher than that reported in the current study of 30.4%. Nonetheless, the isolates in Yoon, Jo (23) and Lu, Li (19) studies retained 100% susceptibility to penicillin, ampicillin and vancomycin. On the other hand, the resistance of GBS to clindamycin reported in our study is higher than that reported by Bolukaoto, Monyama (17) of 17.2%, even so, Bolukaoto, Monyama (17) isolates retained 100% susceptibility to penicillin, ampicillin, vancomycin and high level gentamycin even though they had high resistance to tetracycline of 86.7% (17).
Increase in resistance of GBS to β-lactam antibiotics has been attributed to alterations in the penicillin binding protein 2 × (21) while resistance to clindamycin has been attributed to presence of erm-methylase gene (24). In his study to determine antibiotic resistance of GBS among pregnant women in Garankuwa, South African, Bolukaoto, Monyama (17) reported methylation of ermB genes to be the single most common mechanism of resistance employed by isolated bacteria. Other mechanisms identified were efflux pump mediated by mefA genes and ermTR genes. This high resistance limits antibiotic use and restricts treatment to clindamycin and vancomycin.
All the known ten serotypes of GBS were found to be occurring among this study’s population. The most occurring serotype was Ia (75.9%) followed by III (62%), V (56%), VI (54%), Ib (53.7%), VII (53.5%), IX (53.3%), II (48%) and VIII (44.2%). Serotype IV was the least occurring at 36%. 66.7% of participants were found to harbour more than one serotype of GBS. These findings agree with Rench and Baker (25) who reported serotypes Ia, Ib, II and III as the most occurring, similarly, they mirror Lu, Li (19) who isolated eight GBS serotypes among pregnant women in his study with the exception of VII and IX. Nine GBS serotypes (Ia, Ib, II-VIII) have been reported to occur in Europe and USA (26). Previous studies conducted in Kenya have reported the occurrence of six GBS serotypes Ia, III, V, VI, VII and VIII in a Mombasa cohort (15). Our findings agree with observations of Dutra, Alves (27) who reported variations in the regional distribution and occurrence of GBS serotypes as this study found serotype Ia to be the most occurring in this population while Cools et al reported serotype III as the most occurring among pregnant women in Mombasa (15, 27). It is however, important to note that serotype distribution in a population could change with time. Yoon, Jo (23) in a GBS study conducted over a 20 years period found the dominant serotypes to change with time (23).
Serotype Ia (80.5%) and VI (81.5%) had significant resistance to penicillin G while serotypes Ia (61%), III (71%), IV (77.8%), VI (70.5), VII (65.25) and VIII (73.7%) registered significant resistant to ampicillin. Colonization with these serotypes was also a predictor of antimicrobial resistance with being colonised with serotype Ia predicting for penicillin G resistance (OR 4.8; CI: 1.265–18.311, p = 0.021) as was serotype VI (OR 3.38; CI:0.947–12.098, p = 0.061). Being colonized with serotypes III (OR 6.84; CI:1.899–24.672, p = 0.003), IV (OR 5.12; CI:1.372–19.077, p = 0.015), VI (OR 4.45; CI:1.353–14.653, p = 0.014), VII (OR 3.48; CI:0.99-12.242, p = 0.052) and VIII (OR 4.67; CI:1.255–17.358, p = 0.022) was a predictor for ampicillin resistance. Our findings differ from those by Yoon, Jo (23), who in their study, found no strains to be resistant to penicillin (23). They also reported a resistance of 93.8% to clindamycin by serotype V. This level of resistance to clindamycin by serotype V is higher than that reported in our study of 22.7%.
Though some of the participants in this study reported having a history of risk factors associated with GBS such as stillbirths, abortions, ectopic pregnancy, history of preterm births, history of neonatal deaths, history of neonatal infection and history of fore water break more than 18 hours to labour, no statistically significant association was found between these factors and GBS colonization in this study. Similarly, there was no association between maternal age, gestational age, parity and number of live birth with GBS colonization. Other studies have reported an association between GBS colonization and some of these risk factors. For examples Salat (9) had previously reported GBS colonization to be associated with a history of stillbirth in the same population as did Seale, Koech (8) in Kilifi and Doare, Jarju (5) in the Gambia who besides still birth, also noted GBS colonization to be associated with early onset neonatal disease and gestational age (5, 9). Even though the current study did not find any association between GBS colonization and associated risk factors, the role of GBS in causing stillbirths and neonatal sepsis cannot be ruled out given the relatively high GBS prevalence, stillbirths and neonatal sepsis in this setting.