The early and late onset GBS incidences in Qatar are comparable to data from different international centres (Table 8). However, we could not find a plausible explanation for the high rate recorded during 2015. In a recently published study from Qatar, Ali et al. (2022) [28] reviewed 196 invasive blood stream infections involving all age groups. They observed an increasing trend from 1.48/100000 population in 2015 to 2.09/100000 population in 2018.
The data also depends on the definition used and the study methodology. Studies with a definition of 72 hours for EOGBSS would naturally miss cases presenting between the fourth and sixth day of life. Single-center-based, NICU-based, and lab-based studies will also likely under-report the incidence. There was no significant decrease in EOGBSS cases after implementing universal maternal GBS screening in 2018 (0.57 vs. 0.52 per 1000 LB, p 0.73). The previous study in Qatar by Abdelmaaboud et al. (2011) [24] also did not observe a significant change in incidence of EOGBSS after shifting to the risk-based approach (2006–2009) from universal screening policy (2003–2005) (0.53 during universal screening vs. 0.57 during risk-based screening, p 0.28). Schrag et al. (2002) [13] and Hasperhoven et al. (2020) [14] have reported a significant reduction in EOGBSS cases after adopting the universal screening strategy. However, this policy is not universally accepted. The RCOG recommends the risk-based approach as a cost-effective strategy for preventing EOGBSS [22]. In short, the strategy should be driven by accurate data obtained through good-quality prospective research conducted in the country concerned.
The incidence of LOGBSS also did not vary significantly over the study period. Except for a relatively higher incidence during 2015, the LOGBSS rate remained between 0.5 and 0.33/1000LB during 2016–2020. The introduction of IAP has not affected the incidence of LOGBSS [16]. The rate of LOGBSS in Qatar is consistent with the incidence worldwide.
Table 8
GBSS incidence: Comparison of the current study to published data.
Authors | Country | Study period | Study method | EOGBSS | LOGBSS | Comments |
Present study | Qatar | 2015–2020 | Retrospective review | 0.58 | 0.33 | • All live-born babies during the study period were screened -Birth registry based |
CDC 2020[9] | USA | 2020 | Prospective surveillance | 0.2 | 0.28 | Definitions: • EOGBSS – 6days or less • LOGBSS- 7 to 90 days |
O’Sullivan et al.,2019[29] | UK | 2014–2015 | Prospective surveillance | 0.57 | 0.37 | |
Sing et al., 2019[30] | Australia | 2002–2012 | Prospective Surveillance | 0.43 | | • Definition EOGBSS: 48 hours, • data collected from centres in major cities, not representative of the more urban population. • There was no significant trend. |
Ding et al., 2023[31] | China | 2016–2022 | Systematic review | 0.28 | 0.13 | • Included 1,314,216 live births in Mainland China. |
Sikias et al., 2022[32] | France | 2019–2021 | Prospective observational study | 0.16 | | • Only 34 weeks and more gestation included, • EOGBSS definition: < 72 hours |
Abdelmaaboud et al., 2011[24] | Qatar | 2003–2009 | Retrospective review | 0.51 | | • EOGBSS definition - <72 hours • Laboratory-based |
Hammoud MS et al., 2017[25] | Middle East (UAE, Kuwait, Saudi Arabia) | 2013–2015 | Multi centre, prospective, observational | UAE: 0.9, Kuwait:2.64 Saudi: 0.4 | | • EOGBSS < 72 hours. • Only cases admitted to NICU were included |
Manroori et al., 2019[26] | Oman | 2006–2016 | Retrospective review | 0.33 | 0.12 | • EOGBSS- up to 6 days • LOGBSS- 7 to 90 days • Only NICU admissions were considered. |
Almudeer et al., 2020[33] | Saudi Arabia | 2011–2018 | Retrospective review | 0.74 | • Overall GBSS, up to 28 days • Only NICU admissions were considered |
CDC – Centers for Disease Control and Prevention, US- United States, UK- United Kingdom, UAE-United Arab Emirates. EOGBSS Early-onset GBS sepsis, LOGBSS Late-Onset GBS Sepsis.
The mean maternal age was 27 ± 5 years. Maternal age below 18 years has been identified as a strong predictor of EOGBSS [34]. Thirty-five percent of mothers were of Qatari nationality, and the remaining patients belonged to 18 countries. Ethnicity and culture are important determinants of GBS carrier state. Mothers of African origin have a significantly increased colonisation rate and EOGBSS [35].
Maternal carriage of GBS is a prerequisite for neonatal infection. Prematurity, prolonged rupture of membranes (> 18 hours), chorioamnionitis, and intrapartum fever are well-known risk factors for EOGBSS [35]. In this study, 72 babies were diagnosed with EOGBSS. The observed risk factors included prematurity (27.6%), clinical chorioamnionitis (11.5%), intrapartum fever (11.5%), premature rupture of membranes > 18 hours (16.8%), and GBS colonisation (58% of 22 mothers who were tested). In a population-based study involving a large cohort of 377 infants with EOGBS, Heath et al. (2004) [36] observed that 42% had no significant risk factors for EOGBSS. In our study, nearly half of the mothers who were screened for GBS were negative, although all tests were performed within five weeks of delivery. In a systematic review, Valkenburg-van et al. (2010) [37] observed positive and negative predictive values of 43–100% and 80–100%, respectively, with higher accuracy when tested closer to delivery.
The pathogenesis and risk factors of LOGBSS are not well defined [17]. In our study, 34% of mothers who tested for GBS in the LOGBSS group were carriers, which was similar to those with EOGBSS. LOGBSS cases were more likely to be premature (39% vs. 27.6%, p 0.017). Spontaneous rupture of membranes (SROM) and chorioamnionitis were less frequent in LOGBSS. Maternal GBS carrier state increases the risk of neonatal carriage and, thus, LOGBSS [17] Pintye et al. (2016)[38] reported that 40–50% of all LOGBSS occur among premature neonates. Berardi et al. (2013) [17] also suggested cross-infection from mothers as an important source of LOGBSS, with a possibility of transmission through breast milk.
We observed that nearly half of the EOGBSS cases were symptomatic at birth, with 86% presenting within 24 hours and 90% within 48 hours of birth. Only 10% presented between 48 hours and six days. Among the seven babies who presented after 48 hours, five had already been discharged home before the onset of symptoms. Heath et al. (2004) [36] reported that half of EOGBSS cases were symptomatic at birth, and 98% presented within 12 hours of age. In the multi state surveillance from the US, Nanduri et al. (2019) [16] reported that 94.8% were diagnosed within 48 hours of life. In our study, 5 of the seven babies who developed symptoms after 48 hours were already discharged home before they presented. The overall outcome of this small group was not different from the rest of the cohort. In the same multi centre surveillance [16], 0.3/100000 live births were discharged home before presenting with EOGBSS.
Respiratory distress (81.9%) and hypo-activity (48.8%) were the most common presenting symptoms in EOGBSS. Fever was present in 22%. Pneumonia and confirmed meningitis were diagnosed in 52% and 18% of cases, respectively. Four babies (5.5%) remained asymptomatic, and a blood culture was sent due to significant risk factors for EOGBSS. In contrast, Nanduri et al. (2019) [16] observed bacteremia without focus as the presenting type in 83% of EOGBSS and meningitis was observed in 9.5%. We included all documented cases of pneumonia, irrespective of the radiology report. Pneumonia was suspected in any case of EOGBSS requiring respiratory support with suspicious x-ray findings in bedside review. The rate of meningitis is comparable to that reported by Nanduri et al. (2019) [16] Madrid et al. (2017) [15] observed meningitis among 16% of babies with EOGBSS.
LOGBSS cases presented at a median age of 21 days (IQR 15–51 days). Fever (85.4%) and hypo-activity (78%) were the most common presenting symptoms. Eight premature infants, who were still not discharged home after post-delivery hospital admission, presented with apnea, bradycardia, and desaturation as the predominant symptoms. The proportion of confirmed meningitis was higher in LOGBSS (34% vs. 18%), but the difference did not reach statistical significance (p 0.054). Overall, meningitis was diagnosed in 48% of all LOGBSS cases. Two cases of LOGBSS had myositis of the thigh; one had orbital cellulitis, and another three babies presented with associated urinary tract infection. Nanduri et al. (2019) [16] reported a median presenting age of 34 days (IQR 20–49 days), and bacteremia without focus was the most common presentation. GBS was isolated from CSF in 20.7%, meningitis was diagnosed in 31.4% of cases, soft tissue infection was reported in 1.8% and 7% of babies had meningitis without bacteremia [16]. Our study included only cases with bacteremia, and we estimate that 2–3 cases of LOGBSS meningitis without bacteremia might have been missed during the 5-year study period.
In a prospective cohort study involving 311893 births, Berardi et al. (2013) [17] described 100 neonates with LOGBSS. Of them, 57% had bacteremia without focus, 34% had meningitis, and 7% had soft tissue infections. In the systematic review involving 6900 cases of neonatal GBSS, Mardrid et al. (2017) [15] observed meningitis among 43% of LOGBSS cases.
Abnormal laboratory markers, including increased CRP, thrombocytopenia, and metabolic acidosis were observed in similar frequency in EOGBSS and LOGBSS. Leukopenia was more frequent in EOGBSS. Bulkowstein et al. (2016) [39] observed a significantly higher incidence of leukopenia and thrombocytopenia in early-onset sepsis and leukocytosis in late-onset sepsis. However, these are non-specific markers with low sensitivity and specificity in diagnosing sepsis [40, 41]. The level of support needed was compared between EOGBSS and LOGBSS. EOGBSS cases were more likely to need respiratory support (83% vs. 36%, p < 0.001). Other requirements, such as central lines, fluid boluses, inotropes, blood products, and mechanical ventilation, were similar between EOGBSS and LOGBSS.
The overall mortality was 8% (11% among preterm and 5% among term babies). The published studies have reported mortality rates of 2–3% and 1–3% among term babies with EOGBSS and LOGBSS, respectively [16, 19]. Premature babies have considerably higher mortality rates ranging from 20–30% [16, 19, 20]. Yeo et al. (2017) [19] observed that the mortality risk persisted beyond hospital discharge. Madrid et al. reported an overall case fatality rate of 8%. We observed that, after adjusting for gender and type of sepsis, confirmed meningitis (adj OR 5.4, 95% CI 1.04-28), pneumonia (adj OR 9.17, 95% CI 1.76-47), birth weight (adj OR 0.4, 95% CI 0.187–0.88), and peak CRP (adj OR 1.022, 95% CI 1.002–1.025) were associated with a significantly higher odds of the combined outcome of death plus neurodisability.
Survivors of GBSS are at risk of adverse neurodevelopmental outcomes. The risk is significantly higher among extremely premature babies and those with GBS meningitis [20, 42].
The files of all enrolled babies were reviewed for documented follow-ups beyond 18 months of age. For babies with at least three visits to a paediatrician after 18 months of age, documentation was reviewed for any evidence of major neurodisability, defined as cerebral palsy, major hearing/visual deficit, or intellectual disability as documented in the file. Overall, 18% of the babies were documented to have major neurodisability. There was no difference in major neurodisability (9.8% with EOGBSS, 13.5% with LOGBSS, P = 0.58) or combined death and major neurodisability (19% in EOGBSS vs. 15.8% in LOGBSS, P = 0.64) between the two types of sepsis. In a systematic review, Kohli-Lynch et al. (2017) [43] observed neurodevelopmental impairment among 18% of GBS meningitis survivors.
Limitations
There are some inherent limitations to the retrospective design of this study that should be acknowledged. We did not have access to current data on maternal GBS colonisation rates, which could have provided important context for interpreting the results. Our study only included cases with GBS bacteraemia, and therefore, isolated cases of meningitis without bacteraemia may have been missed. Nanduri et al. (2019) [16] have reported that 0.1% of EOGBSS and 7.2% of LOGBSS present as meningitis without bacteremia. Although the missed EOGBSS cases would be negligible, the study may have missed 2–3 cases of LOGBSS during the whole 5-year period. Moreover, Qatar has a significant expatriate population, which increases the possibility of missing cases who travelled outside Qatar before three months of age. Therefore, the incidence of LOGBSS was slightly underestimated. Additionally, relying solely on physician documentation to assess neurodevelopmental outcomes might miss milder cases, leading to underestimation. Despite these limitations, our study provides important insights into the epidemiology and outcomes of GBSS in Qatar.