Epidemiology and clinical presentation of community-acquired Staphylococcus aureus bacteraemia in children under 5 years of age admitted to the Manhiça District Hospital, Mozambique, 2001–2019

Staphylococcus aureus bacteraemia (SAB) is one of the most common bloodstream infections globally. Data on the burden and epidemiology of community-acquired SAB in low-income countries are scarce but needed to define preventive and management strategies. Blood samples were collected from children < 5 years of age with fever or severe disease admitted to the Manhiça District Hospital for bacterial isolation, including S. aureus. Between 2001 and 2019, 7.6% (3,197/41,891) of children had bacteraemia, of which 12.3% corresponded to SAB. The overall incidence of SAB was 56.1 episodes/100,000 children-years at risk (CYAR), being highest among neonates (589.8 episodes/100,000 CYAR). SAB declined significantly between 2001 and 2019 (322.1 to 12.5 episodes/100,000 CYAR). In-hospital mortality by SAB was 9.3% (31/332), and significantly associated with infections by multidrug-resistant (MDR) strains (14.7%, 11/75 vs. 6.9%, 14/204 among non-MDR, p = 0.043) and methicillin-resistant S. aureus (33.3%, 5/15 vs. 7.6%, 20/264 among methicillin-susceptible S. aureus, p = 0.006). Despite the declining rates of SAB, this disease remains an important cause of death among children admitted to MDH, possibly in relation to the resistance to the first line of empirical treatment in use in our setting, suggesting an urgent need to review current policy recommendations.


Background
Staphylococcus aureus bacteraemia (SAB) is one of the most common bloodstream infections worldwide [1,2]. In 2019, S. aureus was amongst the top three pathogens responsible for global deaths associated with antimicrobial resistance (AMR), with methicillin-resistant S. aureus (MRSA) causing more than 100,000 annual deaths [3]. While SAB epidemiology is well described in high-income countries [2,4], data from low-income countries, particularly in sub-Saharan Africa, remain scarce [5,6]. This study aims to describe the burden, epidemiological trends and clinical presentation of community-acquired SAB among children aged < 5 years in southern Mozambique, between 2001 and 2019.

Material and methods
The study was conducted at the Manhiça District Hospital (MDH) by the Centro de Investigação em Saúde de Manhiça (CISM). The Manhiça district, a rural area in southern Mozambique, has a subtropical climate, with a warm and rainy season (November to April) followed by a cool and drier season [7]. The geographical and socio-demographic characteristics of the study community are described elsewhere [7,8]. CISM's health and demographic surveillance system (HDSS) documents vital events and migrations in the Manhiça district (2,380 km 2 ) with an estimated population of 201,845 inhabitants (27,560 children aged < 5 years) living in 46,441 households [8]. Since 1997, the CISM and MDH operate a 24h morbidity surveillance, which includes the collection of clinical data of all paediatric patients and a systematic collection of a single venous blood sample for bacterial isolation upon hospital admission for all children aged < 2 years, and for children aged 2 and < 15 years with axillary temperature ≥ 39 ºC or with signs of severe illness [9]. All the data analysed were obtained from the microbiological and morbidity surveillance databases of the study area [7][8][9]. Minimum community-based incidence rates of SAB and 95% CIs were calculated considering individual time at risk for children residing in the CISM study area excluding periods of migration. In calculating person-time, individuals were excluded during a lag period of 15 days after each episode of community-acquired bacteraemia. Negative binomial regression models were estimated to compare incidence rates. Score test for trend of rates with calendar year was assessed by Mantel-Haenszel type method. Logistic regression models were used to compare the prevalence of SAB among all admitted children. Wilcoxon rank sum tests were used for nonparametric comparisons. Nutritional status was assessed using weight-for-age z scores, calculated using the LMS method and the 2000 CDC growth reference charts [10]. In-hospital case fatality rate (CFR) due to SAB was calculated for children with known outcome, excluding patients that left the hospital without medical permission (absconders) or those transferred to Maputo Central Hospital (MCH). Data on antibiotic resistance-MRSA and multidrug-resistant (MDR) phenotypes of this S. aureus collection [11] were compared with clinical and epidemiological features using χ 2 or Fisher's exact test, considering a significance level of 5% (STATA version 17, StataCorp LP, Texas, USA).

Minimum community-based incidence rates
Of the 394 SAB episodes, only 39.6% occurred in children living within the HDSS area, yielding an overall minimum community-based incidence of 56.1 episodes (95% CI, 47.9-65.6) per 100,000 children-years at risk (CYAR

Discussion
We described the epidemiological and clinical data of two decades of surveillance of community-acquired SAB among children admitted at MDH, showing a decline on SAB incidence in our setting. However, this incidence could be underestimated considering that only 39.6% of the cases were from the study area and identified through passive detection (severe cases for which health care was sought). This declining trend is similar to those observed for other invasive bacterial pathogens and clinical malaria previously reported in our setting [12][13][14][15]. These downward trends probably reflect the improved socioeconomic and nutritional status, improved water supply, sanitary and overall hygiene conditions, more effective prevention of mother to child HIV transmission, and better access to antiretroviral treatment by patients infected by the HIV. The highest incidence of SAB detected among neonates was previously reported in our setting [9] and recently in Gambia [16], probably reflecting the relatively immature immune responses [17].
The SAB rate observed in our study (12.3%) was similar to the one previously reported in the same setting between 2001 and 2006 [9]. However, this previous study did not evaluate the clinical characteristics and trend of bacteraemia by pathogen nor evaluated its association with malaria. In our study, SAB patients with co-infections by other bacteria were more likely to present a poor outcome, calling for an early and accurate diagnosis for a proper inpatient clinical management. The rate of SAB plus co-infections with other bacteria (3.5%) and the CFR (7.8%) observed were similar to a previous report from South Africa (7% and 8.8%, respectively), a neighbouring country of Mozambique [18]. However, the CFR reported here might be underestimated considering that more than 50% of paediatric deaths in the study area occur at home, with no previous hospital attendance [19]. The high proportion of deaths reported early on admission suggests that the children sought hospital care in critical conditions, or that the infection was caused by isolates resistant to the initial empirical antibiotic therapy. We observed an association between mortality and infection by MRSA and MDR, in addition to the extended LOS among MRSAinfected patients. These findings corroborate with previous reports showing persistent infection and high mortality among patients infected by MRSA compared to MSSA [20,21], likely due to inappropriate empirical antibiotic therapy. Thus, identifying children with risk factors for MDR and MRSA infection will allow better guidance for empiric antibiotic therapy and a likely improved survival. In our study only a single blood culture was performed per patient, hindering the evaluation of infection persistence, although patients' clinical evolution was reassessed every day. The Mozambican national guidelines recommend parenteral combination of penicillin plus gentamicin or chloramphenicol (less prescribed due to its toxicity, despite occasional use in case of antibiotic stock out) for children > 2 years or ampicillin plus gentamicin as empirical antibiotic therapy in younger children or those cases of severe malnutrition. Treatment is reassessed based on clinical evolution and blood culture results using ceftriaxone for infections by MDR strains or severe disease. Our data demonstrated that these antibiotics were the most used in our setting, with ceftriaxone increasingly prescribed throughout the years. We recently reported low rates of SAB caused by MRSA or gentamicin-resistant strains (< 6%), but high for penicillin-resistant strains (90%) in our setting [11]. This resistance may challenge the empirical therapy based on combination of ampicillin/penicillin plus gentamicin, while ceftriaxone may continue as a valuable alternative. Due to the lack of complete information on treatment duration, number of doses of antibiotic given and prescription alterations, we were unable to assess the quality and appropriateness of antimicrobial therapy and relate it to patient's outcome.

Conclusion
In spite of the general declining rates of SAB in our population, this disease remains an important cause of death among children admitted to MDH, possibly in relation to the resistance to the first line of empirical treatment in use in our setting, suggesting an urgent need to review current policy recommendations. Data availability The datasets generated during and/or analysed during the current study are not publicly available to protect our participants' sensitive data but are available from the corresponding author on reasonable request.

Declarations
Ethics approval The morbidity surveillance described in this study is included in the ongoing morbidity surveillance system established as part of the CISM´s health and demographic surveillance system (HDSS) approved by the Institutional Bioethics Committee for Health at CISM, and from the National Bioethics Committee for Health.
Consent to participate All residents in the Manhiça district have signed an individual informed consent to become part of the ongoing HDSS.

Competing interests
The authors have no relevant financial or nonfinancial interests to disclose.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.