Case-fatality and sequelae following acute bacterial meningitis in South Africa, 2016 through 2020.

OBJECTIVES: Providing country-specific estimates of case fatality and sequelae from bacterial meningitis (BM) is important to evaluate and monitor progress toward the World Health Organization's roadmap to "defeating meningitis by 2030". METHODS: From 2016-2020, GERMS-SA conducted enhanced surveillance at 26 hospitals across South Africa. Episodes of laboratory-confirmed BM due to Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis were included. Risk factors for in-hospital death and sequelae at hospital discharge among survivors were analyzed. RESULTS: Of 12,717 invasive bacterial infections reported nationally, 39% (4980) were from enhanced surveillance sites, including 4159 pneumococcal, 640 H. influenzae, and 181 meningococcal infections. BM accounted for 32% (1319/4159) of pneumococcal, 21% (136/640) of H. influenzae, and 83% (151/181) of meningococcal invasive diseases. Clinical data were available for 91% (1455/1606) of BM: 26% (376/1455) were aged <5 years, 50% (726/1455) were female, and 62% (723/1171) with known HIV results, were HIV-infected. In-hospital case fatality was 37% (534/1455), and 24% (222/921) of survivors had adverse sequelae. Risk factors for death included altered mental status, HIV infection, and comorbidities. Risk factors for adverse sequelae included altered mental status and antimicrobial nonsusceptibility. CONCLUSION: BM in South Africa has a high case fatality, and adverse sequelae frequently occur among survivors. Those with comorbidities (including HIV) are at the highest risk.

Epidemiology of Culture-confirmed Candidemia Among Hospitalized Children in South Africa, 2012-2017.

BACKGROUND: We aimed to describe the epidemiology of candidemia among children in South Africa. METHODS: We conducted laboratory-based surveillance among neonates (≤28 days), infants (29 days to <1 year), children (1-11 years) and adolescents (12-17 years) with Candida species cultured from blood during 2012-2017. Identification and antifungal susceptibility of viable isolates were performed at a reference laboratory. We used multivariable logistic regression to determine the association between Candida parapsilosis candidemia and 30-day mortality among neonates. RESULTS: Of 2996 cases, neonates accounted for 49% (n = 1478), infants for 27% (n = 806), children for 20% (n = 589) and adolescents for 4% (n = 123). The incidence risk at tertiary public sector hospitals was 5.3 cases per 1000 pediatric admissions (range 0.39-119.1). Among 2943 cases with single-species infections, C. parapsilosis (42%) and Candida albicans (36%) were most common. Candida auris was among the 5 common species with an overall prevalence of 3% (n = 47). Fluconazole resistance was more common among C. parapsilosis (55% [724/1324]) versus other species (19% [334/1737]) (P < 0.001). Of those with known treatment (n = 1666), 35% received amphotericin B deoxycholate alone, 32% fluconazole alone and 30% amphotericin B deoxycholate with fluconazole. The overall 30-day in-hospital mortality was 38% (n = 586) and was highest among neonates (43% [323/752]) and adolescents (43% [28/65]). Compared with infection with other species, C. parapsilosis infection was associated with a reduced mortality among neonates (adjusted odds ratio 0.41, 95% confidence interval: 0.22-0.75, P = 0.004). CONCLUSIONS: Candidemia in this setting mainly affected neonates and infants and was characterized by fluconazole-resistant C. parapsilosis with no increased risk of death.

Bacterial infections in HIV-infected children admitted with severe acute malnutrition in Durban, South Africa.

BACKGROUND: Bacterial infections in HIV-infected children admitted with severe acute malnutrition (SAM) contribute to higher mortality and poorer outcomes. This study describes the spectrum of bacterial infections in antiretroviral treatment (ART)-naïve, HIV-infected children admitted with SAM. METHODS: Between July 2012 and February 2015, 82 children were prospectively enrolled in the King Edward VIII Hospital, Durban. Specimens obtained on and during admission for microbiological evaluation, if clinically indicated, included blood, urine (obtained by catheterisation or suprapubic aspiration), induced sputum and cerebrospinal fluid. All positive bacterial cultures between admission and 30 days after enrollment were documented and characterised into samples taken either within 2 days of admission (infections on admission) or within 2-30 days of admission (hospital-acquired infections, HAIs). RESULTS: On admission, 67% of patients had abnormal white blood cell counts (WBCC) (>12 or <4 × 109/L) and 70% had elevated CRP; 65% were classified as severely immunosuppressed according to the WHO immunological classification.1 A pathogen was isolated on the admission blood culture in four patients (6%) and in 27% of urine specimens. HAIs were predominately Gram-negative (39/43), and 39.5% were extended-spectrum ß-lactamase-positive. Mortality was not significantly associated with isolation of a bacterial pathogen. CONCLUSIONS: Routine pre-hospital administration of antibiotics as per the Integrated Management of Childhood Illness (IMCI) guidelines may be responsible for the low rates of positive admission blood cultures. HAIs with drug-resistant Gram-negative organisms are an area of concern and strategies to improve the prevention of HAIs in this vulnerable population are urgently needed.