When first line treatment of neonatal infection is not enough: blood culture and resistance patterns in neonates requiring second line antibiotic therapy in Bangui, Central African Republic.

BACKGROUND: Infectious diseases account for the third most common cause of neonatal deaths. Globally, antibiotic resistance (ABR) has been increasingly challenging neonatal sepsis treatment, with 26 to 84% of gram-negative bacteria resistant to third-generation cephalosporins. In sub-Saharan Africa, limited evidence is available regarding the neonatal microbiology and ABR. To our knowledge, no studies have assessed neonatal bacterial infections and ABR in Central-African Republic (CAR). Therefore, this study aimed to describe the pathogens isolated and their specific ABR among patients with suspected antibiotic-resistant neonatal infection admitted in a CAR neonatal unit. METHODS: This retrospective cohort study included neonates admitted in the neonatal unit in Bangui, CAR, from December 2018 to March 2020, with suspected antibiotic-resistant neonatal infection and subsequent blood culture. We described the frequency of pathogens isolated from blood cultures, their ABR prevalence, and factors associated with fatal outcome. RESULTS: Blood cultures were positive in 33 (26.6%) of 124 patients tested (17.9% for early-onset and 46.3% for late-onset infection; p = 0.002). Gram-negative bacteria were isolated in 87.9% of positive samples; with most frequently isolated bacteria being Klebsiella pneumoniae (39.4%), Escherichia coli (21.2%) and Klebsiella oxytoca (18.2%). All tested bacteria were resistant to ampicillin. Resistance to third-generation cephalosporins was observed in 100% of tested Klebsiella pneumoniae, 83.3% of isolated Klebsiella oxytoca and 50.0% of tested Escherichia coli. None of the tested bacteria were resistant to carbapenems. Approximately 85.7 and 77.8% of gram-negative tested bacteria were resistant to first-line (ampicillin-gentamicin) and second-line (third-generation cephalosporins) treatments, respectively. In hospital mortality, adjusted for blood culture result, presence of asphyxia, birth weight and sex was higher among neonates with positive blood culture (adjusted relative risk [aRR] = 2.32; 95% confidence interval [CI] = 1.17-4.60), male sex (aRR = 2.07; 95% CI = 1.01-4.26), asphyxia (aRR = 2.42; 95% CI = 1.07-5.47) and very low birth weight (1000-1499 g) (aRR = 2.74; 95% CI = 1.3-5.79). CONCLUSION: Overall, 77.8% of confirmed gram-negative neonatal infections could no longer effectively be treated without broad-spectrum antibiotics that are not routinely used in sub-Saharan Africa referral hospitals. Carbapenems should be considered an option in hospitals with surveillance and antibiotic stewardship.

A qualitative awake EEG score for the diagnosis of continuous spike and waves during sleep (CSWS) syndrome in self-limited focal epilepsy (SFE): A case-control study.

PURPOSE: To determine whether awake EEG criteria can differentiate epileptic encephalopathy with continuous spike and waves during sleep (EE-CSWS) at the time of cognitive regression from typical, self-limited focal epilepsy (SFE). METHODS: This retrospective case-control study was based on the analysis of awake EEGs and included 15 patients with EE-CSWS and 15 age-matched and sex-matched patients with typical SFE. The EEGs were anonymised and scored by four independent readers. The following qualitative and quantitative EEG indices were analysed: slow-wave index (SLWI), spike-wave index (SWI), spike-wave frequency (SWF), long spike-wave clusters (CLSW) and EEG score (between grades 0 and 4). Sensitivity and specificity were assessed using receiver operating characteristic (ROC) curves and their reproducibility with a kappa test. RESULTS: Based on a highly sensitive cut-off, EE-CSWS patients were 8.4 times more likely than those with SFE to have an SLWI > 6%, 15 times more likely to have an SWI > 10 % and six times more likely to have a CLSW of ≥ 1 s. There was substantial agreement between readers (with kappa values of 0.64, 0.69 and 0.67). EE-CSWS patients were 13 times more likely to have an SWF of > 11 % and 149 times more likely to have an EEG score of ≥ 3 than typical SFE patients. Agreement about these ratings was almost perfect (kappa 0.91 and 0.86). CONCLUSION: An EEG score of ≥ 3 on a 20-min awake EEG differentiates typical SFE from EE-CSWS at the time of cognitive regression, with good reliability across readers with different levels of expertise.