Fig 1.
Flow chart of enrolled neonates with physician diagnosed suspected and protocol defined sepsis.
Table 1.
Characteristics of infants meeting case definition* for protocol-defined early-onset sepsis (n = 1,231)—SANISA Study: Soweto, South Africa.
Table 2.
Blood culture isolates by pathogen among neonates with protocol defined early-onset sepsis* (n = 1,231)—SANISA Study, Soweto, South Africa.
Table 3.
Pathogens detected using Taqman Array Card among neonates with protocol defined sepsis* who had blood (N = 933) and respiratory specimens (N = 1204) collected.
Table 4.
Concordance between blood culture and Taqman Array Card (TAC) for detecting bacteria in neonates with early-onset sepsis†.
Table 5.
Pathogen detection using Taqman Array Card among healthy infants who had blood (n = 304) and respiratory (n = 303) specimens collected.
Fig 2.
Estimated proportion of total cases caused by each pathogen with 95% confidence interval; Legend: *Pathogen proportion estimated indirectly from the “other blood culture” class.
Footnote: 73.3% (68.2–77.2%) of case episodes could not be attributed to an evaluated pathogen class.
Table 6.
Estimated true positive rate (TPR) and average false positive rate (FPR) in percentages by pathogen and testing method.
Table 7.
Observed and estimated pathogen specific incidences per 1,000 live births (with 95% confidence limits) of sepsis* overall and by HIV exposure.