Antimicrobial resistance
How are newborns with sepsis treated, and who is most likely to die?
Open access · cc by · source: Europe PMC
Across 11 mostly low- and middle-income countries, newborns with sepsis received hundreds of different antibiotic combinations that often departed from WHO guidance, many infections resisted recommended drugs, and a simple bedside score picked out babies at high risk of death.
Study at a glance
- Design
- Cohort — Prospective observational cohort at 19 hospitals in 11 countries (2018-2020) with daily data on signs, antibiotics and microbiology; Cox models for 28-day mortality derived in a random 85% and validated in the remaining 15%.
- N
- N=3204 · 3,204 infants enrolled; 3,141 started empiric antibiotics; 3,195 had baseline blood cultures; score derivation used 2,726 infants and validation 478.
- Population
- Hospitalised infants under 60 days old with clinical sepsis, mainly in low- and middle-income countries in Asia and Africa.
- Outcome
- Empiric antibiotic regimens and switching, blood-culture pathogens and resistance, and 28-day mortality (plus performance of the NeoSep Severity and Recovery Scores).
Structured fields used in claim comparison tables when every cited study has a complete layer.
Key findings
Infants were started on 206 different antibiotic combinations; only about a quarter received the WHO first-line Access regimen, 18.0% began on a carbapenem, and 25.3% of regimens were later escalated, while de-escalation was rare (5.9%). About 17.7% had a pathogen in the blood, mostly gram-negative bacteria; 32.6% of Klebsiella pneumoniae and 71.4% of Acinetobacter isolates were carbapenem-resistant, and 61.1% of Staphylococcus aureus were MRSA. Overall 11.3% of infants died within 28 days, rising to 17.7% with a positive culture. The Severity Score separated risk well (C-index 0.76 in validation), with death rates of 1.6%, 11.0% and 27.3% in low, medium and high score groups, and it outperformed a score based on WHO danger signs (0.63).
Methodology
Hospitals in 11 countries enrolled 3,204 infants under 60 days old with clinically suspected sepsis and followed them daily for 28 days, recording clinical signs, supportive care, every antibiotic given, blood and spinal fluid culture results and survival. The authors grouped the starting antibiotic regimens by the WHO AWaRe categories (Access, Watch, Reserve) and tracked escalation and de-escalation. They then built a baseline NeoSep Severity Score for 28-day death and a daily NeoSep Recovery Score, developing them in a random 85% of infants and testing them in the other 15%.
Limitations
Because antibiotic choice depended on how sick infants were and on local resources, this observational study cannot say which regimen works best; the authors stress that randomized trials are needed. Sites were mostly urban secondary or tertiary hospitals with good microbiology, so resistance and antibiotic choice may differ in district hospitals, and not all eligible infants were screened, so the sickest babies who died quickly may be under-represented. The scores were validated only internally on a 15% sample containing 42 deaths, so external validation is needed. The provided text also omits most of the Methods section, so enrolment criteria and statistical details are only available in summary form.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Last-line drugs are failing in first-line situations.
Carbapenem resistance in Gram-negative bloodstream isolates is already common in several low- and middle-income hospital settings: 39.3% of Gram-negative isolates in a single-hospital adult oncology cohort of 414 bloodstream infections (rising each year), and 32.6% of Klebsiella pneumoniae and 71.4% of Acinetobacter isolates in the NeoOBS neonatal sepsis cohort.
Evidence for the claim as stated.
Broad empiric use is itself a driver of the resistance it responds to.
Empiric prescribing often drifts to broad drugs and rarely steps back down. In NeoOBS, 3,204 infants received 206 different antibiotic combinations, only about a quarter got the WHO first-line Access regimen, 18.0% started on a carbapenem, 25.3% were escalated and only 5.9% de-escalated.
Evidence for the claim as stated.
History
When this study was placed
Dated entries from the concept change log — when this paper was added or removed as support, challenge, or qualifier on a claim.
Placed as supporting evidence on Antimicrobial Resistance
Carbapenem resistance in Gram-negative bloodstream isolates is already common in several low- and middle-income hospital settings: 39.3% of Gram-negative isolates in a single-hospital adult oncology cohort of 414 bloodstream infections (rising each year), and 32.6% of Klebsiella pneumoniae and 71.4% of Acinetobacter isolates in the NeoOBS neonatal sepsis cohort.
Placed as supporting evidence on Antimicrobial Resistance
Empiric prescribing often drifts to broad drugs and rarely steps back down. In NeoOBS, 3,204 infants received 206 different antibiotic combinations, only about a quarter got the WHO first-line Access regimen, 18.0% started on a carbapenem, 25.3% were escalated and only 5.9% de-escalated.
Related papers in this topic
Same topic cluster — not a recommendation engine.
- Does antibiotic resistance kill children with blood infections?
- Which bacteria cause blood infections in cancer patients, and resist drugs?
- How common is isoniazid-resistant TB, and which mutations cause it?
- How has childhood shigellosis in Vietnam changed over 14 years?
- Who gets carbapenem-resistant Klebsiella, and does it kill more?