New research by an international team of scientists raises questions about the recommended treatment for neonatal sepsis in low- and middle-income countries (LMICs).
Published last week in The Lancet Infectious Diseases, the analysis of newborns diagnosed with sepsis in seven LMICs across Africa and South Asia found extremely high rates of resistance to ampicillin and gentamicin, the combination therapy recommended by the World Health Organization (WHO) for empiric treatment of neonatal sepsis. The study also found that children treated with a different combination therapy—ceftazidime and amikacin—had lower death rates.
"Our data have immediate implications at every level of health care and suggest that WHO might need to revise their antibiotic guidelines for neonatal sepsis within LMICs, where antibiotic resistance to currently recommended treatments is extremely high," the study authors wrote.
Higher mortality, higher resistance
The findings are from a sub-study of the larger BARNARDS (Burden of Antibiotic Resistance in Neonates from Developing Societies) study, a Bill & Melinda Gates Foundation–funded observational study that aimed to assess the common sepsis-causing pathogens in newborns, along with their antibiotic resistance profiles.
Sepsis is a significant threat for young children in LMICs. The Global Burden of Disease study published last year found that sepsis caused an estimated 2.9 million deaths in children under 5 in 2017, with the highest incidence and mortality in newborns and the biggest impact seen in LMICs in sub-Saharan Africa and Asia.
Because LMICs have limited access to lab facilities for assessing sepsis-causing pathogens and antibiotic resistance levels, antibiotic treatment is often empiric. But the main BARNARDS study, conducted from Nov 12, 2015, to Feb 2, 2018, found high rates of resistance to ampicillin and gentamicin, so the researchers leading this study set out to examine the effectiveness of the combination therapy and potential alternative treatments.
The study analyzed the outcomes for the most common antimicrobial treatments in newborns (up to 60 days old) with suspected sepsis at 12 clinical sites in Bangladesh, Ethiopia, India, Nigeria, Pakistan, Rwanda, and South Africa. In addition, blood samples were collected from newborns who showed clinical signs of sepsis, and antibiotic susceptibility was determined for bacteria isolated from clinically confirmed sepsis cases.
Of the 36,285 newborns enrolled in the main study, 9,784 had clinically diagnosed sepsis, 5,749 had antibiotic data available, and 2,483 had culture-confirmed sepsis. The four most common antibiotic treatments were ampicillin-gentamicin, ceftazidime-amikacin, piperacillin-tazobactam and amikacin, and amoxicillin-clavulanate and amikacin.
Assessment of 476 antibiotic prescriptions for 446 newborns showed that reported mortality for those treated with ceftazidime-amikacin was lower than for those treated with ampicillin-gentamicin (9.3% vs 16.2%; adjusted hazard ratio, 0.32; 95% confidence interval [CI], 0.14 to 0.72).
The authors note that the mortality effect may have been influenced by country-specific factors. But analysis of the 390 gram-negative bacterial isolates from culture-confirmed cases found that 97.2% were resistant to ampicillin and 70.3% were resistant to gentamicin, with similar levels of resistance across all sites.
Only 28.5% of the gram-negative isolates were susceptible to at least one of the antibiotics in the ampicillin-gentamicin combination therapy, compared with 77.2% for ceftazidime-amikacin, 73.3% for amoxicillin-clavulanate-amikacin, and 80% for piperacillin-tazobactam-amikacin.
"The results in this study suggest that ceftazidime–amikacin might be an effective potential alternative to ampicillin–gentamicin in LMICs," the authors wrote. "Further studies are urgently required to evaluate alternatives on neonatal outcomes."
Cost might be a factor
One of the problems with the WHO's empiric treatment recommendations for neonatal sepsis, the authors say, is that they are based on data from high-income nations, because the data from LMICs is so scarce. But while ampicillin remains effective against pathogens commonly found in wealthier nations, such as Group B Streptococci and Listeria, those species were not found in the blood samples in the study.
"While this may be a suitable empirical treatment for neonatal sepsis in high income countries, this data showcases that it is not an effective option for LMICs, who have different common pathogens and vastly increased resistance against these antibiotics," lead study author Kathryn Thomson, MSc, of Cardiff University and the Ineos Oxford Institute of Antimicrobial Research, said in a University of Oxford press release.
But cost may also be a factor influencing antibiotic choice. The study shows that ampicillin and gentamicin are the cheapest options among those used to treat neonatal sepsis in the countries studied, and only sites in India and South Africa cover the cost of second- and third-line antibiotics. Sites in six of the seven countries, except South Africa, said that the cost of the antibiotics would influence treatment.
The authors of an accompanying commentary say cost should be among the factors analyzed as researchers try to determine the most effective antibiotic treatment for neonatal sepsis.
"The findings from the BARNARDS study call for randomised trials comparing mortality benefit and cost efficiency of different antibiotic combinations and management algorithms to safely reduce unnecessary antibiotic exposure for neonatal sepsis," they wrote.