A large cluster randomized trial from Bangladesh illustrates the value of case management for neonatal sepsis and some of the problems.20,21 Clusters of pregnant women were randomly allocated to home care, community care or comparison arms. In the home-care arm, female community health workers made antenatal home visits, post-natal home visits and they treated or referred sick neonates. In the community-care arm, qualified providers promoted birth and newborn-care preparedness and care seeking solely through group sessions. In the last 6 months of the 30-month intervention, neonatal mortality was reduced in the home-care arm by 34% (95% CI 7–53) compared with the comparison arm but mortality was not reduced in the community-care arm.20 The case fatality rate for infected neonates was 4.4% if treated by community health workers, 14.2% by qualified medical providers and 28.5% by other unqualified providers or untreated. Compared with neonates who were untreated or treated by unqualified providers, neonates treated by community health workers had a 78% reduction in mortality (95% CI 29–93) and those treated by qualified providers had a 39% reduction (95% CI 1–63). However, referral compliance even for very severe diseases was only 34% and home-treatment acceptance was 43%,21 showing that one problem will be putting effective interventions into practice.22

22.4 Pathogens

The predominant pathogens causing early-onset infections in many developing countries in Africa, Asia, Latin America and the Middle East are Gram-negative enteric bacilli, particularly Klebsiella species and E. coli, and Staphylococcus aureus.23 GBS is uncommon in many developing countries23 but an important cause of early sepsis in some Asian countries.24

There are limited data on home deliveries. The 4 commonest pathogens reported with almost equal frequency from 7–28 days of age are S. aureus, GBS, Streptococcus pneumoniae and non-typhoidal Salmonella species.14

There are very limited data on antibiotic susceptibility of community-acquired pathogens in Africa and Asia.25–27 Available data suggest common pathogens such as S. aureus and Klebsiella have reduced susceptibility to almost all commonly used antibiotics.25 S. pneumoniae are usually susceptible to penicillin but not co-trimoxazole.25 One review of 10 studies found that 72% of E. coli were resistant to ampicillin, 78% to cotrimoxazole (trimethoprim–sulfamethoxazole) and 19% resistant to third generation cephalosporins.26 Klebsiella species were almost 100% resistant to ampicillin, 45% to cotrimoxazole and 66% to third generation cephalosporins.26 Methicillin resistance of S. aureus was rare but 46% were resistant to cotrimoxazole.26 A subsequent study from Kolkata, India reported that over 80% of Gram-negative bacilli causing early-onset sepsis exhibited multi-drug resistance to ampicillin, third generation cephalosporins and gentamicin indicating that these multi-resistant organisms circulate widely in the community. The mortality from Gram-negative sepsis was high.27 A more recent non-Cochrane systematic review and meta-analysis of 19 studies from 13 countries found up to 40% of cases were resistant or had reduced susceptibility to ampicillin/penicillin, gentamicin and third generation cephalosporins.28

22.5 Antibiotics

The appropriate choice of empiric antibiotics depends on the antibiotic susceptibility of local organisms causing neonatal sepsis. This makes it unwise to be prescriptive about empiric antibiotic regimens, which may anyway be dictated in part by cost. As discussed in Chapter 5 (see Section 5.1.3) the WHO currently recommends the use of ampicillin or penicillin G plus gentamicin to treat suspected community-acquired late-onset sepsis in developing countries but up to 40% of infections are caused by organisms resistant or with reduced susceptibility to ampicillin/penicillin and gentamicin and to third generation cephalosporins.28An editorial on combating antibiotic resistance sub-titled ‘The war against error’ advised against one size fits all recommendations and recommended that antibiotic regimens be tailored to local epidemiology. The editorial also stressed that bacteria have evolved resistance genes over millions of years, so escalation to use increasingly broad-spectrum antibiotics to chase resistance is doomed to failure.29

Oral antibiotics are effective in the case management of pneumonia.30 Parenteral antibiotics are needed for serious infections. A review suggested that procaine penicillin G plus gentamicin or ceftriaxone alone were the preferred regimens for serious neonatal infections in the community and primary care because they could be given once daily and were safe and effective.31

There are very few RCTs of antibiotics in neonatal infection.32 An RCT compared failure rates of three clinic-based antibiotic regimens in 0–59 day-old infants with possible serious bacterial infection whose families refused hospitalization in Karachi communities with high neonatal mortality rates >45/1000 live births.33 Eligible infants were randomly assigned to 7 days of: (i) procaine penicillin plus gentamicin or (ii) ceftriaxone alone, or (iii) oral cotrimoxazole (trimethoprim–sulfamethoxazole) plus gentamicin. The primary outcome was treatment failure (defined as death or either, deterioration or no improvement after 2 days on antibiotics). There were 13 of 145 failures (9%) with penicillin–gentamicin, 22 of 145 with ceftriaxone (15.1%) and 26 of 143 with cotrimoxazole–gentamicin (18.2%). Treatment failure was significantly higher with cotrimoxazole–gentamicin compared to penicillin–gentamicin (RR 2.0, 95% CI 1.1–3.8). The difference between the ceftriaxone and penicillin–gentamicin groups approached statistical significance (RR 1.7, 95% CI 0.9–3.2). By 14 days, there were 2 deaths in the penicillin–gentamicin group, 3 in the ceftriaxone group, and 11 in the cotrimoxazole–gentamicin group. The authors concluded that outpatient therapy with injectable antibiotics is an effective option when hospitalization of sick infants is not possible, that procaine penicillin–gentamicin was superior to cotrimoxazole–gentamicin and that ceftriaxone may be less effective.33

Stay updated, free articles. Join our Telegram channel

Join