Community management of neonatal infections

www.thelancet.com Published online April 2, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60204-5 1 The substantial reduction of mortality in children younger than 5 years during the past decade is one of the most notable recent achievements in global health. The total number of deaths among children in this age group decreased from 9·88 million in 2000 to 6·28 million in 2013. However, the reduction in neonatal mortality during the same period has been less impressive. Neonatal mortality decreased at an annual rate of 2·9% compared with 4·9% in children aged 1–59 months. This comparatively small decrease has contributed to the global failure to achieve Millennium Development Goal 4. Severe bacterial infection (ie, sepsis, pneumonia, and meningitis) in neonates is an important cause of child morbidity and mortality. Estimates suggest that, in 2012, 6·9 million such cases occurred and 557 000 neonates died as a result. Furthermore, the risk of impairment in survivors is high. Presentation is typically with nonspecifi c symptoms and signs that suggest severe disease, and clinical distinction between sepsis, pneumonia, and meningitis is very diffi cult. In resource-poor settings, many cases never reach a health facility. Thus, treatment of young infants with suspected severe bacterial infection in developing countries has been based on clinical signs. Clinical approaches to identify and manage these young infants, such as WHO’s Integrated Management of Childhood Illnesses (IMCI), have deemed these children to have possible severe bacterial infection, and traditionally targeted the fi rst point of contact with the health system—ie, fi rst-level trained health workers. Challenges exist in the diagnosis of young infants with severe bacterial infections. Bacteriological tests have poor sensitivity and most studies of causation are from tertiary care settings, which are not truly representative of cases in the general population. Thus, data for common bacterial pathogens and their antimicrobial resistance patterns are scarce at the community level. A 2013 systematic review of 13 studies from developing countries identifi ed Staphylococcus aureus, Klebsiella spp, and Escherichia coli in 55% (39–70%) of bacteraemic specimens, and reported that only 57% of isolates were susceptible to recommended antibiotics. The two African Neonatal Sepsis Trial (AFRINEST) studies in The Lancet, from diverse settings in east, central, and west Africa, are important and underscore challenges associated with management of children with possible severe bacterial infection in resource-scarce settings. Existing WHO guidelines recommend hospital admission and procaine benzylpenicillin (or ampicillin) and gentamicin as fi rst-line antibiotics. However, access to hospitals is often restricted, and parents can be unwilling to accept hospital treatment and adhere to treatment regimens that include injectable antibiotics. The results of previous studies in rural India, Bangladesh, and Nepal have shown that community-based case management of neonatal infection by trained health workers can substantially decrease neonatal mortality. Eff ective community-based treatment of possible severe bacterial infection with injectable antibiotics might not be feasible in settings that are remote, have inadequate numbers of community health workers, or where health workers are not licensed to provide treatment with injectable antibiotics. Therefore, community-based studies that help to identify eff ective simplifi ed treatment regimens, which can help improve coverage and adherence, are laudable and could have important programmatic implications. The new studies investigate two subgroups within possible severe bacterial infection—infants with mild disease (fast breathing alone); and infants with severe but non-critical disease (ie, poor feeding, lethargy, temperature ≥38°C, and severe chest wall indrawing, with or without fast breathing). The primary outcome in both community-based studies was treatment failure by day 8 after enrolment. In one of the AFRINEST studies, Antoinette Tshefu and colleagues compared oral amoxicillin twice per day with intramuscular gentamicin and procaine benzylpenicillin once per day in 2333 young infants aged 0–59 days. This study was an open-label equivalence trial with individual randomisation, and the authors conclude that young infants with fast breathing alone can be eff ectively treated in outpatient settings when referral to a hospital or hospital admission is not possible. 234 (22%) infants in the injectable gentamicin and procaine benzylpenicillin group failed treatment compared with 221 (19%) in the oral amoxicillin group (risk diff erence –2·6%, 95% CI –6·0 to 0·8). Very few deaths occurred in either group (four [<1%] infants died in the intramuscular gentamicin and procaine benzylpenicillin group, and two [<1%] in the Community management of neonatal infections