Outpatient treatment for neonates and young infants with clinically suspected severe infection

The reduction in mortality in children younger than 5 years over recent decades has been impressive, but shortfalls still exist in achievement of the MDG4 targets. Acceleration of progress for child survival would need increased emphasis on quality care at and around birth, and timely and effi cacious treatment for neonatal infections, diarrhoea, and pneumonia in regions where these causes contribute substantially to mortality in children younger than 5 years. The challenge in provision of standard care is increased wherever health systems are weak, socioeconomic conditions are suboptimum, and settings are remote and inaccessible. Innovative implementation research is being widely promoted to provide solutions that improve equitable delivery of established interventions without compromising safety or eff ectiveness. This type of research is challenging and often complex. In The Lancet Global Health, Abdullah Baqui and colleagues report their fi ndings from a large equivalence trial in Bangladesh, in which two antibiotic regimens were compared with the standard regimen (intramuscular injections of procaine benzylpenicillin and gentamicin once per day for 7 days) for treatment of clinically suspected serious infection in young infants (aged 0–59 days) who were treated in an outpatient setting after their parents sought help but refused hospital admission. The two alternative regimens assessed were: intramuscular gentamicin once per day and oral amoxicillin twice per day for 7 days; and intramuscular procaine benzylpenicillin and intramuscular gentamicin once per day for 2 days, followed by oral amoxicillin twice per day for 5 days. Thus, compared with the 14 injections in the reference treatment, these regimens had reduced numbers of injections (seven and four, respectively). The primary outcome was a composite measure of treatment failure by day 8 of enrolment based on several individual indicators, and deaths were also carefully recorded. The risk of treatment failure was 8% in each of the two alternative regimens, compared with 10% in the standard treatment group. Risk of death by day 15 of follow-up was 2% in all treatment groups. This result was similar to the number of deaths in infants who opted for hospital admission (2% in hospital and another 1% within one week of discharge). We believe that the equivalence of these regimens for the primary outcome was satisfactorily established in the type of patients enrolled. The case fatality rates were low overall and consistent with the similar failure rates across treatment arms. What are the implications of these fi ndings for child health programmes? In an ideal world, serious infections in neonates and young infants would be treated in medical facilities. However, the barriers to such care are overwhelming for millions of families in Asia and Africa. Facilities can be far away, often the services provided are of poor quality, and families might not be able to aff ord the related direct and indirect costs of care. Therefore, simplifi ed outpatient treatment regimens in settings close to home make for a strong programmatic rationale for treatment of serious infections in young infants whose caregivers are unwilling or unable to accept hospital admission. Widespread acceptance of such a strategy needs rigorous and conclusive evidence for its feasibility, effi cacy, and safety. In this respect, Baqui and colleagues should be complimented for designing and undertaking a high quality study that paves the way for adoption of this approach in health programmes. Some limitations of the study, however, should be noted. The trial was not masked for practical reasons. The trial enrolled fewer infants younger than 7 days than older infants, and this subgroup might diff er in terms of causes and outcomes. The study individuals were young infants whose families sought care but did not accept admission to hospital; the infants in the study might therefore represent a group with milder disease than that of those who accepted hospital admission. In this effi cacy trial, high compliance was achieved for all regimens, which were delivered at home through regular physician visits. Regardless of the regimen, compliance with treatment and monitoring might be better in hospitals than at home in real-life programmes. Our view is that referral and hospital admission should be promoted for sick neonates and young infants. The study results do, however, provide good evidence that when hospital admission is not feasible, and if simplifi ed Lancet Glob Health 2015