Letter from Bombay: On editing a book in India

treatment, could reduce deaths among children aged 6-35 months.Even assuming a good individual efficacy, the impact of such interventions on infant and child mortality may be less than the sum of their isolated effects.'Children in areas with a high mortality are exposed to many risks, concurrent and sequential.We must also be concerned with issues of equitable distribution, cost, and operational constraints.The conditions of the experiment in The Gambia, which indicated an im- pressive impact of curative services on infant and maternal mortality, cannot be replicated here.2'While we do not ignore the potential of socioeconomic progress and environmental changes to reduce infant and maternal mortality, we have deliberately focused on health services because they seem less difficult to implement in the short term than socioeconomic reforms.Although our conclusions apply only to the area in which this study was done, we believe that the approach used, despite its exclusive focus on mortality, can help to define the required balance between preventive and curative interventions.