The family is (still) the patient.

摘要

IN THIS ISSUE OF ARCHIVES, SlLVERMAN ET AL PRE-sent data from the 2004 Bangladesh Demographic Health Survey, revealing that more than 40% of Bangladeshi mothers with children younger than 5 years of age experience intimate partner violence (IPV) and that their children sustain significantly elevated rates of 2 leading causes of mortality in Bangladeshi children, respiratory infection and diarrhea. Although 2 in 5 mothers may seem to be an extraordinary prevalence of IPV to some, it is a value com- mensurate with those in neighboring Asian countries and in some American states.3 In the current study, younger Bangladeshi mothers and the mothers of very young children were more likely to have been assaulted, and IPV was more prevalent among households with lower levels of maternal education and without sanitary resources. Importantly, ascertainment of IPV was based on husbands' reports of violent incidents during the past year, while mothers independently described children's illnesses during the 2 weeks prior to the study interview. Although men's underdisclosure of IPV could have biased the reported association, such bias would operate in a conservative direction, producing an overestima-tion of childhood illness rates among families without IPV and obscuring the IPV-related differences. Furthermore, the association persisted after adjustment for many possible confounders, such as maternal age, education, household wealth and size,-number of children, sanitation, cooking fuels, and crowding.