Correlates of social behavior change communication on care-seeking behaviors for children with fever: an analysis of malaria household survey data from Liberia.

BACKGROUND: In 2010, malaria was responsible for an estimated 41% of deaths among children under the age of five years in Liberia. The same year, the Rebuilding Basic Health Services Project launched "Healthy Baby, Happy Mother," a social and behavior change communication campaign. The campaign encouraged caregivers to take children under the age of five years to a health facility as soon as children developed fever. This study investigated correlates of two case management outcomes: care-seeking for children under five with fever during the past two weeks and administration of an artemisinin-based combination therapy (ACT) the same or next day as fever onset. METHODS: Data from a 2014 cross-sectional household survey from four counties was used to investigate correlates of two case management outcomes. Using multilevel analysis, the association between these outcomes and a caregiver's recall of the campaign, her sociodemographic characteristics, and unmeasured characteristics of the community she lived in was investigated. RESULTS: Caregivers living in Grand Kru County were less likely (OR = 0.21, 95% CI 0.073, 0.632) to take a child to a health facility than those in Bong County. Caregiver recall of the campaign was positively associated with the odds that a child received an ACT promptly (OR 3.62, 95% CI 1.398-9.372), but not with the odds of a caregiver taking a child in their care to a health facility. While unmeasured community-level factors accounted for 19.0% of the variation in the odds that a caregiver's child was brought to a health facility, they did not play a role in the odds of prompt ACT treatment. CONCLUSIONS: Recalling the "Healthy Mother, Happy Baby" campaign was positively associated with the odds that children received ACT promptly, even in the absence of other malaria prevention and treatment messaging. While caregiver exposure was not associated with care-seeking during the two weeks before interview, prompt care-seeking likely preceded prompt receipt of ACT since most ACT came from health facilities. Unmeasured community-level factors, such as distance from the health facility, may play a role in determining the odds that a caregiver takes a child to a health facility.