Correlates of infant death from infectious diarrhea in the southeastern United States

Abstract

Infectious diarrhea (ID) caused 242 infant deaths in South Carolina between 1970 and 1978, and is still a significant cause of death in the southeastern United States. Therefore, using South Carolina death certificates from 1970 to 1978 and the state's linked birth/death certificate file, we sought prenatal and postnatal characteristics that might predict death from gastroenteritis in the first year of life. Accuracy of diagnoses was verified from the original death certificates and in a sample of hospital charts. Most deaths caused by ID occurred after the first 28 days of life and were significantly concentrated in the winter. Low birth weight (odds ratio 7.09), nonwhite race (7.00), inadequate prenatal care (4.35), small size of hospital (0.31), and illegitimate birth (2.35) were all significantly associated with postneonatal death. Infants with all of the three strongest predictor characteristics were 14 times as likely to die of ID, but the "predictive value" of this combination for ID death was only 0.54%. Although mothers of these high-risk infants should be given special education about prevention of dehydration, and the families would be logical candidates for a rotavirus vaccine trial, more expensive interventions should not be instituted without a cost-benefit analysis.