ABSTRACT
Abstract:Perinatal infection significantly contributes to neonatal morbidity and mortality. There are no reliable rapid diagnostic tests. Drug resistance is increasing in organisms acquired in hospital. There are little data on the indications for antibiotic use and the prevalent organisms in lower resource settings; and none in regional hospitals in South Africa. We conducted a retrospective cohort study of risk factors; indications for and drugs used at Worcester Provincial Hospital Neonatal Unit. A systematic sample of every alternate neonate listed in the admissions register from 1 July 2005 to 30 June 2006 was taken. Charts for all cases were reviewed. Early antibiotic use was defined as therapy within 72 hours of life. One hundred and ninetyfive infants where included; 144 (74) had 194 antibiotic events. Antibiotic events occurred at a rate of 99 events per 100 neonates. Prematurity was common (83 of admissions) and; in conjunction with prolonged rupture of membranes; was the major driver of early antibiotic use. Ceftriaxone use within 72 hours of birth was significantly associated with subsequent antibiotic events; compared with penicillin alone or in combination with an aminoglycoside (p 0.04). Longer duration of treatment for early events was associated with subsequent empiric need for antibiotics (p 0.02). Prematurity is the major driver for antibiotic use at this unit. Antibiotics are prescribed appropriately but earlier discontinuation; which may be complicated by the inability to confirm/refute infection; should be practised. Alternatives to third generation cephalosporins should be available to treat hospital infection at secondary level