One general hospital requested evaluation of a long-standing problem of conjunctivitis among neonates. The hospital staff believed that the infection was acquired during phototherapy. We investigated all newborns hospitalized from January to May 1992 to identify risk factors for conjunctivitis and to recommend control measures to the hospital.
Methodology
We defined a case of neonatal conjunctivitis (NC) as any infant born in this hospital who presented with an eye discharge within six weeks of birth. We reviewed the laboratory and infection control registers for conjunctivitis. For each case in each month we chose one control-infant at random from the list of all infants admitted to the nursery. We reviewed medical files for birth circumstances and exposures in the nursery, including phototherapy and nurse assignment.
Results
We detected 55 cases of NC (attack rate [AR] = 24/1000). Forty were in the special care baby unit (SCBU), which had a higher AR (89/1000) (p<.001, Fisher's exact test) than the ward, where 15 cases were found (AR = 8/1000). A variety of organisms were isolated, including skin flora (Staphylococcus aureus) (AR = 6/1000), intestinal flora (Escherichia coli, other coliforms, fecal streptococci and Candida) (AR = 10.4/1000), and S. epidermidis or no organism (AR = 10/1000). NC patients were not more likely than control infants to have had phototherapy (OR=.75, 95% CI 0.33-1.7), whether the comparison was made for all NC or for the subgroups. There was no association of NC with handling by any specific nurse, with the length of hospitalization or with the type of delivery. The presence of a congenital defect was the only condition associated with NC (14.5%) when compared with control-infants (p<0.01, Fisher's exact test). A review of nursing practices in the SCBU revealed that nurses did not always wash their hands between patient contacts and that two rooms had no wash basins.
Conclusion
Admission to the SCBU was the principal risk factor for conjunctivitis. There, infants are handled by several people, compared with the ward, where infants are handled mostly by their mothers. The association of NC with congenital defects is probably due to more frequent handling of infants with such problems. The multiple organisms isolated suggest inadequate handwashing in the SCBU.