March 19, 1982

Confusing Correlates of Chlamydial Infection

Author Affiliations

Centers for Disease Control Emory University School of Medicine Atlanta

JAMA. 1982;247(11):1606-1607. doi:10.1001/jama.1982.03320360056037

Associating a tenfold increased risk of perinatal death with prenatal endocervical chlamydial infection, as Martin et al have described in this issue of The Journal (p 1585), is remarkable. The authors themselves caution against overinterpretation of their surprising findings. They point out several limitations to their study: the small numbers of women, unequal numbers of uninfected mothers matched to each infected mother, the absence of data on simultaneous infections, and the lack of detailed pathological examinations of the placentas and products of conception. Their prospective study divided pregnant women into groups defined by presence or absence of endocervical infection with Chlamydia trachomatis. Each group was followed up to determine the incidence of perinatal death. The two study groups differed in a number of ways that could have confounded the findings. The authors took this into account by attempting "to find two matched uninfected controls for each of the Chlamydia-positive