Universal screening for tuberculosis identifies significantly more cases of the disease than symptom–based testing and significantly reduced early infant mortality, according to a Johns Hopkins study comparing these two methods of screening in HIV–infected pregnant women in South Africa. The work was presented at 9th International AIDS Conference on HIV Science by Neil Martinson, MD, MPH, adjunct associate professor of medicine at the Johns Hopkins University Center for Tuberculosis Research and executive director of the Perinatal HIV Research Unit in Soweto, South Africa. The results support testing for tuberculosis in all HIV–infected pregnant women in high–burden areas, regardless of symptoms, to significantly increase detection of the disease and lower mortality rates for their infants.

The current symptom–based screening method puts pregnant women and their infants at risk, as pregnancy may mask symptoms of active tuberculosis, delaying treatment. This potential diagnostic disparity prompted Martinson, Richard Chaisson, MD, director of the Center for Tuberculosis Research at the Johns Hopkins University School of Medicine, and their team of researchers to conduct a study in which 16 public–sector antenatal clinics in two health districts in South Africa were randomly assigned to use either a universal or symptom–based screening method.

The study enrolled 2,032 HIV–infected pregnant women and revealed that universal tuberculosis screening identified 10 times as many cases as symptom–based testing and reduced early infant mortality by more than 50 percent.

The investigators recommend that all pregnant women with HIV infection should be tested for TB, regardless of symptoms.