[Pediatric AIDS at the University Hospital Center-Tokoin (Lomé): the role of protein-calorie malnutrition and a formulation trial of a clinical diagnostic score].

: Two cross-sectional studies were carried out in the pediatric ward of the Tokoin Teaching Hospital, Lome. One study determined the prevalence of HIV infection in the 49 malnourished patients treated in the ward in February to March and between August and December 1994. The other was carried out between July 1994 and January 1995 and included 57 other hospitalized children fulfilling at least one of the WHO's pediatric AIDS criteria. The aim was to draw up a screening system for pediatric AIDS based on clinical scores that would be more sensitive than and as specific as the WHO criteria. We tested these criteria and the other signs used in the suggested scoring system using the reference test, HIVchek. The seroprevalence of HIV was 28.6% in malnourished children and transmission was probably exclusively from mother to child. It was difficult to distinguish pediatric AIDS from protein energy malnutrition on clinical grounds, although some of the associated morbidities, including anemia, adenopathy and splenomegaly, were highly suggestive of pediatric AIDS. The second study showed that: 1) the sensitivity of the WHO criteria was low; 2) the best positive predictive values were obtained in cases of polyadenopathy and confirmed HIV infection of the mother. Both these criteria were relatively infrequent; 3) there were 6 criteria significantly associated with HIV infection, each being given a point score according to its Yule coefficient: chronic cough (4 points), chronic diarrhea (3 points), chronic fever (2 points), oropharyngeal candidiasis (2 points) and marasmus (1 point). A score of 4 points was the threshold for suspicion of pediatric AIDS. Our scoring system was more sensitive than the WHO criteria and had similar specificity and positive predictive value. We stress the importance of preventive measures against HIV infection, particularly for women of child-bearing age and suggest a new score test and appropriate clinical definitions for infants and older children.