Abstract 981 Poster Session IV, Tuesday, 5/4 (poster 151)

AIMS To determine the association of RV infection with diarrhea (D), vomiting (V), and/or fever (F) during a RV season among hospitalized children. METHODS At 3 children's hospitals between Nov 97 and June 98, we actively assessed RV infection upon admission of all hospitalized children 15 d to 5 yrs old who had D, V without respiratory or structural GI cause, and/or unexplained F. Acute stool and blood specimens were collected. Stool specimens were tested for RV by EIA. Convalescent blood was collected from children whose stool tested RV-negative. Serum pairs were tested for anti-RV IgA and IgG antibody. Discharge diagnoses were assembled from hospital information systems. RESULTS A total of 710 children were assessed, of whom 621 (87%) had a stool specimen tested for RV. RV detection rates were similar (40 to 48%) for each symptom when present alone or in combination (Table).

Table 1 No caption
Full size table

RV detection rates were highest when all three symptoms occurred in the same child, lower when two symptoms occurred together, and lowest when a symptom occurred alone (Table). Among children whose stool specimen tested negative for RV, and who had an acute/convalescent serum pair, 9/54 (17%) had a >4-fold rise in antibody to RV. A total of 274 assessed children received a discharge diagnosis that is a surveillance marker for RV-associated hospitalization (J Infect Dis 1990;162:598). By adjusting for unsampled children, a total of 362 (32% more) assessed children had RV infection. CONCLUSIONS RV detection rates were least when D, V, or F occurred alone and greatest when D, V, and F occurred in combination. Active assessment for RV infection that included measuring seroresponses detected 32% more RV-associated hospitalizations than passive, diagnosis-based methods.

Authors supported by Wyeth/Lederle, Merck, and NIH, all with financial interest in RV vaccine development.