The observation that astrovirus is detected more frequently with diarrhea of increasing duration suggests the need for further studies to determine whether astrovirus plays a causative role in persistent diarrhea or is a secondary agent.

Group A rotavirus strains from 3 locations in Bangladesh collected over one year were examined. Serotypes 1-4 were found throughout in 2 locations, whereas an epidemic of RV diarrhoea due to a single rotavirus strain occurred in one location.

The stools were analyzed to determine the distribution of serotypes of group A rotaviruses (RV). A total of 898 stool samples (15.5%) contained RV, as determined by using an enzyme-linked immunosorbent assay. RV isolates from 855 of these samples were serotyped by using serotype-specific synthetic oligonucleotide probes. A total of 558 (65.3%) could be assigned to specific serotypes: 166 (19.4%), 228 (26.7%), 39 (4.6%), and 125 (14.6%) belonged to serotypes 1 through 4, respectively; 12 (1.4%) hybridized with more than one serotype; and 285 (33.3%) failed to hybridize. RV diarrhea was evident throughout the year, with peaks in the dry winter months and in September 1988, coinciding with a major flood. RV was isolated from 46.6% of patients between 7 and 12 months old. Among children under 24 months of age with RV diarrhea, 1.2% (10 of 828) died. The corresponding percentage for children with diarrhea from all causes is 0.9% (29 of 3,301).

A study of neonatal group A rotavirus (RV) infection in 3 hospitals of urban Bangladesh identified 60 infected babies of 100 studied in 2 hospitals. Preliminary evidence based on serotyping and electropherotyping suggest that these RVs differ from community strains.