Without a virus culture system, genetic analysis becomes the principal method to classify norovirus (NoV) strains. Currently, classification of NoV strains beneath the species level has been based on sequences from different regions of the viral genome. As a result, the phylogenetic insights of some virus were not appropriately interpreted, and no consensus has been reached to establish a uniform classification scheme. To provide a consistent and reliable scientific basis for classifying NoVs, we analyzed the amino acid sequences for the major capsid protein of 164 NoV strains by first using an alignment based on the predicted 3D structures. A Bayesian tree was generated, and the maximum likelihood pairwise distances of the aligned sequences were used to evaluate the results from the uncorrected pairwise distance method. Analyses of the pairwise distances demonstrated three clearly resolved peaks, suggesting that NoV strains beneath the species level can be classified at three levels: strain (S), cluster (C), and genogroup (G). The uncorrected pairwise distance ranges for S, C, and G were 0-14.1%, 14.3-43.8%, and 44.9-61.4%, respectively. A scheme with 29 genetic clusters [8 in genogroup 1 (G1), 17 in G2, 2 in G3, and 1 each in G4 and G5] was defined on the basis of the tree topology with the standards provided and was supported by the distance analysis. Of these, five clusters in G2 and one in G1 are newly described. This analysis can serve as the basis for a standardized nomenclature to genetically describe NoV strains.

Between July 1997 and June 2000, fecal specimens from 284 outbreaks of nonbacterial gastroenteritis were submitted to the Centers for Disease Control and Prevention for testing for "Norwalk-like viruses" (NLVs). Specimens were examined by reverse-transcription polymerase chain reaction and direct electron microscopy for the presence of NLVs. Adequate descriptive data were available from 233 of the outbreaks, and, of these, 217 (93%) were positive for NLVs. Restaurants and events with catered food were the most common settings, and contaminated food was the most common mode of transmission. Genogroup II (GII) strains were the predominant type (73%), with genogroup I strains causing 26% of all NLV-positive outbreaks. Certain GII clusters (GII/1,4,j) were more commonly associated with outbreaks in nursing home settings than with outbreaks in other settings. Strain diversity was great: one potential new sequence cluster was implicated in multiple outbreaks, and strains belonging to a tentative new genogroup were identified.