However, in contrast to these and our findings, Xerry et al. described 100% identity in P2 regions of strains in different nosocomial clusters, the largest of which involving four wards in two different hospitals over a seven weeks�� period where 24 samples were examined. A possible epidemiological explanation for outbreaks consisting of identical strains could be a point source such as contaminated food or continuous transmission of the same strain from a contaminated environment, the extent of which is still debated. In our study, no nucleotide changes were found within the sequences in four clusters representing outbreaks confined to one distinct department within a one month��s period. In two clusters closely related sequence patterns were found temporarily related in the same department. Our study does not permit a definite conclusion whether these outbreaks with identical or nearly identical sequences were due to environmental transmission or shorter person-to-person transmissions between individuals with acute infections without evidence of in vivo evolution. In immunocompromised patients, the in vivo evolution of the virus leading to the development of quasispecies has been shown. In accordance to this, we found the highest sequence diversity in the prolonged outbreak in the hematology department, where the majority of 8-Chloroadenosine patients was immunosuppressed either due to their underlying disease or induced by chemotherapy. This outbreak included four of the six patients, where sequences from paired samples were available. The high frequency of mixed bases in these patients�� second samples shows that viral quasispecies had evolved and there was indication for subsequent person-to-person transmission. The frequencies of sequence variation in paired samples from immunocompromised patients with chronic NoV infection in our study are in accordance with previous studies. Repeated analysis of the data by generation of a Maximum Likelihood tree allocated the sequences to the same clusters, demonstrating the method��s robustness despite of the low bootstrap value of cluster 1. However, one of the Den Haag 2006b singleton sequences was obtained from a sample taken on the same day and in the same department as a sample belonging to the heterogeneous cluster 1. As these two sequences only differ by one nucleotide substitution, nosocomial transmission is very likely. This underlines that manual inspection of the sequences, eventually supplemented with epidemiological information, is 5-OMe-UDP trisodium salt helpful in interpreting phylogenetic analysis in areas of uncertainty. This study was designed to investigate sequence variation as an indicator for NoV evolution to study transmission routes.