Stool samples were collected from children attending 6 public hospitals located in different healthcare areas throughout Spain. These hospitals intentionally represented the geographic, climatic, and ethnic diversity of Spain. Their respective catchment areas are shown in Table 1. The study was conducted between January 2005 and January 2006 and included children <5years of age who were hospitalized with acute gastroenteritis and from whom a stool sample was obtained.

Acute gastroenteritis was defined as >3 looser-than-normal stools within a 24-hour period or an episode of forceful vomiting and any loose stool. To enable reporting of test results to hospitals, stool specimens were labeled with the date of collection and a unique surveillance identification number. Permission for enrollment in the study was obtained from children's legal guardians, and ethical approval was obtained from the institutional review board of the Hospital de La Ribera.

Whole stool specimens were collected and transported immediately to hospital laboratories and stored at 4°C until processing. All fecal samples were screened for enteropathogenic bacterial agents by conventional culture methods previously described (14).

Each month, specimens were sent to the reference laboratory (Viral Gastroenteritis Unit, National Center for Microbiology, Instituto de Salud Carlos III, Madrid, Spain). A 10% suspension in 0.1 mol/L phosphate-buffered saline (pH 7.2) was prepared and tested by reverse transcription (RT)-PCR for rotavirus, astrovirus, norovirus, and sapovirus (11,15,16) and by an immunochromatographic method for enteric adenoviruses (14).

Viral RNA was extracted from 250 μL of the 10% fecal suspension by using the guanidine isothiocyanate method and the Rnaid Spin Kit (BIO 101, Anachem Bioscience, Bedfordshire, UK) according to the manufacturer's instructions, with slight modifications (16). RNA was eluted in 50 μL of RNase-free distilled water and stored at –20°C. To determine the G/P type patterns present in children hospitalized from 2005 through 2006, a total of 98 rotavirus strains were P typed. G and P rotavirus genotyping were performed by using RT-PCR methods as previously reported (11,17).

Rotavirus amplicons were genetically characterized by nucleotide sequencing of both strands of the amplified PCR products. These products were purified by using QIAquick PCR Purification kit (Qiagen, Valencia, CA, USA) and then sequenced using an ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA) on an ABI automated sequencer (Applied Biosystems, model 3700). Data analysis was performed by using Clustal for multiple alignments and neighbor-joining and maximum parsimony methods for phylogenetic analysis (Bionumerics, Kortrijk, Belgium). Spanish strains were submitted to GenBank under accession numbers DQ440613 through DQ440624.

A total of 656 hospitalized children were enrolled. Enteropathogenic bacterial strains were detected in 5.0% of samples (Table 2). Astrovirus and norovirus RNA was detected in 3.2% and 12.0% samples, respectively, and adenovirus antigen in 1.8% samples.

A total of 403 rotavirus strains were detected. Rotavirus was found alone in 362 (55.2%) samples but was found in another 41 samples (6.3%) as a coinfection with other viruses. The percentage of children with gastroenteritis caused by rotavirus as unique agent ranged from 36.7% in Leon to 68.2% in Valencia (Table 2).

G typing RT-PCR for rotavirus alone was performed on 362 samples positive for rotavirus but could not be determined in 10 (2.8%) samples. The G types detected, including mixed infections with multiple rotavirus strains, are shown in Table 3. Briefly, the most predominant G type was G9 (50.6%), followed by G3 (33.0%), G1 (20.2%), and G2 (7.1%); the least common G type was G4 (0.6%). G1, previously reported as the most common G type in Spain, was found in only 20.2% of rotavirus infections. With the exceptions of Valencia and Albacete, where G1 and G3, respectively, were the predominant G types, the results from all other regions showed a predominance of G9. However, even in these 2 areas, G9 was the second most common strain detected when cases with coinfection were added (26.7% and 31.6%, respectively).

Common G/P combinations, infrequent patterns, and mixed-infection combinations were all detected (Table 4). G9P[8] (40%) and G3P[8] (31%) were the most common combinations detected, but G types in combination with P[6] and P[9] were also detected.

Using DNA sequencing and phylogenetic analysis of partial sequences of the gene encoding VP7, we compared 2 G3 strains from this study with 9 G3 strains isolated previously in Spain. All G3 strains from Spain shared >99.0% homology and were more closely related to each other than to strains isolated in Italy, United Kingdom, India, and China.