Table 3. Methodological characteristics of health impact assessments of rotavirus vaccination programs conducted in Latin American and Caribbean countries, based on primary data collection.

Author/Year	Country	Study design	Data source	Clinical syndrome	Outcome	Main results
Primary data
Gouvea21 (2009)	Brazil (Rio de Janeiro, RJ)	Hospital-based survey including three years and a half before and one year after vaccine introduction	Primary data collection at the emergency room of one hospital. Data gathered from medical bulletins and patient records. Vital statistics obtained from the Brazilian Ministry of Health	All-cause diarrhea and laboratory-confirmed rotavirus diarrhea	Number of emergency room visits, hospitalizations and deaths, and genotype distribution	The study was unable to clearly show the impact of vaccination. Gastroenteritis visits and hospitalizations showed significant year-to-year variation. A gradual decrease in rotavirus strain diversity was observed in the prevaccination years
Safadi48 (2010)	Brazil (Sao Paulo, SP)	Prospective cohort including years before and after vaccine introduction	Prospective primary data collection in a private hospital, with routine rotavirus testing for all under-five children hospitalized for acute gastroenteritis	All-cause diarrhea and rotavirus-positive diarrhea	Number of hospitalizations and genotype characterization	Reduction in the number of all-cause and rotavirus-related diarrhea hospitalizations; delay in the rotavirus seasonal peak; and predominance of G2P[4] genotype in the postvaccination period
Borges8 (2011)	Brazil (Goiania, GO)	Cross-sectional. Data collection restricted to the postvaccination period was compared with prevaccination studies conducted in the same region Limitations: small sample size, data collection in a period shorter than 1 year	Primary data collection in seven day care centers. Children were enrolled independent of gastrointestinal symptoms	Rotavirus-positive diarrhea, genotype characterization	Proportion of rotavirus-positive samples	Presence of rotavirus in 3.6% of all samples and 10.4% of samples from children with diarrhea were rotavirus-positive, which is less than what was previously observed by other studies in the region (14.4%-37.2%). G2P[4] was the predominant circulating genotype
Assis2 (2013)	Brazil (Juiz de Fora, MG)	Cross-sectional, including pre- and postvaccination years. Limitations: small sample size, compromising genotype distribution analyses. Data cannot be generalized to the entire Country	A university virology laboratory	Rotavirus-positive diarrhea	Frequency of rotavirus-diarrhea and genotype characterization	Decrease in the proportion of rotavirus-positive diarrhea after vaccine introduction. The G1P[6] genotype was most frequent before vaccine introduction and replaced by the G2P[4] genotype in the year of vaccine introduction and after
Leboreiro30 (2013)	Mexico	Case series. Retrospective (pre-vaccination) and prospective (post-vaccine introduction) analysis of a case series	Hospital medical records of children treated in one hospital, including the emergency room and hospital wards	All-cause diarrhea and rotavirus-positive diarrhea	Frequency and severity of diarrhea	Reduction in rotavirus-related diarrhea; reduction in rotavirus diarrhea severity among vaccinated children