Table 4. Methodological characteristics of health impact assessments of pneumococcal conjugate vaccination programs conducted in Latin American and Caribbean countries.

Author/Year	Country	Study design	Data source	Clinical syndrome	Outcome	Main results
Secondary data
Afonso1 (2013)	Brazil (five capitals)	Ecological (interrupted time-series analysis) Control: non-respiratory causes Limitations: short observation period after vaccine introduction (one year)	National hospitalization information system (public health system). Data from five capitals that had good data quality and high PCV-10 vaccination coverage	All-cause pneumonia	Hospitalization rates among children aged two months to two years	Significant declines in the hospitalization rates for pneumonia in three capitals (Belo Horizonte, Curitiba and Recife), but not in the other two (Sao Paulo and Porto Alegre). Prevaccination hospitalization rates for pneumonia varied substantially by city. Hospitalization rates for non-respiratory causes also decreased in all cities, but at a lower rate
Nieto Guevara35 (2013)	Panama	Descriptive (retrospective), comparing a three-year period including pre- and postvaccination years. Involved indigenous population. Limitations: results cannot be generalized	Medical records of a secondary-level referral hospital, based on discharge diagnosis of pneumonia given by the treating physician and coded according to ICD	All-cause pneumonia	Hospitalization rates and transfers to the regional hospital among under-five children	Reduction in hospitalization rates and referrals for pneumonia were observed after vaccine introduction. Results cannot be generalized
Pírez41 (2011)	Uruguay	Descriptive (retrospective), comparing a three-year period before vaccine introduction with a one-year period after	A national tertiary referral pediatric hospital database, complemented by medical records, laboratory databases and reports from the national information system on notifiable diseases	All-cause pneumonia, pneumococcal pneumonia, pneumococcal meningitis	Hospitalization rates among children aged one month to 14 years	Reduction in hospitalization rates for pneumococcal pneumonia and pneumococcal meningitis after PCV-7 introduction. Non-vaccine serotypes 1, 5, 7F, 19A, and 24F became the most frequent causes of pneumococcal pneumonia after vaccine introduction. There were changes in the national hospital admissions aimed to decrease hospitalizations during the study. These changes did not affect the rates of hospital discharges for acute gastroenteritis, the control disease analyzed
Pirez42 (2014)	Uruguay	Descriptive (retrospective), comparing years before (2003-2007) and after (2009-2012) vaccine introduction	Microbiology laboratory database and patient records of a single site	All-cause pneumonia, pneumococcal pneumonia, pneumococcal serotypes	Hospitalization rates among children aged zero to 14 years	Significant reduction in hospitalization rates. A clear two-step reduction in hospitalization rates for pneumonia after each introduction of PCV (PCV-7 and PCV-13). Significant reduction in PCV-13 vaccine serotypes and increase in non-vaccine serotypes after the vaccination program implementation
Becker-Dreps7 (2014)	Nicaragua (León)	Descriptive, comparing a period before (2008-2010) and after (2011-2012) vaccine introduction. Control: healthcare visits due to diarrhea	Epidemiological database of 107 public health facilities (93 healthcare centers, 13 primary care centers, 1 public referral hospital)	Pneumonia	Hospitalization rates, outpatient visits among children aged 0 to 14 years and infant mortality	Reduction in hospitalization and outpatient visits for pneumonia and decrease in infant mortality after vaccine introduction. No changes in overall healthcare visits for diarrhea during the study period
Surveillance data
Hortal25 (2007); Hortal26 (2012)	Uruguay	Two prospective cohorts – a three-year study before27 and a three-year study after28 vaccine introduction	Population-based surveillance system carried out in two municipalities. The study was conducted in four hospitals (two public and two private)	All-cause pneumonia	Annual hospitalization rates; serotype distribution among under-five children	Hospitalization rates for pneumonia did not decline in the three-year prevaccination study. Significant reduction in hospitalization rates for pneumonia and changes in serotypes after vaccine introduction
Primary data
Santos50 (2013)	Brazil (Sao Paulo, SP)	Case series including years before and after vaccine introduction	Prospective data collection at one university hospital that attends a population of approximately 408,000 inhabitants	Invasive pneumococcal disease	Number of cases/1,000 hospital admissions; antibiotic resistance and serotype distribution	Decrease in invasive pneumococcal disease cases among children under 2 years of age and decrease in vaccine serotypes after vaccine introduction
Parra38 (2013)	Colombia	Case series on invasive pneumococcal disease and two transversal studies (before and after vaccine introduction) on nasopharyngeal carriage. Limitations: short observation period after vaccine introduction; children’s vaccination status unavailable	Laboratory surveillance (SIREVA II) data on invasive pneumococcal disease and primary data collection on nasopharyngeal carriage	Invasive pneumococcal disease and nasopharyngeal carriage serotype distribution	Serotype distribution	Decrease in vaccine serotypes and increase in non-vaccine serotypes in invasive pneumococcal disease isolates after vaccine introduction

ICD: International Classification of Diseases; PCV: pneumococcal conjugate vaccine; PCV-7: 7-valent pneumococcal conjugate vaccine; PCV-10: 10-valent pneumococcal conjugate vaccine; PCV-13: 13-valent pneumococcal conjugate vaccine