FormalPara Key Summary Points

Why carry out this study?

Rotavirus (RV) is the most common cause of diarrhea among children, and the leading cause of childhood mortality in Latin America before rotavirus vaccination.

This study evaluated the impact of RV vaccine implementation on the burden of acute diarrheal disease (ADD) and RV positive cases.

The study also evaluated the impact of RV vaccination on hospitalizations among ADD and RV positive cases.

What was learned from the study?

Significant decrease in notified ADD cases and RV cases in children < 2 years after RV vaccine introduction in Argentina in 2015.

Hospital discharges due to ADD and RV infection reduced in children < 2 years after RV vaccine introduction.

Digital Features

This article is published with digital features, including a graphical abstract, to facilitate the understanding of the article. To view digital features for this article go to https://doi.org/10.6084/m9.figshare.21221468.

Introduction

Rotavirus (RV) infection is the most common cause of diarrhea among children aged 5 years or less worldwide [1]. Globally, between 2013 and 2017, RV was the leading etiology of nearly 122,000–215,000 RV-related diarrhea deaths among children under the age of 5 years [2,3,4]. In Latin America and the Caribbean, RV is the leading cause of childhood mortality, and it was estimated to cause 8000 annual deaths prior to the introduction of RV vaccination [5, 6].

RV vaccination is recommended to prevent RV disease; about nine out of ten children who receive the vaccine are protected from its severe form [7]. In 2009, the World Health Organization (WHO) recommended its use among infants to reduce associated fatality rates, especially in countries with childhood mortality rates > 10% [8]. Currently, four WHO-approved RV vaccines are available for infant vaccination: RotaTeq (Merck & Co. Inc., Whitehouse Station, NJ, USA), Rotarix (GSK, Rixensart, Belgium), Rotavac (Bharat Biotech International Ltd, India), and ROTASIIL (Serum Institute of India, India) [9]. While 114 countries have introduced RV vaccines in their National Immunization Programs (NIPs), approximately 59 million children still lack access to RV vaccines [10]. Seven Latin American countries that had introduced the RV vaccines in their NIPs notified vaccine coverage rates ranging from 39% to 91.8% in 2007 [11]. As of January 2011, 14 Latin American countries have included RV vaccines in their NIP’s [6]. According to the 2021 estimates from the Pan American Health Organization, RV vaccine coverage in Latin America ranged from 56% in Paraguay to 100% in Nicaragua [12]. Argentina incorporated the RV vaccine (Rotarix, GSK, Belgium) in its NIP in 2015 [13]. This monovalent vaccine, made from the human viral strain G1P [8], is administered orally as a two-dose regimen to infants aged 2 and 4 months [14, 15]. After introduction of the vaccine, the vaccine coverage rates in Argentina were notified to be 61% [16]. The latest data from the WHO (2021) estimates an increased vaccination coverage rate of 74% [17].

To date, several studies have evaluated the burden of RV disease in Latin American countries after the inclusion of the vaccine in their NIPs. Data showed a significant reduction in mortality associated with acute diarrheal disease (ADD) in the post-vaccination period (22–54%), in both infants aged less than 1 year and children aged less than 5 years [18,19,20,21,22]. The same trend was observed in the number of hospitalizations due to ADD, which decreased by 17–52%, and specifically for RV cases, by 59–81% [19, 22,23,24,25,26,27,28,29].

In the province of San Luis, Argentina, RV vaccination was implemented 2 years earlier than in the rest of the country [30]. A study conducted in this province to evaluate the use of the vaccine showed a reduction of 20–25% in ADD cases and 55–60% in hospitalizations in children younger than 5 years [30]. However, to date, a similar analysis has not been performed for the rest of the country [31, 32]. The objective of this study, therefore, is to evaluate how the introduction of RV vaccination in the Argentinian NIP affected the number of notified ADD cases, RV cases, ADD hospitalizations, and RV hospitalizations (in different datasets of the Ministry of Health) in children aged less than 5 years.

Methods

A counterfactual time-series analysis was performed to study the impact of RV vaccination on the number of notified ADD cases (January 2013–May 2018), notified RV cases (January 2012–December 2018), ADD hospital discharges (January 2011–December 2017), and RV hospital discharges (January 2011–December 2017) in children aged less than 5 years in Argentina. Data were collected prior to (2012/2013–2014) and post (2015–2017/2018) RV vaccine introduction in the Argentinian NIP on 1 January 2015. Vaccine impact was measured estimating the difference between a predicted time trend (simulated using 2011/2013–2014 data) and the actual post-vaccination data (2015–2017/2018). This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Database Description

ADD is a mandatory notifiable disease in Argentina. The National Healthcare Surveillance System (SNVS, Sistema Nacional de Vigilancia de la Salud) gathers all the data relating to the ADD cases. The SNVS constitutes two units: clinical and laboratory.

The clinical surveillance unit (C2) of the SNVS notifies ADD cases from all “medical consultations,” making no distinction between inpatients and outpatient visits, and etiology [33, 34]. Data were collected from January 2013 to May 2018 by epidemiological week and were classified by age group: < 1 year, < 2 years, 2–5 years.

The laboratory surveillance unit (SIVILA, Sistema Nacional de Vigilancia por Laboratorios) of the SNVS gathers information from several sentinel hospital laboratories to report the number of tested cases and their etiology (e.g., RV) [34, 35]. Data were collected from January 2012 to December 2018 by epidemiological week and were classified by age group: < 1 year, < 2 years, 2–5 years.

Moreover, all notified RV cases were also regularly analyzed by the National Administration of Laboratories and Health Institutes (ANLIS, Administración Nacional de Laboratorios e Institutos de Salud) [36].

Hospital discharge data (due to ADD and RV) were retrieved from the Health Statistics and Information Office (DEIS, Dirección de Estadística e Información en Salud) [37]. Data were collected monthly from January 2011 to December 2017 and were classified by age group: < 1 year, < 2 years, 2–5 years. Hospital discharge cases were classified based on the International Classification of Diseases 10th revision (ICD-10) codes A08 “viral and other specified intestinal infections” and A09 “other gastroenteritis and colitis of infectious and unspecified origin” [38]. Hospital discharge data were missing for several provinces, and thus not included in the analysis (Table S1).

Samples were gathered from all regions of Argentina, except the province of San Luis, which was excluded from the analysis because vaccination was implemented in 2013.

Detailed information about the databases can be found in the Supplementary Material, Table S1.

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Statistical Analysis

A counterfactual Bayesian time-series analysis was performed to assess the impact of the vaccine on the number of notified ADD and RV cases and hospital discharges [39]. A predicted scenario in which the vaccine was not introduced was modeled using data prior to the introduction of the RV vaccine in the NIP. The effect of vaccination was derived from the difference between the model-predicted scenario and the actual data.

A validation analysis of the model was performed as indicated in the Supplementary Material, Table S2.

Outcomes

Four outcomes were defined:

  • The first outcome was to assess the impact of the vaccine on the number of notified ADD cases in the age groups < 1 year, < 2 years, and 2–5 years.

  • The second outcome was to assess the impact of the vaccine on the number of notified RV cases in the sentinel hospitals for the age groups < 1 year, < 2 years, and 2–5 years.

  • The third outcome was to assess the impact of the vaccine on the number of hospital discharges following ADD in the age groups < 1 year, < 2 years, and 2–5 years.

  • The fourth outcome was to assess the impact of the vaccine on the number of hospital discharges following RV in the age groups < 1 year, < 2 years, and 2–5 years.

Results

Overall results are presented in Table 1. Since vaccine introduction in January 2015, a decrease in the number of cases and hospital discharges per month was observed for both ADD and RV cases.

Table 1 Impact of RV vaccination on the number of notified ADD and RV cases, and the number of ADD and RV hospital discharges per month by age group
Full size table

Impact of RV Vaccination on the Number of Notified ADD Cases

A statistically significant ADD case reduction was observed in Argentina’s surveillance system for all age groups analyzed, being lowest in the age group least exposed to the vaccine (i.e., 2–5 years). Percentages of cases averted per month due to vaccination by outcome and age group varied between 12.9% and 26.3% (see Table 1).

The cumulative reduction in the number of notified ADD cases per year was greater for the age group < 2 years (Fig. 1). A total of 233,947 ADD cases were averted after vaccine introduction in this age group (Table 1 and Fig. 1). Similarly, in both age groups of < 1 year and 2–5 years of age, a reduction in notified ADD cases after vaccine incorporation was found (Figs. S1, S2). The cumulative reduction is more pronounced in the < 1 year age group than in the 2–5 year age group (Fig. S2).

Fig. 1
figure 1

Impact of RV vaccination on the number of notified ADD cases in children < 2 years. The graphic at the top shows the number of predicted and observed notified ADD cases over time. The dotted line shows the predicted values (January 2013–May 2018 without vaccination). The solid line represents the observed values (January 2013–December 2014 without vaccination and January 2015–May 2018 with vaccination). The graphic in the middle shows the trend over time for the difference between the predicted and observed ADD cases for the age group of < 2 years of age by month. The graphic at the bottom shows the cumulative difference between the predicted and observed data over time. Blue shades are confidence intervals. The vertical dotted line is the date of vaccine inclusion in the NIP (January 2015). ADD acute diarrheal disease, RV rotavirus

Full size image

Impact of RV Vaccination on the Number of Notified RV Cases

As presented in Table 1, the number of notified RV cases decreased in all age groups in the years after the implementation of the vaccine. Moreover, these estimations were higher than those observed for notified ADD cases (the RV outcome is more specific to assess the vaccine’s effects), being 54.0% in the age group < 2 years. A total of 3323 RV cases were averted in the selected sentinel hospitals after vaccine introduction in this age group (Table 1 and Fig. 2). A statistically significant difference in the number of notified RV cases attributable to vaccination was observed in the age groups < 1 year and < 2 years. No significant differences were observed for the age group of 2–5 years.

Fig. 2
figure 2

Impact of RV vaccination on the number of notified RV cases in children < 2 years. The graphic at the top shows the number of predicted and observed notified RV cases over time. The dotted line shows the predicted values (January 2012–December 2018 without vaccination). The solid line represents the observed values (January 2012–December 2014 without vaccination and January 2015–December 2018 with vaccination). The graphic in the middle shows the trend over time for the difference between the predicted and observed notified RV cases for the age group < 2 years of age by month. The graphic at the bottom shows the cumulative difference between the predicted and observed data over time. Blue shades are confidence intervals. The vertical dotted line is the date of vaccine inclusion in the NIP (January 2015). ADD acute diarrheal disease, RV rotavirus

Full size image

Again, the age group with the biggest reductions in the number of notified RV cases per year was the group < 2 years (Fig. 2, Fig. S3). A reduction in the number of RV cases is observed following vaccine introduction, together with a loss of seasonality over the study period for the age groups < 1 year and < 2 years (Fig. 2, Fig. S3).

Impact of RV Vaccination on the Number of Hospital Discharges Following ADD

After implementation of the RV vaccination, a decrease in hospital discharges following ADD was observed for the age groups < 1 year and < 2 years (Table 1, Fig. 3, Figs. S4, S5). These reductions varied between 19.0% and 15.4% in children < 1 years old and < 2 years old, respectively (Table 1). A total estimate of 4263 hospital discharges following ADD were averted after vaccine introduction in the group < 2 years of age (Table 1 and Fig. 3). In the 2–5 years age group, the trend observed was the opposite, with a slight nonsignificant increase (Table 1, Fig. 3, and Fig. S5).

Fig. 3
figure 3

Impact of RV vaccination on the number of hospital discharges following ADD in children < 2 years. The graphic at the top shows the number of predicted and observed ADD hospital discharges over time. The dotted line shows the predicted values (January 2011–December 2017 without vaccination). The solid line represents the observed values (January 2011–December 2014 without vaccination and January 2015–December 2017 with vaccination). The graphic in the middle shows the trend over time for the difference between the predicted and observed ADD hospital discharges for the age group < 2 years of age by month. The graphic at the bottom shows the cumulative difference between the predicted and observed data over time. Blue shades are confidence intervals. The vertical dotted line is the date of vaccine inclusion in the NIP (January 2015). ADD acute diarrheal disease, RV rotavirus

Full size image

Impact of RV Vaccination on the Number of Hospital Discharges Following RV

A statistically significant reduction in hospital discharges following RV was observed for the age groups < 1 year and < 2 years, where the percentages of cases averted per month were about 40% (40.8% and 39.3%, respectively) (Table 1). The cumulative number of prevented RV hospital discharges after vaccine introduction was greater for the age group < 2 years, with an estimated 479 cases of RV averted in the analyzed dataset (Table 1, Fig. 4, Fig. S6).

Fig. 4
figure 4

Impact of RV vaccination on the number of hospital discharges following RV in children < 2 years. The graphic at the top shows the number of predicted and observed RV hospital discharges over time. The dotted line shown the predicted values (January 2011–December 2017 without vaccination). The solid line represents the observed values (January 2011–December 2014 without vaccination and January 2015–December 2017 with vaccination). The graphic in the middle shows the trend over time for the difference between the predicted and observed RV hospital discharges for the age group < 2 years of age by month. The graphic at the bottom shows the cumulative difference between the predicted and observed data over time. Blue shades are confidence intervals. The vertical dotted line is the date of vaccine inclusion in the NIP (January 2015). RV rotavirus

Full size image

Discussion

To our knowledge, this is the first time series analysis that evaluates the impact of RV vaccination on the number of notified ADD and RV cases and hospital discharges, following its introduction in the Argentinian NIP. During the period analyzed, a decreasing trend was observed, being more pronounced for the age groups < 1 year and < 2 years than for the age group 2–5 years. The impact was in all cases greater for RV disease than for ADD. For the age groups < 1 year and < 2 years, the reduction in the number of notified ADD cases was in line with the number of ADD hospital discharges; the same effect was observed for notified RV cases and hospital discharges.

A statistically significant reduction in notified ADD cases was also observed in the 2–5 years age group. This age group may include a proportion of children who were not eligible to be vaccinated based on age. This may suggest indirect community-based protection or herd immunity as a consequence of infant RV vaccination. A similar observation has been notified in El Salvador, where a significant reduction in RV-related hospitalizations was observed in children ineligible for the vaccine [28]. Indirect community-based protection was also observed in Europe [40, 41] and the USA [42].

In contrast, RV vaccination did not show any noticeable impact on the number of notified RV cases and hospital discharges in the 2–5 years age group. It is possible that no significant reduction was seen due to the overall low number of RV cases notified.

This study also reports a loss of seasonal rotavirus epidemic patterns after introduction of the RV vaccine. This is in line with a previous study that has shown that seasonal trends of RV were less pronounced after vaccine introduction [43]. Currently, in the USA, high RV vaccine coverage of 85% among infants resulted in annual RV epidemic patterns [44].

The RV vaccine coverage rates have been steadily increasing in Argentina since the introduction of the vaccine. Coverage rates have risen from 61% in 2015 to 75%, 78%, and 80% in 2016, 2017, and 2018, respectively [10, 16]. An increase in vaccine coverage rates may have contributed to the decreased incidence of ADD, as observed in the current study, and the previously performed analysis in the province of San Luis, Argentina [30]. Most recently, a vaccine impact of 25% and 23% in the number of ADD cases was estimated for children aged < 1 year and < 2 years, respectively, in line with our estimations (26.3% and 22.1%, respectively). However, the impact on the number of ADD hospitalizations was greater in the San Luis study (55–60% in the San Luis study and 15–19% in the current study) [30].

Another study in Argentina evaluated the early impact of RV vaccine implementation on the ADD burden. This observational study included post-vaccination data from 2016 and compared them with previous years (2011–2015). They found a decrease of 20.8% in ADD cases and 61.7% in RV cases in children < 5 years [32]. Our study showed lower vaccination impact, although the results are aligned. This difference may be explained by the number of years post-vaccination analyzed: this study only included the first year (2016), while our study evaluated three years (2016–2018). But it may also be explained by the different strategies used to calculate the impact of RV vaccination; Degiuseppe et al. compared 2016 data with the mean number of cases during the pre-vaccination period (2011–2014) [32], while in our study the impact was derived from the difference between the predicted (calculated through simulations) and the actual data (2015–2018).

Since 2006, several Latin American countries have implemented RV vaccination in their NIPs. Brazil was one of the first to include this vaccine in 2006, altogether with El Salvador, Mexico, Nicaragua, Panama, and Venezuela [1]. Since then, several publications in Brazil have estimated 17–52.5% of ADD hospitalizations averted and a 42.2% reduction in RV samples [19, 23, 27, 29]. Panama followed the same trend, with a decrease of 22–37% in ADD hospitalizations among children < 5 years [18, 24]. Mexico, after vaccine implementation, saw a decline in ADD hospitalizations of 40–47%, with the greatest decline in infants < 1 year (52%) and children < 2 years old (43–49%) [25, 45]. In El Salvador, RV hospitalizations were reduced by 69–81%, again with the highest decline observed in infants < 1 year, although sizable reductions were also observed in children < 2 years [28]. After introduction of the vaccine in Bolivia, El Salvador, Honduras, and Venezuela, the reduction of diarrhea-related hospitalizations and deaths ranged from 15.7% to 56.8% and from 5.6% to 17.9% respectively [22]. Although estimates vary widely between these countries, partially influenced by vaccine uptake, they are consistent and in agreement with our results in estimating a decreasing trend in the number of cases and hospitalizations after RV vaccine implementation. Although the number of cases and hospitalizations followed an overall decreasing trend, this effect was statistically significant only for the age groups < 1 year and < 2 years. A small nonsignificant increase was observed in the number of ADD hospitalizations in the 2–5 years age group (8.6% increase), which may be attributed to inconsistencies in data registries. In addition, RV vaccination resulted in similar reduction rates for the number of notified ADD cases and ADD hospital discharges. For instance, in infants < 1 years, reductions of 26.3% [95% confidence interval (CI) 17.4–36.0%] and 19.0% (95% CI 8.6–29.3%) were observed in the number of notified ADD cases and ADD hospital discharges, respectively; and reductions of 59.4% (95% CI 28.6–88.9%) and 40.8% (95% CI 20.2–61.6%) were observed in the number of notified RV cases and RV hospital discharges.

This study has several limitations. First, the number of ADD cases is expected to be under-reported, even though ADD represents a mandatory notifiable event in Argentina. Under-notification occurs mainly because people do not seek medical care owing to the mildness and self-limited nature of the symptoms. Also, institutional commitment to national health information systems may vary between different institutions and regions in the country. RV data is only provided by certain sentinel hospitals participating in the laboratory surveillance system, and even in these institutions RV testing is limited to the demand of local physicians and is not a population-based surveillance strategy [32].

Second, complete hospital discharges datasets were not available for some Argentinian provinces (i.e., Catamarca, Corrientes, Entre Ríos, Formosa, Jujuy, La Pampa, La Rioja, Misiones, Rio Negro, San Juan, Santiago del Estero, and Tierra del Fuego). These provinces were thus excluded from the analysis to avoid inconsistency of the final results; however, this might have affected the observed effect of the RV vaccine. Previously, total numbers of RV hospitalizations per year in Argentina of around 18,000 and 24,000 were estimated [46, 47]. Although this dataset contains a very much reduced number of RV cases analyzed, we were able to demonstrate a significant impact observed after vaccine introduction, even when vaccine coverage was 61–80% during the studied period [17].

Conclusion

The introduction of the RV vaccine in the NIP of Argentina has contributed to the decline in the number of notified ADD and RV cases, especially among children < 2 years (22.1% and 54.0%, respectively). The reduction in the number of ADD and RV hospital discharges after vaccine implementation was in line with the reduction in the number of notified cases. This study paves the way for future studies to assess the effect of RV vaccination on the economic burden of RV and ADD in addition to gastroenteritis-related (GE) mortality. Beyond these limitations, this evaluation of the actual impact of the vaccine may complement clinical trials, providing important information for decision-making, as well as for the assessment of public health programs and interventions.