Review Reports - Persistent Varicella Transmission in a High-Coverage Single-Dose Vaccination Program: Costa Rica, 2015–2024

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript is interesting but requires major revisions. Please address all reviewers' comments point by point. The main issues are outlined below.

1) The authors state that the reduction in cases is attributable to the vaccination program; however, the period analyzed is 2015–2024—many years after the vaccine was introduced in 2007—so no true comparison exists. Please clarify this point.

2) The authors do not present data on the individual vaccination status of the cases, seroprevalence, or population susceptibility; stating that a second dose is needed without providing supporting data is incorrect.

3) Denominators are missing in some instances—for example, regarding the 8,715 outpatient visits and 19,604 emergency department visits. What are the denominators?

4) Diagnosis is primarily clinical, and laboratory confirmation is not systematic; this represents a significant methodological limitation.

5) Instead of "varicella-associated death," given the methodology used, it would be more appropriate to refer to "deaths occurring during hospitalizations with a varicella diagnosis," or the criteria should be better clarified.

6) It is unclear why the authors highlighted the 2023 outbreak. Please provide further explanation.

7) The concept of a 95% threshold appears repeatedly in the manuscript. The authors themselves correctly note that there is no universally applicable threshold and that population protection depends on transmissibility, vaccine efficacy, and the proportion of the population that is immune.

8) Confidence intervals for the rates are missing; including them would certainly facilitate better comparisons.

9) The authors do not perform an inferential analysis in the study; it reads more like a report than a scientific article. Did the authors consider conducting such an analysis? Do they have access to socio-demographic data, for instance? It would be worth attempting this analysis to identify factors associated with a higher risk of disease and/or hospitalization.

10) In the discussion, better compare the obtained results with those found in the literature.

11) The conclusions should provide more precise guidance to policymakers.

Kind regards

Author Response

Thank you for pointing this out. We agree with this observation. Universal varicella vaccination was introduced in Costa Rica in 2007, whereas the present study analyzes epidemiological trends beginning in 2015. A previous study conducted in Costa Rica described changes in varicella epidemiology by comparing the pre-vaccination period with the years following vaccine introduction through 2014; however, those data were not included or reanalyzed in the present study. Therefore, our analysis itself does not include a pre-vaccination comparator and cannot establish a causal relationship between the vaccination program and the reductions observed during our study period.

Accordingly, we have revised the Abstract, Discussion, and Conclusions to describe the epidemiological changes observed in our study without directly attributing them to vaccination. The previous Costa Rican study is used only to provide historical context regarding the period before and following vaccine introduction. We have also clarified that trends observed during the COVID-19 pandemic may reflect changes in transmission, healthcare utilization, and case reporting, whose individual contributions cannot be quantified using our data.

All corresponding revisions in the Abstract, Discussion, and Conclusions are clearly identified in red using Track Changes in the revised manuscript.

2. The authors do not present data on the individual vaccination status of the cases, seroprevalence, or population susceptibility; stating that a second dose is needed without providing supporting data is incorrect.

Thank you for this important observation. We agree that our data do not allow us to establish the need for a second dose of varicella vaccine in Costa Rica. The present study is a retrospective analysis of aggregated surveillance and administrative data. These data sources do not routinely link individual vaccination status to reported varicella cases, and no individual medical record review was performed. Furthermore, seroprevalence data were not available. Therefore, individual vaccination status and population susceptibility could not be directly assessed.

We have revised the manuscript accordingly to avoid implying that our findings demonstrate the need for a second dose. Instead, we now emphasize that additional information on individual vaccination history and population susceptibility is needed before the potential contribution of a two-dose schedule can be adequately evaluated in Costa Rica. The recommendation has therefore been reframed as a topic for further evaluation rather than as a conclusion supported directly by the present study.

Corresponding revisions have been made in the Abstract, Discussion, Limitations, and Conclusions and are clearly identified in red using Track Changes in the revised manuscript.

3. Denominators are missing in some instances—for example, regarding the 8,715 outpatient visits and 19,604 emergency department visits. What are the denominators?

Thank you for this observation. The 8,715 outpatient visits and 19,604 emergency department (ED) visits represent absolute numbers of varicella-related healthcare encounters recorded between 2019 and 2024 and were not calculated as population-based rates. Therefore, no population denominator was used for these totals.

For age-specific distributions, the denominator was the total number of healthcare encounters included in the corresponding analysis. In the combined outpatient and ED analysis presented in Figure 2, the denominator was 28,319 encounters (8,715 outpatient visits and 19,604 ED visits). These records represent healthcare encounters and not necessarily unique patients.

We have clarified the distinction between absolute encounter counts and population-based rates, as well as the denominator used for the age-specific distribution, in the Methods, Results, and Figure 2 legend. These revisions are clearly identified in red using Track Changes in the revised manuscript.

4) Diagnosis is primarily clinical, and laboratory confirmation is not systematic; this represents a significant methodological limitation.

Thank you for this important observation. We agree that the predominantly clinical diagnosis of varicella and the lack of systematic laboratory confirmation represent an important methodological limitation. As described in the Methods, routine surveillance primarily relies on clinically diagnosed cases, laboratory confirmation is not systematically required, and laboratory confirmation status was not consistently available in the analyzed datasets. In addition, individual medical records were not reviewed and no case-level diagnostic validation was performed.

We have therefore expanded the Limitations section to explicitly acknowledge the potential for diagnostic misclassification and to clarify that the records analyzed in this study should not be interpreted as systematically laboratory-confirmed cases. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Thank you for this important clarification. We agree that the term “varicella-associated death” may imply a causal relationship that cannot be established from the available data. The records correspond to deaths occurring during hospitalizations in which varicella (ICD-10 B01.0–B01.9) was recorded as a principal or secondary diagnosis. Individual medical records were not reviewed, and the cause of death was not independently verified; therefore, we cannot determine whether varicella caused or contributed to these deaths.

The operational definition has been clarified in the Methods, and the terminology has been revised throughout the Abstract, Results, Discussion, and Conclusions to avoid implying a causal relationship between varicella and death. These revisions are clearly identified in red using Track Changes in the revised manuscript.

6) It is unclear why the authors highlighted the 2023 outbreak. Please provide further explanation.

Thank you for this observation. The 2023 outbreak was highlighted because it was the only institutional outbreak identified in the surveillance and administrative sources available for this study and because it documented varicella transmission among adults in a closed setting. Its inclusion was intended to complement the national epidemiological and healthcare utilization data by illustrating a documented institutional transmission event.

We have clarified that this finding does not imply that it was the only varicella outbreak that occurred in Costa Rica during the study period, as community-, school-, and other non-institutional outbreaks could not be systematically identified from the available data. We have also limited the interpretation of this event to transmission in a closed institutional setting and have avoided extrapolating it to the frequency or risk of varicella outbreaks in the general population. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Thank you for this important observation. We agree that 95% should not be presented or interpreted as a universally applicable threshold for population protection against varicella. In our study, percentages around or above 95% reflect observed administrative vaccination coverage values rather than a threshold established by our analysis.

We have therefore removed the term “threshold” and revised references to 95% throughout the manuscript to avoid implying that this value guarantees population-level protection or represents a sufficient criterion for considering the introduction of a second vaccine dose. We have also removed the 95% reference line from the vaccination coverage figure. The revised Discussion now emphasizes that population-level protection depends on factors including pathogen transmissibility, vaccine effectiveness, and the proportion of the population that is immune, rather than on a universally applicable coverage threshold. These revisions are clearly identified in red using Track Changes in the revised manuscript.

8) Confidence intervals for the rates are missing; including them would certainly facilitate better comparisons.

Thank you for this observation. We agree that confidence intervals would provide additional information regarding the precision of the incidence estimates. Incidence rates in the present study are reported as point estimates. Although population denominators were available for the calculation of incidence rates, the datasets currently available for this revision do not contain the complete underlying annual and stratum-specific case counts and denominators required to retrospectively calculate valid confidence intervals for all reported rates. Therefore, we were unable to incorporate confidence intervals retrospectively.

We have acknowledged this limitation in the revised manuscript and have ensured that comparisons between incidence estimates are interpreted descriptively rather than as statistically demonstrated differences. This revision is clearly identified in red using Track Changes in the revised manuscript.

Thank you for this important observation. The objective of this study was to provide a descriptive epidemiological analysis of national varicella trends and healthcare utilization using aggregated surveillance and administrative data. Accordingly, the analyses were designed to characterize temporal, demographic, and geographic patterns rather than to identify individual-level risk factors for disease or hospitalization.

Although the available datasets included aggregated information on variables such as age, sex, and geographic location, we did not have access to linked individual-level data on vaccination status, comorbidities, socioeconomic characteristics, or other potential determinants required to appropriately evaluate factors associated with disease or hospitalization. Individual medical records were not reviewed. Therefore, a multivariable analysis of individual-level risk factors could not be reliably performed using the available datasets.

We have revised the study objective and Methods to more clearly define the descriptive scope of the analysis and have explicitly acknowledged the absence of individual-level data and inferential analyses as a limitation. Future studies linking surveillance, immunization, clinical, and sociodemographic data would be valuable for identifying factors associated with varicella infection and hospitalization. These revisions are clearly identified in red using Track Changes in the revised manuscript.

10) In the discussion, better compare the obtained results with those found in the literature.

Thank you for this important suggestion. We agree that comparison with previously published evidence is essential for interpreting our findings. The Discussion has been reviewed and strengthened to ensure that the principal findings are directly contextualized within the available literature.

Specifically, our findings are compared with the previous national study from Costa Rica [8], which evaluated the pre-vaccination and early post-vaccination periods, as well as with regional evidence from Latin America and the Caribbean [9–12]. The observed pediatric burden and hospitalization patterns are discussed in relation to previous Costa Rican reports [13,14], while healthcare utilization is contextualized using published evidence from other vaccination settings [15]. We also compare the broader epidemiological experience in Costa Rica with findings from universal vaccination programs in Uruguay [22–24], Brazil [25–27], Panama [28,30], and Argentina [29], as well as with evidence from two-dose vaccination programs [31,32].

These comparisons are interpreted in light of differences in observation periods, vaccination schedules, data sources, and study designs. In particular, we distinguish studies that evaluated changes before and after vaccine introduction from our descriptive analysis of an exclusively post-introduction period, thereby avoiding causal interpretation of temporal changes observed in our study. These revisions are clearly identified in red using Track Changes in the revised manuscript.

11) The conclusions should provide more precise guidance to policymakers.

Thank you for this important suggestion. We have revised the Conclusions to provide more specific guidance for policymakers while remaining within the scope of our descriptive findings. The revised Conclusions emphasize the following priorities: maintaining high, timely, and equitable first-dose vaccination coverage; improving the completeness and quality of vaccination and surveillance data; and strengthening preparedness and response to varicella transmission in closed institutional settings.

We also emphasize the need for better information on individual vaccination history and population susceptibility to inform the evaluation of targeted catch-up strategies and the potential contribution of a two-dose vaccination schedule in Costa Rica. These recommendations are presented as priorities for policy evaluation and future evidence generation rather than as interventions whose necessity or effectiveness has been established by the present study. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

Hidalgo-Mora et al collected epidemiological data of varicella infections in humans in the period of 2015-2024 in Costa Rica. Detailed data of human cases (hospitalized cases, reported cases) are shown. What I missed are: a., no data about sold vaccine units are given, no details of the vaccination program are given. e.g. which vaccine was used? when the program was started? etc, b., effect of the COVID pandemic on their data and conclusions are not discussed, c., the clear conclusions of the study in some points are also needed. d., the word „virus” is missing from the paper. Some features nomenclature name of the virus, description of the illness, symptoms should also be mentioned in teh discussion chapter. e.g. How were the human patients infected? Are there any data about this?

I also missed data about the distribution of human varicella cases on geographical map of Costa Rica. It would also be benefitable.

Abstract

line 24 – incidence means human cases with clinical signs?

The results here do not say too mch without some data of the vaccination (regular, compulsory, age of vaccinated people etc.

Introduction

The main information about the pathogen and the caused illness must be given in the introduction chapter of a paper with such a topic. The agent is virus, bacteria, fungi? Main symptoms, proportion of deaths, way of pathogen spreading etc. No such data are given for the readers. Some epidemiological data must be shown in a paragraph, and main features of the virus.

lines 57-58. Age groups of vaccinated children? Compulsory?

When the vaccination profgram started? How many vaccination units were sold? How number of used vaccinated changed over the years? Regional geograohical differences in vaccine usage in the country? Shown on a map or in a figure? Without thise, such data of the vaccination program other data given in this manuscript can not be evaluated.

Results

line 159 – when the vaccination program started? Without this, the data can not be evaluate.

3.1. - Mitigation of human cases during 2019-2022 was evidently the result of the epidemiologocal measures taken during the COVID pandemic. This decrease of varicella cases can not be considered as an isolated varicella epidemiological feature.

line 169 – Figure 1. The number of the sold vaccines annually should be shown here in Figure 1 parallel with human annual varicella cases data.

lines 175-180. These results can not be evaluate without theexact data of vaccination of age groups. Were these youngs vaccinated? What proportion of them? Again lack of vaccination data.

lines 201-203 – reasons?

Figure 3. Decrease of hospital discharges is sharp just in the COVID epidemic period. Should be mentioned and (probable) explained.

line 213 – detailes of the outbreak? Town, village, adults/children, hospitalization, how long didi t take etc.?

line 220. 3.6. - One of the main problems of this paper is details of vaccination. What does „coverage” mean? e.g. 95% of born infants were inoculated in the certain year? 95% of whole population of the country? Whithout this teh whole data of the manuscript can not be understand, evaluate. Who were vaccinated and when? Which vaccine?

Discussion

The discussion chapter is extremely long. 46% of all lines of the paper are in the Discussion/conclusion chapters. Authors usually tend to hide the lack of real message behind long pages of discussion. The authors should select their 2-3-4 main conclusions/messages, and in 1-1 paragraph discuss them.

I could not find in the dscussion chapter a logical line.

Lines 242-250 – the authors discuss „ national vaccination program” and geograhical distribution of cases, when no details about them were shown before.

line 252 -260. In a paragrapgh the authprs indicate a shift from children to adolescent and adult patients, in the next paradraph they say varicella infection is still a pediatric disease.

line 274- Are any data about vaccination background of the hospitalized patients avaiable?

319-328 – In the discussion chapter authors discuss their results. This part shuld be part of the introduction. Where a concise summary of the details and history of the varicella vaccination program in Costa Rica is missing. This program is not for discussing, as it is not result of the authors, but facts from the past.

Conclusions

line 380 – here main message(s) of the study should be summarized. Even in points. Not general meaningless empty sentences.

Spelling:

line 15 vaccination program

line 15 – Al-though

line 83 – Ja-nu-a-ry

line 182 – re-cor-ded

line 234 be-ha-vi-or

line 358 – va-ri-ed

Author Response

a., no data about sold vaccine units are given, no details of the vaccination program are given. e.g. which vaccine was used? when the program was started? etc,

Thank you for this observation. We have added further context regarding the national vaccination program. Costa Rica introduced universal childhood varicella vaccination in 2007 as a one-dose schedule, with vaccination provided free of charge at 15 months of age through the CCSS National Immunization Program [8]. As previously described by Ávila-Agüero et al. [8], vaccines have been procured through the PAHO Revolving Fund; the program initially used the GSK varicella vaccine and subsequently the MSD vaccine.

Data on vaccine sales were not available among the sources analyzed in the present study. Moreover, the objective of this study was to describe national epidemiological trends and healthcare utilization rather than vaccine procurement or program implementation. Because vaccine units sold or procured are not equivalent to doses administered or administrative vaccination coverage, we did not attempt to infer these values from the available coverage data. The manuscript has therefore been revised to provide the programmatic information necessary to contextualize the epidemiological analysis, while detailed procurement and sales data remain outside the scope of the study.

Referencia [8]: Ávila-Agüero ML, et al. Expert Rev Vaccines. 2017;16:229–234. doi:10.1080/14760584.2017.1247700.

b., effect of the COVID pandemic on their data and conclusions are not discussed,

Thank you for this important observation. We have explicitly addressed the potential influence of the COVID-19 pandemic on the observed epidemiological trends in the revised Discussion and Limitations. The marked decline in varicella incidence and healthcare utilization during 2020–2021 coincided with the pandemic and may have reflected multiple factors, including reduced social contact and school closures, changes in healthcare-seeking behavior, and potential underreporting.

Because the available aggregated data do not allow the relative contribution of these factors to be distinguished or quantified, we have interpreted the pandemic-period decline cautiously and have avoided attributing these changes directly to vaccination. We have also acknowledged the COVID-19 pandemic as an important limitation when interpreting temporal trends during the study period. These revisions are clearly identified in red using Track Changes in the revised manuscript.

c., the clear conclusions of the study in some points are also needed.

Thank you for this observation. We have revised the Conclusions to more clearly distinguish the principal findings of the study from their implications for public health policy. The revised Conclusions emphasize that, despite substantial reductions in reported incidence during the study period, varicella cases and healthcare utilization continued to be recorded, while the distribution of disease burden differed by outcome: reported cases remained concentrated in younger age groups, whereas hospitalizations became increasingly concentrated among adults.

We have also clarified the practical implications of these findings, including maintaining high, timely, and equitable first-dose vaccination coverage and improving the completeness, quality, and integration of vaccination and surveillance data. At the same time, we explicitly maintain the descriptive scope of the study and avoid causal attribution of the observed trends to vaccination. The potential contribution of a two-dose vaccination schedule and targeted catch-up strategies is therefore presented as an area requiring additional evidence on vaccination status and population susceptibility rather than as a need established by the present study. These revisions are clearly identified in red using Track Changes in the revised manuscript.

d., the word „virus” is missing from the paper. Some features nomenclature name of the virus, description of the illness, symptoms should also be mentioned in teh discussion chapter. e.g. How were the human patients infected? Are there any data about this?

Thank you for this observation. We have added the formal viral nomenclature and a brief description of the clinical presentation and transmission of varicella to provide additional biological and clinical context. Varicella is the primary infection caused by varicella-zoster virus (VZV), a human alphaherpesvirus, and is transmitted primarily through the respiratory route and direct contact with vesicular fluid [3]. The disease typically presents with fever and a generalized pruritic vesicular rash, although clinical severity and complications vary with age and host factors [3].

Regarding the source of infection in the patients included in our study, the national surveillance and administrative datasets did not contain individual-level information on exposure history or circumstances of transmission, and individual medical records were not reviewed. Therefore, we were unable to determine how individual patients acquired infection or identify specific transmission sources. We have clarified this limitation and have distinguished established general mechanisms of varicella transmission from exposure information directly available in our study. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Reference: Gershon AA, Breuer J, Cohen JI, Cohrs RJ, Gershon MD, Gilden D, et al. Varicella zoster virus infection. Nat Rev Dis Primers. 2015;1:15016. doi:10.1038/nrdp.2015.16.

I also missed data about the distribution of human varicella cases on geographical map of Costa Rica. It would also be benefitable.

Thank you for this helpful suggestion. We agree that the geographic distribution of reported varicella cases within Costa Rica is epidemiologically relevant. Geographic patterns were evaluated as part of our analysis, and regional differences in reported incidence were identified and summarized in the manuscript.

However, the primary objective of the present study was to provide an integrated national overview of contemporary varicella epidemiology and healthcare burden in Costa Rica and to contribute country-level evidence that can be interpreted within the broader Latin American context. Detailed geographic data are already routinely available to national public health authorities and are important for local surveillance and planning; however, presenting the complete regional distribution, including a geographic map and detailed province-level results, would substantially expand the manuscript beyond its principal objective. We therefore prioritized the epidemiological findings that we considered most relevant to the national and international interpretation of the study.

For this reason, geographic findings are presented at a summary level rather than as a detailed spatial characterization of varicella within Costa Rica. Nevertheless, the underlying geographic data can be provided as Supplementary Material if the Editor or reviewers consider that this additional level of detail would be useful to readers.

Abstract

line 24 – incidence means human cases with clinical signs?

The results here do not say too mch without some data of the vaccination (regular, compulsory, age of vaccinated people etc.

Thank you for these observations. We have revised the Abstract to clarify both the definition of reported incidence and the context of the national varicella vaccination program. The incidence estimates correspond to varicella cases reported through the national surveillance system, where diagnosis is primarily clinical and laboratory confirmation is not systematically required. Therefore, these estimates represent the incidence of reported varicella cases rather than all varicella-zoster virus infections in the population or exclusively laboratory-confirmed cases.

To provide the appropriate context for interpreting these findings, we have also clarified that Costa Rica introduced universal childhood varicella vaccination in 2007 as a one-dose schedule administered at 15 months of age through the publicly funded National Immunization Program. More detailed information on the vaccination program is provided in the Introduction, given the word limit of the Abstract. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Introduction

lines 57-58. Age groups of vaccinated children? Compulsory?

Thank you for these detailed observations. We agree that additional clinical, virological, and programmatic context is useful for interpreting the epidemiological findings. We have therefore expanded the Introduction to identify varicella-zoster virus (VZV) as the causative agent of varicella and to briefly describe its principal routes of transmission, typical clinical manifestations, complications, and greater severity of primary infection in adults. Relevant epidemiological context has also been retained to characterize the burden of varicella and the experience following vaccine introduction [1–4].

We have also clarified the characteristics of the Costa Rican vaccination program. Universal childhood varicella vaccination was introduced in 2007 as a one-dose schedule administered routinely at 15 months of age through the publicly funded National Immunization Program. The manuscript reports official first-dose administrative vaccination coverage over the years for which these data were available, allowing vaccination coverage trends to be considered alongside the epidemiological findings.

Data on vaccine sales were not among the sources analyzed in the present study. Moreover, vaccine units sold or procured are not equivalent to doses administered or administrative vaccination coverage; therefore, we did not attempt to infer these values from the available coverage data. Detailed vaccine procurement and product-specific utilization were not included because the objective of this study was to characterize national varicella epidemiology and healthcare burden rather than vaccine procurement or program implementation.

Results

line 159 – when the vaccination program started? Without this, the data can not be evaluate.

Thank you for this observation. Universal childhood varicella vaccination was introduced in Costa Rica in 2007. We have clarified this information in the Abstract and Introduction. Importantly, the present study covers 2015–2024 and therefore evaluates an exclusively post-introduction period. It does not provide a direct comparison between pre-vaccination and post-vaccination periods, and the observed temporal trends should not be interpreted as an independent estimate of vaccine program impact. These revisions are clearly identified in red using Track Changes in the revised manuscript

Thank you for this important observation. We agree that the marked decline observed during 2020–2021 must be interpreted in the context of the COVID-19 pandemic. We have therefore explicitly identified the pandemic period in the Results and expanded the Discussion to consider potential effects of reduced social contact and school closures, as well as changes in healthcare-seeking behavior and case reporting.

However, because these factors were not measured individually, our data do not allow us to determine their relative contributions or to attribute the observed decline exclusively to pandemic mitigation measures. We therefore describe the temporal coincidence while avoiding causal attribution to either the pandemic or vaccination. These revisions are clearly identified in red using Track Changes in the revised manuscript.

line 169 – Figure 1. The number of the sold vaccines annually should be shown here in Figure 1 parallel with human annual varicella cases data.

Thank you for this suggestion. Figure 1 presents the annual incidence of reported varicella cases by sex. Data on vaccine sales were not available among the sources analyzed in the present study and, importantly, vaccine units sold or procured are not equivalent to doses administered or to vaccination coverage. We therefore did not attempt to infer vaccine utilization from commercial sales data.

Vaccination is instead characterized using official first-dose administrative coverage reported by the CCSS National Immunization Program, which is presented separately in Figure 4. We consider administrative vaccination coverage the more appropriate population-level indicator for the objectives of this study and have therefore retained the incidence and vaccination coverage figures separately.

lines 175-180. These results can not be evaluate without the exact data of vaccination of age groups. Were these youngs vaccinated? What proportion of them? Again lack of vaccination data.

Thank you for this important question. Individual vaccination status was not routinely linked to reported varicella cases in the datasets available for this study, and individual medical records were not reviewed. We therefore cannot determine which of the reported cases had previously been vaccinated or calculate the proportion of cases occurring among vaccinated individuals.

Administrative vaccination coverage describes vaccination at the population level and cannot be used to infer the vaccination status of individual cases. Accordingly, we have avoided classifying these cases as breakthrough infections and have explicitly acknowledged the absence of linked individual vaccination histories as a limitation. These revisions are clearly identified in red using Track Changes in the revised manuscript.

lines 201-203 – reasons?

Thank you for this observation. The descriptive design of the study allows us to characterize temporal changes but does not allow their underlying causes to be established. We have therefore reviewed the Results to ensure that observed changes are presented descriptively, while potential explanations are addressed separately in the Discussion.

In particular, changes during the pandemic period may have reflected differences in transmission opportunities, healthcare-seeking behavior, and case reporting. However, these factors were not directly measured, and their individual contributions cannot be quantified using the available data. We have therefore avoided causal interpretations that are not directly supported by the study design. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Figure 3. Decrease of hospital discharges is sharp just in the COVID epidemic period. Should be mentioned and (probable) explained.

Thank you for this observation. We agree that the temporal relationship between the decline in hospitalizations and the COVID-19 pandemic should be explicitly acknowledged. We have revised the Results to indicate that hospitalizations declined to their lowest level in 2021, during the pandemic period.

Potential explanations, including reduced opportunities for varicella transmission and changes in healthcare-seeking behavior, hospital admission practices, and reporting, are considered in the Discussion. Because these factors were not directly measured, their relative contributions cannot be determined and the decline cannot be attributed exclusively to either the pandemic or vaccination. These revisions are clearly identified in red using Track Changes in the revised manuscript.

line 213 – detailes of the outbreak? Town, village, adults/children, hospitalization, how long didi t take etc.?

Thank you for this suggestion. We have expanded the description of the institutional outbreak to include the information available in the surveillance records. The outbreak occurred in 2023 in a correctional facility and involved 25 adult cases, with no reported deaths.

Additional characteristics, including the duration of the outbreak, number of hospitalizations, and individual vaccination histories, were not consistently available in the dataset analyzed for this study. We have therefore limited the description to information that could be verified from the available records and have avoided inferring unreported outbreak characteristics. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Thank you for this important observation. We have clarified the definition of vaccination coverage in the Methods. In this study, first-dose varicella vaccination coverage refers to the annual percentage of the target childhood population reported as having received the first dose of varicella vaccine through the national immunization program; it does not represent the proportion of the entire Costa Rican population vaccinated against varicella.

Varicella vaccine is routinely administered at 15 months of age as a one-dose schedule through the publicly funded National Immunization Program. Coverage estimates were obtained directly from official CCSS National Immunization Program reports and were not independently calculated by the investigators. We have revised the Methods accordingly to make the population represented by these estimates explicit. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Discussion

I could not find in the dscussion chapter a logical line.

Thank you for this important observation. We agree that the previous version of the Discussion was overly long and that the flow of ideas could be improved to make the main messages of the study more apparent. We have therefore substantially reorganized and streamlined the Discussion, removing or consolidating repetitive material and restructuring the sequence of ideas to provide a clearer progression from the principal findings to their interpretation and implications.

To further improve readability and organization, we have introduced thematic subheadings addressing: (1) trends in reported varicella incidence; (2) age distribution and healthcare burden; (3) vaccination coverage and future vaccination strategies; (4) the institutional outbreak and its implications for closed settings; and (5) strengths and limitations. Within these sections, the Discussion has been refocused on the main messages of the study, while background information already presented in the Introduction and unnecessary repetition of numerical results have been reduced. The strengths and limitations have also been consolidated and shortened while retaining the methodological considerations necessary for interpreting the findings.

We believe that these changes provide a clearer logical flow and make the principal findings and their implications more readily identifiable. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Lines 242-250 – the authors discuss „ national vaccination program” and geograhical distribution of cases, when no details about them were shown before.

Thank you for this observation. We agree that background information on the history and characteristics of the national varicella vaccination program is more appropriately presented in the Introduction. We have therefore consolidated this information there rather than discussing it as a finding of the present study.

Regarding geographic variation, geographic patterns were evaluated as part of our descriptive analysis and are now more clearly summarized in the Results. The Discussion has been revised to include only geographic interpretations directly supported by these findings and to avoid extending the interpretation beyond the level of geographic detail presented in the study. These revisions are clearly identified in red using Track Changes in the revised manuscript.

line 252 -260. In a paragrapgh the authprs indicate a shift from children to adolescent and adult patients, in the next paradraph they say varicella infection is still a pediatric disease.

Thank you for this important observation. We agree that the previous wording could be interpreted as suggesting a general shift of varicella toward older age groups. We have revised the Discussion to distinguish more clearly between the different outcomes evaluated in the study. Reported varicella cases remained concentrated in younger age groups, whereas adults, particularly those aged 20–44 years, accounted for a substantial proportion of hospitalizations and became the leading hospitalized age group by the end of the study period.

We have therefore avoided describing these findings as evidence of a general “age shift” in varicella epidemiology. We also acknowledge that the age categories differ in width and that hospitalization counts and proportions alone do not provide age-specific hospitalization risks. Accordingly, the findings are now interpreted as differences in the observed age distribution of reported cases and hospitalizations rather than as evidence of a higher individual risk of hospitalization among adults. These revisions are clearly identified in red using Track Changes in the revised manuscript.

line 274- Are any data about vaccination background of the hospitalized patients avaiable?

thank you for this important question. Individual vaccination status was not routinely linked to hospitalization records in the administrative datasets available for this study, and individual medical records were not reviewed. We therefore cannot determine the proportion of hospitalized patients who had previously received varicella vaccine or evaluate hospitalization according to vaccination status.

Accordingly, we have avoided attributing hospitalizations to lack of vaccination or inferring individual vaccination status from patient age or birth cohort. The absence of linked individual vaccination histories is also explicitly acknowledged as a limitation of the study. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Thank you for this observation. We agree that background information describing the introduction and characteristics of the national varicella vaccination program is more appropriately presented in the Introduction. We have therefore consolidated this information in the Introduction, where it provides the necessary context for interpreting the study findings. The Discussion has been revised to focus on the interpretation of the observed epidemiological patterns, comparison with previously published evidence, and the methodological limitations of the analysis. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Conclusions

line 380 – here main message(s) of the study should be summarized. Even in points. Not general meaningless empty sentences.

Thank you for this important observation. We have revised the Conclusions to focus more specifically on the principal findings and implications of the study. The revised Conclusions emphasize that reported varicella cases and healthcare burden persisted during the study period; that reported cases remained concentrated in younger age groups, whereas adults accounted for a substantial proportion of hospitalizations; and that the documented institutional outbreak illustrates the potential for transmission in closed settings.

We have also clarified the implications for future policy and research, emphasizing the importance of maintaining high and equitable first-dose vaccination coverage, strengthening surveillance and immunization data, and obtaining individual-level information on vaccination status and residual susceptibility before evaluating potential modifications to the national vaccination strategy, including a two-dose schedule. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors should explain why the data relating to the different outcomes have such different time or are unavailable (particularly for in-hospital deaths associated with varicella).
The Authors stated that first-dose vaccination coverage exceeded 100% in certain regions due to "numerator-denominator discrepancies". The Authors should better explain the reasons for this discrepancy.
The Authors should calculate the age-specific hospitalization rate (e.g., cases per 100,000 adults) to explain the observed increase.
The authors should discuss in more depth the lack of recent data on the incidence.
In the discussion section, the Authors should better clarify what do they mean by intersectoral public health response.
The need for catch-up vaccination campaigns for adults should be emphasized in the discussion.
In the conclusions section it should be highlighted that the findings on outbreaks support the conclusion (outbreak in a single center) only for closed facilities and not for the general community.

Author Response

The authors should explain why the data relating to the different outcomes have such different time or are unavailable (particularly for in-hospital deaths associated with varicella).

Thank you for this important observation. The study integrates several national surveillance and administrative datasets, and the period of availability differed according to the outcome. Specifically, reported incidence data were available for 2015–2022; outpatient and emergency department visits for 2019–2024; hospitalizations for 2015–2024; and in-hospital deaths with a recorded varicella diagnosis for 2018–2020. Analyses for each outcome were therefore restricted to the period for which the corresponding institutional data were available.

These differences reflect limitations inherent to the use of routinely collected surveillance and administrative data, for which availability, completeness, and continuity were not uniform across data sources and years. Changes in surveillance priorities and data collection practices, particularly during the COVID-19 pandemic, as well as other operational limitations, may have contributed to these gaps; however, the specific reason for the absence of data for each outcome and year could not be established from the available datasets.

We have revised the Methods and Limitations to make these different observation periods explicit and to clarify that the absence of data outside these periods should not be interpreted as the absence of events. In particular, mortality data were available to the investigators only for 2018–2020; therefore, mortality trends before 2018 or after 2020 cannot be evaluated using the present data. These revisions are clearly identified in red using Track Changes in the revised manuscript.

The Authors stated that first-dose vaccination coverage exceeded 100% in certain regions due to "numerator-denominator discrepancies". The Authors should better explain the reasons for this discrepancy.

Thank you for this important observation. We have clarified the interpretation of administrative vaccination coverage values exceeding 100%. These estimates are based on the relationship between the number of vaccine doses recorded for the target population and the estimated target population used as the denominator. Values exceeding 100% may therefore arise when the numerator and denominator do not correspond perfectly to the same underlying population.

Potential contributing factors include population mobility, vaccination occurring outside the area of residence, differences between population projections and the population actually served, inclusion of doses administered to individuals outside the estimated target cohort, and other reporting or data-quality issues. However, the available aggregated data do not allow us to determine which of these factors accounted for the values exceeding 100% in specific years or regions. We have therefore revised the manuscript to present these factors as possible explanations rather than established causes. These revisions are clearly identified in red using Track Changes in the revised manuscript.

The Authors should calculate the age-specific hospitalization rate (e.g., cases per 100,000 adults) to explain the observed increase.

thank you for this important observation. We agree that age-specific population-based hospitalization rates would allow a more appropriate comparison of hospitalization burden across age groups and over time. However, the data available for the present analysis did not include the complete age-specific population denominators required to calculate valid hospitalization rates for each year and age group. We were therefore unable to retrospectively calculate age-specific hospitalization rates.
We have revised the manuscript to clarify that the observed changes represent differences in absolute numbers of hospitalizations rather than changes in population-based hospitalization rates or individual hospitalization risk. Specifically, hospitalizations among adults aged 20–44 years increased by approximately 35% when comparing the 2015–2018 and 2021–2024 periods, while the increase from 7 hospitalizations in 2015 to 21 in 2024 represents a 200% increase between the first and last years of the study period. These findings should therefore be interpreted descriptively. The absence of age-specific hospitalization rates has also been explicitly acknowledged as a limitation. These revisions are clearly identified in red using Track Changes in the revised manuscript.

The authors should discuss in more depth the lack of recent data on the incidence.

Thank you for this important observation. National reported incidence data available for this study extended through 2022; therefore, we cannot determine the national incidence trend during 2023–2024 or directly compare incidence with the vaccination coverage estimates reported for those years. Although more recent data were available for outpatient visits, emergency department visits, and hospitalizations, these outcomes represent healthcare utilization and cannot be used as substitutes for population-based incidence.
We have clarified this distinction throughout the manuscript and expanded the Limitations to emphasize that incidence trends should not be extrapolated beyond 2022. Accordingly, vaccination coverage estimates for 2023–2024 are presented descriptively and are not interpreted as corresponding to contemporaneous national incidence trends. These revisions are clearly identified in red using Track Changes in the revised manuscript.

In the discussion section, the Authors should better clarify what do they mean by intersectoral public health response.

Thank you for this observation. We agree that the term “intersectoral public health response” was insufficiently specific. In this context, the term was intended to refer to coordinated outbreak management involving health authorities and the administration of the correctional facility. However, the available aggregated data do not provide sufficient detail to characterize the specific components or effectiveness of this response. We have therefore revised the Discussion to avoid implying specific interventions that were not directly documented in the data available for this study and to focus instead on the need for coordinated preparedness and outbreak response in closed institutional settings. These revisions are clearly identified in red using Track Changes in the revised manuscript.

The need for catch-up vaccination campaigns for adults should be emphasized in the discussion.

Thank you for this important suggestion. We agree that the substantial representation of adults among hospitalizations and the documented outbreak in a correctional facility warrant consideration of catch-up vaccination strategies for adults without evidence of immunity, particularly in closed or other higher-risk settings. However, individual vaccination histories and seroprevalence data were unavailable in our study; therefore, we cannot estimate the proportion of susceptible adults or establish the need for a population-wide adult catch-up vaccination campaign.

We have revised the Discussion and Conclusions to emphasize the need to evaluate targeted catch-up vaccination strategies for adolescents and adults without evidence of immunity, while recognizing that their implementation should be informed by improved characterization of vaccination history and residual susceptibility. These revisions are clearly identified in red using Track Changes in the revised manuscript.

In the conclusions section it should be highlighted that the findings on outbreaks support the conclusion (outbreak in a single center) only for closed facilities and not for the general community.

Thank you for this important observation. We agree that the documented outbreak, which occurred in a single correctional facility, supports conclusions regarding the potential for varicella transmission in closed institutional settings but should not be generalized to the broader community. We have revised the Conclusions to explicitly distinguish this finding from the national epidemiological and healthcare utilization findings. The revised text clarifies that this single institutional event cannot be used to infer the frequency or risk of outbreaks in the general community. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

See attachment for comments

Comments for author File: Comments.pdf

Author Response

Thank you for the favorable assessment of our manuscript and for recognizing its contribution to the epidemiological evaluation of varicella in Latin America. We agree that the findings should be interpreted within the context of Costa Rica’s long-standing universal one-dose vaccination program. However, we have clarified throughout the manuscript that the present study describes epidemiological trends during a period entirely after vaccine introduction and was not designed to estimate the causal effect of the vaccination program. Specifically, the approximately 84% reduction in reported varicella incidence represents the observed change between 2015 and 2022 and should not be interpreted as a pre- versus post-vaccination reduction.

We have also clarified the interpretation of the hospitalization findings among adults aged 20–44 years. Hospitalizations in this age group increased from 37 during 2015–2018 to 50 during 2021–2024 (approximately 35%), while annual hospitalizations increased from 7 in 2015 to 21 in 2024, representing a 200% increase between the first and last years of the study period. Because individual vaccination histories were unavailable and hospitalization outcomes were analyzed using counts rather than age-specific population-based rates, our data cannot establish that hospitalized adults were unvaccinated or that their individual risk of hospitalization increased. The manuscript has been revised to make these distinctions explicit. These revisions are clearly identified in red using Track Changes in the revised manuscript.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript is of little interest and does not represent an advancement in knowledge, but following revisions, it is clearer and more transparent.

Reviewer 2 Report

Comments and Suggestions for Authors

If authors could not get access to some important data of vaccination program and geographical distribution of human cases, it should be mentioned somewhere in the paper. As I see unfortunately whithout these data vaccinia infections of people of Costa Rica cen not be evaluated properly. But as they answered my suggestions and remarks, and in many parts the authors made changes in the manuscript, I accept their answeres and I suggest publication of the manuscript in its present form.

Reviewer 3 Report

Comments and Suggestions for Authors