Data on dengue incidence in South-eastern Brazil, 2014–2018

Data from the routine surveillance systems have been extensively used to estimate the incidence of dengue. However, routine surveillance data frequently underestimate the diseases’ incidence. Underreporting of dengue cases is related to the varying spectrum of its clinical presentation, with a large proportion of mild and asymptomatic infections, to its unspecific signs and symptoms, to the limitations of access to health care, and to the performance of the surveillance system itself [1–3]. In order to obtain accurate figures on dengue incidence, a cohort of children and adolescents was set up and followed during four years. The incidence of reported cases was used as a reference for the sample size calculation, which was stratified by age groups. A two-stage procedure was used to select the participants: census tracts were randomly selected, and within each one, a pre-determined number of children of each age group was randomly selected. The parents or legal guardians of the participating children and adolescents provided a written informed consent. In the first home visit, they responded to a questionnaire containing data on socio-demographic characteristics, housing, access to water, sewage, and garbage collection. Also, during the first visit a blood sample of the participating child/adolescent was collected for dengue baseline serology. Beginning in the week after the enrolment, the parent or legal guardian that was designated in the first visit received weekly phone calls for fever surveillance. If the child/adolescent had fever during the week, a nurse was dispatched to the family's home to collect more detailed data on the fever episode and collect a blood sample for dengue diagnosis (IgG, IgM, NS1 and PCR). If the dengue diagnosis was confirmed, a medical appointment was scheduled, and another blood sample for confirmatory tests was collected. It was also agreed that in every anniversary of their participation, they would receive another visit for a blood collection for dengue serology, regardless if they had a fever episode or a confirmed dengue diagnosis during the previous year. This article contains the description of the cohort's dataset. It is associated with the article published in Acta Tropica, under the title “A cohort study to assess the incidence of dengue, Brazil, 2014–2018” [4]. The associated article focused on the seroprevalence and incidence of dengue, and explored some associations between both outcomes and some explanatory variables.

serology. Beginning in the week after the enrolment, the parent or legal guardian that was designated in the first visit received weekly phone calls for fever surveillance. If the child/adolescent had fever during the week, a nurse was dispatched to the family's home to collect more detailed data on the fever episode and collect a blood sample for dengue diagnosis (IgG, IgM, NS1 and PCR). If the dengue diagnosis was confirmed, a medical appointment was scheduled, and another blood sample for confirmatory tests was collected. It was also agreed that in every anniversary of their participation, they would receive another visit for a blood collection for dengue serology, regardless if they had a fever episode or a confirmed dengue diagnosis during the previous year. This article contains the description of the cohort's dataset. It is associated with the article published in Acta Tropica, under the title "A cohort study to assess the incidence of dengue, Brazil, 2014 e2018" [4]. The associated article focused on the seroprevalence and incidence of dengue, and explored some associations between both outcomes and some explanatory variables.
©  Table 1 presents the description of the dataset, with the list of variables and categories/types of responses. The variables were grouped according to the moment when data were collected: at baseline interview, annual visits on the anniversary of participation, and during the fever episodes. At the baseline interview, the procedures were presented to parents or legal guardians, and an invitation was made for their son/daughter to participate. In the event of agreement on participation, the Informed Consent Form was read and signed. For older children and adolescents their signature of the Assent Form was also collected. The questionnaire contained data on the characteristics of the household, number of people living in the household, number of rooms, and sanitation. A blood sample was drawn from the participating child/adolescent for dengue serology. The family was asked to designate the member that would receive the weekly phone calls for fever surveillance. Beginning in the week after enrolment, this family member started receiving weekly phone calls for fever surveillance. If the participating child/adolescent had fever, a nurse would come to the family's home to collect a blood sample for dengue diagnosis, and to fill a questionnaire on dengue signs and symptoms. If the child/ adolescent had a dengue diagnosis, a medical appointment was set. During that appointment, an -RNA was extracted from respectively 0.4 mL of plasma on the automated platform iPrep using the PureLink Virus Kit (Life Technologies, Brazil) or 0.5 mL in the NucliSENS Easy-Mag (Biomerieux, Brazil). -Extracted nucleic acids were submitted to a one-step dengue generic real-time polymerase chain reaction employing primers and probe previously described and TaqMan Fast Virus 1-Step Master Mix (ThermoFisher, Brazil). -Reactive samples were typed by submitting extracted RNA to four type-specific reactions, using the same conditions as above but replacing generic primers/probe by type-specific reagents [5]. Annual visits on the anniversary of participation: blood sample for IG antibodies, as described above. Data  Value of the Data Dengue is one of the major emerging infectious diseases worldwide. Global warming has led to the spread of insect vectors to more northern and southern latitudes. The present dataset tracks the incidence of dengue fever at a site initially characterized as low endemic. These data may be useful to the dengue research community, to policymakers in the public health field, and to the industry developing products related to dengue prevention and control.

Data description
Our dengue cohort has provided data on the disease prevalence, incidence, diagnosis, clinical presentation and explored some associations between both outcomes and some explanatory variables. These data can be compared to similar studies, and can be aggregated with other similar datasets, increasing the sample size.
The outcome of a large outbreak of dengue was documented additional blood sample was collected. Every anniversary of their participation they would receive a visit for a blood sample collection, regardless of having had a fever episode during the year. The dataset contains also some new variables that were created from the original ones, to facilitate the analysis. They are: a) Confirmed dengue case e 0/1 e if the participant had a fever episode in which acute dengue was diagnosed. b) DENV 1e0/1 e for the ones who had an acute dengue diagnosis, if the genotype was dengue virus 1 (DENV1). c) DENV 2e0/1 e for the ones who had an acute dengue diagnosis, if the genotype was dengue virus 2 (DENV2). d) Asymptomatic dengue infection e 0/1 e if the participant had a dengue seroconversion detected in any of the annual serologic surveys, and did not have a fever episode in which the dengue diagnosis was confirmed. e) Final date e necessary to calculate the follow up time. Table 2 presents the cohort's planned sample size. The data on age-specific incidence of reported cases were used to estimate the sample size, stratified into four age groups (2e5 years of age, 5 to 9, 10 to 13, and 14 to 16).
Three criteria were applied confirm dengue as the aetiology of the fever episodes. Table 3 presents the distribution of the confirmed cases according to them. The majority of cases were confirmed by the detection of dengue virus RNA by RT-PCR. A small proportion was confirmed by the detection of dengue virus NS1 antigen in serum samples, and just four cases were confirmed by seroconversion of IgM and IgG antibodies. Table 4 presents the frequency of symptoms and signs among the 308 confirmed dengue cases. The more frequently reported symptoms were loss of appetite, myalgia, and abdominal pain. The frequency of warning signs, such as intense vomiting and fluid accumulation was low.

Design
In order to overcome the limitations of routine surveillance systems in estimating the burden of dengue [1e3] a cohort was set up to determine dengue incidence in children and adolescents from 2 to 16 years of age [4].

Site
It was conducted in Araraquara (Population 212,617 in 2012), a city located in the central region of the State of São Paulo (21 47 0 S; 48 10' W), a location classified as a mid level endemic location for dengue. The first autochtonous dengue case in the city was reported in 1995, and the largest outbreak prior to the period of observation occurred in 2011, when 2500 confirmed dengue cases were reported.

Sample size
Dengue incidence in Brazil peaks in young adults [6] and is low among children. The age-specific incidence of dengue reported cases was used as a reference for the sample size calculation. The planned sample size can be seen in Table 2.

Sample selection procedure
A two-stage strategy was used to select the actual participants. In the first stage census tracts were randomly selected. In the selected census tracts the households with children and/or adolescents in the age range of interest were identified. Then, a pre-determined number of children/adolescents was randomly selected.

Ethical issues
The protocol was approved by the research ethics review board of the Hospital das Clínicas, of the University of São Paulo's Medical School, and is registered at the National Research Ethical Evaluation System, approval number CAAE25706913.6.1001.0065. The parents or legal guardians of the participating children and adolescents provided a signed Informed Consent. Older children and adolescents provided a written Informed assent.

Procedures
Selected households received a first home visit in which the cohort's aims and procedures were presented to the parents or legal guardians, and the invitation to the participation of their son/daughter was made. When they agreed to participate, informed consent and assent were obtained, and a questionnaire was filled, as previously described.
After enrolment, the designated family member received weekly phone calls asking if the child/ adolescent has or had a fever during the week. A clinical thermometer was given to the participating family. From the second year on, they could opt for weekly text messages, instead of calls. When the participant reported a fever episode, a nurse would visit the family's home to collect a blood sample for   acute dengue diagnosis, and fill a questionnaire on the signs and symptoms accompanying the fever. When the dengue diagnosis was confirmed, a medical appointment was scheduled. During that appointment another blood sample was collected. On every anniversary of their participation they received another nurse's visit for a blood collection for dengue serology.

Definitions
Dengue suspect case: any fever case, with axillary temperature above 37.5 . Symptomatic dengue confirmed case: a suspect case with at least one of the following criteria fulfilled in acute and/or convalescent serum sample: 1: Detection of dengue RNA by real time polymerase chain reaction (RT-PCR). 2: Detection of NS1 protein. 3: Detection of IgM antibodies with IgG seroconversion. The number of cases confirmed by each criterion is presented in Table 3. The signs and symptoms presented by the confirmed dengue cases are presented in Table 4.
Asymptomatic dengue infections -Naïve subjects for dengue antibodies at baseline that presented seroconversion (at the one year serology sample) that did not report a febrile episode during the year were considered unapparent dengue infections. The same procedure was applied for those who seroconverted during the four-year follow up.

Outcomes
The main outcomes of the cohort follow-up were the dengue baseline seroprevalence, the annual seroprevalence, the cumulative incidence and incidence density of symptomatic confirmed dengue cases. Dengue baseline seroprevalence was calculated as the proportion of positive results in the recruitment sample. Yearly seroprevalence was calculated as the proportion of positive results in the yearly serologic surveys. Symptomatic confirmed dengue incidence was calculated as the number of symptomatic confirmed dengue cases divided by the population at risk: in the first year, the number of enrolled participants. For the following years, the number of subjects that participated in the yearly surveys was used as the denominator. Incidence density was calculated as the number of confirmed dengue cases in the numerator divided by the person-time units in the denominator. Asymptomatic dengue infections incidence was calculated as the number of asymptomatic dengue infections divided by the population at risk.

Laboratory methods
Dengue IgM and IgG antibodies were tested, using an ELISA assay (DengueVirus IgM Capture DxSelect TM and Dengue ELISA IgG, FOCUS Technologies, Cypress, CA, USA), according to the manufacturer instructions.
Blood was drawn in Plasma Preparation Tubes (PPT, Beckton-Dickinson, Brazil) and centrifuged within 4 hours. RNA was extracted from respectively 0.4 mL of plasma on the automated platform iPrep using the PureLink Virus Kit (Life Technologies, Brazil) or 0.5 mL in the NucliSENS Easy-Mag (Biomerieux, Brazil). On both methods, total nucleic acids were eluted into 50uL of buffer.
Extracted nucleic acids were submitted to a one-step dengue generic real-time polymerase chain reaction employing primers and probe previously described [7] and TaqMan Fast Virus 1-Step Master Mix (ThermoFisher, Brazil). Forty cycles of 95 C 10 seconds -60 C 40 seconds were performed in the thermocyclers models 7300 or 75-00 Fast (Applied Biosystems). Ct values below 37 were considered reactive for dengue RNA. Reactive samples were typed by submitting extracted RNA to four typespecific reactions using the same conditions as above but replacing generic primers/probe by typespecific reagents [5].