There is currently no system to track the emergence of Zika virus (ZIKV) subtypes. We developed a surveillance system able to retrieve sequence submissions and further classify distinct ZIKV genotypes in the world. This approach was able to detect a new occurrence of ZIKV from an African lineage in Brazil in 2019.

Zika virus was responsible for the microcephaly epidemic in Brazil which began in October 2015 and brought great challenges to the scientific community and health professionals in terms of diagnosis and classification. Due to the difficulties in correctly identifying Zika cases, it is necessary to develop an automatic procedure to classify the probability of a CZS case from the clinical data. This work presents a machine learning algorithm capable of achieving this from structured and unstructured available data. The proposed algorithm reached 83% accuracy with textual information in medical records and image reports and 76% accuracy in classifying data without textual information. Therefore, the proposed algorithm has the potential to classify CZS cases in order to clarify the real effects of this epidemic, as well as to contribute to health surveillance in monitoring possible future epidemics.

Background Prenatal exposure to ZIKV has potential teratogenic effects with a wide spectrum of clinical presentation called congenital Zika syndrome (CZS). There are limited data on survival of children with CZS, we estimated mortality comparing live births with and without CZS. Methods A population-based cohort study using linked routinely collected data in Brazil, from January 2015 to December 2018. Kaplan-Meier and survival analyses were performed adjusted for confounding and stratified by gestational age, birth weight and small for gestational age. Results We followed 11,737,554 live births for up to 36 months. The mortality rate among live births with CZS was 52.6 (95% confidence interval [CI] 47.6-58.0) and among those without CZS it was 5.6 (5.6-5.7) per 1000 person-years. The mortality rate ratio was 11.3 (95%CI 10.2-12.4) times higher among live births with CZS than those without CZS up to the age of 36 months. For infants born before 32 weeks' gestation or with birth weight less than 1500g, the risk of death was similar regardless of CZS. Infants with CZS born at term (mortality rate with and without CZS—38.4 vs 2.7) or with birth weight greater than 2499g (mortality rate with and without CZS—32.6 vs 2.5) were 14.3 (95%CI 12.4-16.4) and 12.9 (95%CI 10.9-15.3) times more likely to die than those without CZS. The burden of congenital anomalies, diseases of the nervous system and infectious diseases, as recorded causes of deaths, were higher among the CZS group. Conclusion There is a higher mortality risk in live births with CZS than live births without CZS that persists throughout the first three years of life.

The co-circulation of different arboviruses in the same time and space poses a significant threat to public health given their rapid geographic dispersion and serious health, social, and economic impact. Therefore, it is crucial to have high quality of case registration to estimate the real impact of each arboviruses in the population. In this work, a Vector Autoregressive (VAR) model was developed to investigate the interrelationships between discarded and confirmed cases of dengue, chikungunya, and Zika in Brazil. We used data from the Brazilian National Notifiable Diseases Information System (SINAN) from 2010 to 2017. There were three peaks in the series of dengue notification in this period occurring in 2013, 2015 and in 2016. The series of reported cases of both Zika and chikungunya reached their peak in late 2015 and early 2016. The VAR model shows that the Zika series have a significant impact on the dengue series and vice versa, suggesting that several discarded and confirmed cases of dengue could actually have been cases of Zika. The model also suggests that the series of confirmed and discarded chikungunya cases are almost independent of the cases of Zika, however, affecting the series of dengue. In conclusion, co-circulation of arboviruses with similar symptoms could have lead to misdiagnosed diseases in the surveillance system. We argue that the routinely use of mathematical and statistical models in association with traditional symptom-surveillance could help to decrease such errors and to provide early indication of possible future outbreaks. These findings address the challenges regarding notification biases and shed new light on how to handle reported cases based only in clinicalepidemiological criteria when multiples arboviruses co-circulate in the same population.