Keywords
Congenital hereditary and neonatal diseases and abnormalities
Prediction algorithms
Health planning
How to Cite
Copyright © | All rights reserved.
The journal holds the exclusive copyright for the publication of this article under the terms of Brazilian law 9610/98.
Partial reproduction
The use of parts of the texts, figures and questionnaire of the article is free, being mandatory the citation of the authors and journal.
Total reproduction
It is expressly prohibited and must be authorized by the journal.
Abstract
Objective: To estimate the occurrence of new cases of congenital syphilis in southern Brazil, using the Seasonal Autoregressive Integrated Moving Average (SARIMA) model, the Long Short-Term Memory (LSTM) Neural Network, the Complex Exponential Smoothing (CES) model and a hybrid CES-SARIMA-LSTM model. Methods: Data from the Notifiable Diseases Information System were used, referring to monthly notifications registered between January 2014 and June 2024. The evaluation of the models was conducted using the metrics Mean Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE), implemented in Python. Results: The Augmented Dickey-Fuller test showed a p-value of 0.008, indicating a stationary time series; the Mann-Kendall test indicated an increasing trend, and the Ljung-Box Q test revealed a p-value of 2.51e-55, confirming autocorrelation in the time series data. The SARIMA (0,1,1) (0,0,1) model [12] demonstrated superiority to the LSTM, CES, and triple model in all error metrics. SARIMA's forecasts indicate a steady relatively linear reduction in monthly cases from July 2024 to June 2027, going from 243 to 228 reported cases. Conclusion: Government actions are still needed to prevent an increase in cases and predictive models can be an excellent auxiliary tool for this purpose.
References
ALVES CC, et al. Aplicação de métodos estatísticos com suavização exponencial dupla e tripla para previsão de demanda na gestão de estoques. Revista Produção Online: Revista Eletrônica de Engenharia de Produção e Correlatas. 2019; 19(3): 1001-26.
ARAUJO TKA, et al. Tendência e predição da sífilis congênita no Brasil no século XXI: Um estudo de séries temporais. Revista Gaúcha de Enfermagem, 2026; 47: e20240166.
AZIZ AA, et al. A hybrid simple exponential smoothing-barnacles mating optimization approach for parameter estimation: Enhancing COVID-19 forecasting in Malaysia. MethodsX, 2025; 14: 103347.
BERLATTO KC, et al. Sífilis Congênita na Região Sul do Brasil (2018-2022): Perfil Epidemiológico com Discussão das Possíveis Alterações Anatômicas Associadas. Brazilian Journal of Implantology and Health Sciences, 2025; 7(1): 357-368.
BRASIL. 2026. Sífilis congênita. Disponível em: https://www.gov.br/saude/pt-br/assuntos/saude-de-a-a-z/s/sifilis/gestantes/congenita. Acesso em: 8 maio 2026.
CARNEIRO, DO, et al. Utilização do modelo de suavização exponencial para avaliação temporal da situação epidemiológica da sífilis congênita no Brasil. Diálogos & Ciência, 2023; 3(1): 19-34.
CHEN L. Application progress of ensemble forecast technology in influenza forecast based on infectious disease model. Frontiers in Public Health, 2023; 15(11): 1335499.
COOPER JM, et al. In time: the persistence of congenital syphilis in Brazil - More progress needed!. Em tempo: a persistência da sífilis congênita no Brasil - Mais avanços são necessários!. Revista paulista de pediatria, 2016; 34(3): 251–253.
ECHEGARAY F, et al. Repercussions of the COVID- 19 pandemic on maternal and congenital syphilis in South Brazil: a time series analysis 2010-2022. BMC Infectious Diseases, 2025; 25(528).
FENG S, et al. Development and validation of an individualised nomogram to predict mother-to-child transmission in pregnant women with syphilis in China: a retrospective cohort study. BMJ Open, 2024; 14(4): e080211.
KUMAR S, et al. Forecasting the spread of COVID‐19 using LSTM network. BMC Bioinformatics, 2021; 22(6): 316.
LIN Y, et al. The trend analysis of HIV and other sexually transmitted infections among women of childbearing aged 15 to 49 years from 1990 to 2021 and its forecasting from 2022 to 2030. Front. Public Health, 2025; 13: 1702289.
OLIVEIRA E, CRUZ DM, et al. Inteligência epidemiológica, investimento em tecnologias da informação e as novas perspectivas para o uso de dados na vigilância em saúde. Cadernos de Saúde Pública, 2024; 40(8): e00160523
ORSSATTO AJ, et al. Aplicação de séries temporais na estimação do número de casos de dengue em Cascavel-PR. Revista Brasileira de Computação Aplicada, 2022; 14(3): 37-50.
OTHMAN M, et al. Model Forecasting Development for Dengue Fever Incidence in Surabaya City Using Time Series Analysis. Processes, 2022; 10(11): 2454.
PAVINATI G, et al. Temporal analysis of gestational and congenital syphilis indicators in Brazil: toward the elimination of vertical transmission by 2030? Revista Brasileira de Epidemiologia, 2025; 28: e250028.
RAIMUNDO DML, et al. Análise espaço-temporal da sífilis congênita no Rio Grande do Norte, Brasil, de 2008 a 2022. Revista de Epidemiologia e Controle de Infecção, 2025; 15(3).
RAIMUNDO DML, et al. Sífilis congênita: análise de tendência temporal e projeção de casos. Acta Paulista de Enfermagem, 2025; 38.
SALOMÃO R. Infectologia: Bases Clínicas e Tratamento. 2 ed. Guanabara Koogan: Rio de Janeiro - RJ, 2023; p77.
SILVA WS, et al. Seleção de um método de previsão de demanda oriundo da aquicultura na Bacia do Rio doce em Minas Gerais (MG) e Espírito Santo (ES). Revista Produção Online, 2025; 25(3): 5336.
SOUZA IG. Análise temporal das internações por complicações da sífilis na Bahia, no período de 2000 a 2020. Revista Eletrônica Acervo Saúde, 2024; 24(6): 1-10.
SVETUNKOV I, et al. Complex exponential smoothing. Naval Research Logistics, 2022; 69(8): 1108-1123.
TANG X, et al. The evaluation of preventive and control measures on congenital syphilis in Guangdong Province, China: a time series modeling study. Infection, 2022; 50: 1179–1190.
TEIXEIRA IV, et al. Predicting congenital syphilis cases: A performance evaluation of different machine learning models. PLoS ONE, 2023; 18(6): 1-25.
WANG ZD, et al. Analysis and forecasting of syphilis trends in mainland China based on hybrid time series models. Cambridge University Press of Epidemiology and Infection, 2024; 152(93): 1-11.
YAGNIK H, et al. Deep Learning for Air Pollution Prediction: A Systematic Review of CNN, LSTM, GNN, and Transformer Architectures. Arch Computat Methods Eng, 2026.
YOUNG SD, et al. Using social media as a tool to predict syphilis. Preventive Medicine, 2018; 109: 58-61.