Time series analysis (2015-2022): dengue risk scenario in Goiás, Brazil

Oliveira Junior, Ronaldo Rodrigues de; Modesto, Ana Carolina Figueiredo; Assunção, Leandro do Prado; Pena, Lindomar José; Feres, Valéria Christina de Rezende

doi:10.1590/1980-549720260010

Tabla de contenido Anterior Actual Siguiente

Tabla de contenido

ORIGINAL ARTICLE • Rev. bras. epidemiol. 29 • 2026 • https://doi.org/10.1590/1980-549720260010 linkcopy

Time series analysis (2015-2022): dengue risk scenario in Goiás, Brazil

Authorship SCIMAGO INSTITUTIONS RANKINGS

Objective: To analyze confirmed cases of dengue in Goiás, Brazil between 2015 and 2022 and to estimate the risk of new outbreaks until 2026.

Methods: A time series study using data from the Notifiable Diseases Information System (SINAN) was conducted. Monthly records of cases confirmed by laboratory or clinical-epidemiological criteria were included. The Seasonal Autoregressive Integrated Moving Average (SARIMA) model was applied using the R software (v.4.2.1). Stationarity, trend, seasonality, residual autocorrelation, and model fit were evaluated, with estimates obtained by maximum likelihood and 95% confidence intervals.

Results: During the study period, 709,270 confirmed cases were recorded. Epidemics occurred cyclically every two years, with peaks in 2015 (101,261), 2016 (82,077), 2018 (70,794), 2019 (107,589), and 2022 (189,998), interspersed with years of lower incidence such as 2017 and the COVID-19 pandemic years (2020-2021). Serotype replacement was observed preceding major outbreaks. The SARIMA model showed good fit (Akaike Information Criterion - 1768.9; Bayesian Information Criterion - 1786.8) and predicted new peaks in 2025 (177,775 cases) and 2026 (224,100 cases).

Conclusion: Dengue in Goiás displayed recurrent epidemic cycles, pointing to an increase in cases and reinforcing the need for integrated strategies based on prevention and control. The SARIMA model proved useful for surveillance and public health planning, although its accuracy may be influenced by external factors.

Keywords:
Dengue; Epidemiological models; Time series; Forecasting; Epidemiology

Age group (years)	SARIMA*	AIC	BIC	Statistical test	p-value^{†,‡,§,//}
0-9	(0,1,2)(2,1,3)	1,290.088	1,313.167	Dickey-Fuller	0.010
				Mann-Kendall	0.079
				Kruskal-Wallis	0.000
				Box-Pierce	0.996
10-19	(0,1,3)(3,1,3)	1,403.171	1,428.815	Dickey-Fuller	0.010
				Mann-Kendall	0.942
				Kruskal-Wallis	0.000
				Box-Pierce	0.463
20-29	(0,1,1)(1,1,3)	1,450.105	1,468.055	Dickey-Fuller	0.010
				Mann-Kendall	0.543
				Kruskal-Wallis	0.000
				Box-Pierce	0.517
30-39	(0,1,1)(1,1,3)	1,429.412	1,447.362	Dickey-Fuller	0.010
				Mann-Kendall	0.697
				Kruskal-Wallis	0.000
				Box-Pierce	0.508
40-49	(0,1,1)(1,1,3)	1,404.614	1,420.000	Dickey-Fuller	0.010
				Mann-Kendall	0.423
				Kruskal-Wallis	0.000
				Box-Pierce	0.954
50-59	(0,1,2)(1,1,3)	1,372.779	1,390.729	Dickey-Fuller	0.010
				Mann-Kendall	0.478
				Kruskal-Wallis	0.000
				Box-Pierce	0.864
60-69	(0,0,3)(3,0,0)	1,293.124	1,311.074	Dickey-Fuller	0.010
				Mann-Kendall	0.482
				Kruskal-Wallis	0.000
				Box-Pierce	0.807
70-79	(0,0,3)(3,0,0)	1,175.009	1,195.524	Dickey-Fuller	0.010
				Mann-Kendall	0.271
				Kruskal-Wallis	0.000
				Box-Pierce	0.709
80-89	(0,0,3)(3,0,0)	961.397	981.912	Dickey-Fuller	0.010
				Mann-Kendall	0.142
				Kruskal-Wallis	0.000
				Box-Pierce	0.911
90-100	(0,1,1)(3,1,1)	682.687	700.638	Dickey-Fuller	0.010
				Mann-Kendall	0.077
				Kruskal-Wallis	0.000
				Box-Pierce	0.429

location_on

Associação Brasileira de Saúde Coletiva Av. Dr. Arnaldo, 715 - 2º andar - sl. 3 - Cerqueira César, 01246-904 São Paulo SP Brasil , Tel./FAX: +55 11 3061-8055 - São Paulo - SP - Brazil
E-mail: epidemio_submissao@abrasco.org.br

rss_feed Acompañe los números de esta revista en su lector de RSS

[TFN1] SARIMA: Seasonal Autoregressive Integrated Moving Average; AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion; *SARIMA Model; ^†Dickey-Fuller Test; ^‡Mann-Kendall Test; ^§Kruskal-Wallis Test; ^//Box-Pierce Test.