[Downloaded free from http://www.ijph.in on Tuesday, June 29, 2021, IP: 183.171.126.

85]

Original Article

Forecasting Incidence of Dengue in Rajasthan,

Using Time Series Analyses
Sunil Bhatnagar1, *Vivek Lal2, Shiv D. Gupta3, Om P. Gupta4
1
OSD (ME), Government of Rajasthan, 2Assistant Professor, Institute of Health Management Research, 3Director, Institute of Health
Management Research, Jaipur, 4Director (Public Health), Directorate of Medical Health and Family Welfare, Government of Rajasthan, India

Abstract
Aim: To develop a prediction model for dengue fever/dengue haemorrhagic fever (DF/DHF) using time series
data over the past decade in Rajasthan and to forecast monthly DF/DHF incidence for 2011. Materials and
Methods: Seasonal autoregressive integrated moving average (SARIMA) model was used for statistical modeling.
Results: During January 2001 to December 2010, the reported DF/DHF cases showed a cyclical pattern with seasonal
variation. SARIMA (0,0,1) (0,1,1)12 model had the lowest normalized Bayesian information criteria (BIC) of 9.426 and
mean absolute percentage error (MAPE) of 263.361 and appeared to be the best model. The proportion of variance
explained by the model was 54.3%. Adequacy of the model was established through Ljung–Box test (Q statistic 4.910
and P-value 0.996), which showed no significant correlation between residuals at different lag times. The forecast for
the year 2011 showed a seasonal peak in the month of October with an estimated 546 cases. Conclusion: Application
of SARIMA model may be useful for forecast of cases and impending outbreaks of DF/DHF and other infectious
diseases, which exhibit seasonal pattern.

Key words: Dengue, SARIMA model, Time series analyses

Introduction In India, the disease reflects cyclic patterns, which over

An estimated 50 million dengue infections occur
annually and approximately 2.5 billion people live in
dengue endemic countries.1 Dengue fever (DF) inflicts
a significant health, economic, and social burden on the
populations of these endemic areas. The World Health
Organization (WHO) South-East Region and Western
Pacific Region bear nearly 75% of the global disease
burden due to dengue.2
the years have increased in frequency and geographical
extent. Over the past decade, the cases of dengue
have increased more than 20 times; from 650 cases
in 2000 to 15,535 in 2009.2 The case fatality rate is
significantly high compared with other infectious
diseases. Although, available data is largely derived
from hospitalized cases, which represent dengue
hemorrhagic fever (DHF) and dengue shock syndrome
(DSS), the burden due to uncomplicated DF is
nevertheless considerable.

*Corresponding Author: Dr. Vivek Lal,

Current dengue prevention strategies are weak as they
Assistant Professor, Institute of Health Management Research, 1, are reactive rather than anticipatory. As a result, they
Prabhu Dayal Marg, Sanganer Airport, Jaipur - 302 011, India. may often be implemented late, thereby reducing the
E-mail: doc.lal@gmail.com
opportunities for preventing transmission and controlling
the epidemic. The Asia Pacific Dengue Strategic Plan
Access this article online (2008–15) has been prepared to aid countries to reverse
Quick Response Code: the rising trend of dengue by enhancing their preparedness
Website: www.ijph.in
to detect, characterize, and contain outbreaks rapidly and
to stop the spread to new areas.3 Detailed information
DOI: 10.4103/0019-557X.106415
about when and where DF/DHF outbreaks occurred in
PMID: ***
the past can be a useful guide to the potential magnitude
and severity of future epidemics. Forecasting incidence

Indian Journal of Public Health, Volume 56, Issue 4, October-December, 2012

[Downloaded free from http://www.ijph.in on Tuesday, June 29, 2021, IP: 183.171.126.85]
282 Bhatnagar S, et al.: Forecasting Incidence of Dengue
of DF/DHF enables suitable allocation of resources for
improved public health interventions.
The outbreaks of DF/DHF can be predicted by
epidemiological modeling thus enabling the health
systems to be in readiness to manage outbreaks. Time
series analysis has been increasingly used in the field of
epidemiological research on infectious diseases, such as
influenza4 and malaria5-7 and dengue.8-13
The objective of the present study was to develop a
prediction model for DF/DHF using time series data over
the past decade in Rajasthan and to forecast the monthly
DF/DHF incidence for the year 2011.
Materials and Methods
Reported monthly DF/DHF cases from all the districts of
Rajasthan for the period January 2001 through December
2010 were obtained from the Directorate of Health and
Family Welfare, Government of Rajasthan.
Autoregressive integrated moving average (ARIMA)
modelshave been used for statistical modeling and
analyzing time series data containing ordinary or seasonal
trends to develop a predictive forecasting model.14 The
ARIMA approach was first popularized by Box and
Jenkins,15 and such models are often referred to as
Box–Jenkins models. The ARIMA procedure provides
a comprehensive set of tools for univariate time
series model identification, parameter estimation, and
forecasting, and it offers great flexibility in analysis,
which has contributed to its popularity in several areas
of research and practice. An ARIMA model may possibly
include autoregressive (p) terms, differencing (d) terms
and moving average (q) operations and is represented
by ARIMA (p, d, q).
The ARIMA models can be extended to handle seasonal
components of a data series. Seasonal ARIMA (SARIMA)
is an extension of the method to a series in which a pattern
repeats seasonally over time and is represented as SARIMA
(p, d, q) (P, D, Q)s. Analogous to the simple ARIMA
parameters, these are: Seasonal autoregressive (P), seasonal
differencing (D), and seasonal moving average parameters
(M); s defines the number of time periods until the pattern
repeats again (for a monthly data it is 12).
Statistical forecasting methods are based on the
assumption that the time series can be rendered
Indian Journal of Public Health, Volume 56, Issue 4, October-December, 2012
approximately stationary. A stationary time series is one
whose statistical properties such as mean and variance
are constant over time. Seasonality usually causes the
series to be nonstationary because the average values at
some particular times within the seasonal span may be
different than the average values at other times. 
SPSS version 19.0 was used to determine the best-fitting
model. The stationarity of the series was made by means
of seasonal and nonseasonal differencing. The order of
autoregression (AR) and moving average (MA) were
identified using autocorrelation function (ACF) and
partial autocorrelation function (PACF) of the differenced
series.
Several logical combinations of criteria to look for
better models were considered. From among several
models, the most suitable was selected based on three
measures, namely, normalized Bayesian information
criteria (BIC), mean absolute percentage error (MAPE),
and stationary R-squared. Whereas, lower values of BIC
and MAPE were preferred, a higher value of stationary
R-squared suggested a greater proportion of variance of
the dependent variable explained by the model.
Before using the model for forecasting, it was checked
for adequacy. A model is adequate if the residuals left
over after fitting the model are simply white noise. This
was done through examining the ACF and PACF of the
residuals. Further, Ljung–Box test was used to provide an
indication of whether the model was correctly specified.
A significant value less than 0.05 was considered to
acknowledge the presence of structure in the observed
series, which was not accounted for by the model;
therefore, we ignored the model if it had significant value.
After the best model was identified, forecast for monthly
values of the year 2011 were made.
Results
The time series plot of the reported DF/DHF cases
displayed seasonal fluctuations and therefore deemed
nonstationary. Large autocorrelations were recorded
for lags 1, 12, and 24 with values 0.6, 0.4, and 0.3,
respectively. The sharp decrease in autocorrelation values
after lag 1 indicated no evidence of a long-term trend;
consequently, there was no need to include a first-lag
difference term in the SARIMA model structure (d = 0).
In contrast, large autocorrelation values were registered
[Downloaded free from http://www.ijph.in on Tuesday, June 29, 2021, IP: 183.171.126.85]

Bhatnagar S, et al.: Forecasting Incidence of Dengue 283

at annual lags (and its multiples), which indicated the Discussion and Conclusion
need to include a 12-month difference term in the models
(S = 12, D = 1) [Figure 1]. The ACF and PACF plots ARIMA models are useful in modeling the temporal
of the differenced series provided further support for dependence structure of a time series as they explicitly
these conclusions [Figure 1]. Therefore, a SARIMA assume temporal dependence between observations.16
(p,0,q) (P,1,Q)12 was selected as the basic structure of Particularly for seasonal diseases, ARIMA models have
the candidate model. been shown to be adequate tools for use in epidemiological
surveillance.17 Our study provides an example of applying
Among the statistical models, SARIMA (0,0,1) (0,1,1)12 a SARIMA model to forecast incidence of DF/DHF.
was selected as the best model, with the lowest normalized Although these models have been utilized to forecast
BIC of 9.426 and a MAPE of 263.361 [Table 1]. The DF/DHF incidence in several countries,8-13 such analyses
model explained 54.3% of the variance of the series has not been undertaken in an Indian situation before.
(stationary R-squared). The model parameters were
significant (P-value <0.001) with MA in the model, Among all candidate models, SARIMA (0,0,1) (0,1,1)12
seasonal lag 1 of β = 0.756 (SE = 0.135). was the most suitable predictive model in our study, which
showed the highest stationary R-squared and the lowest
Ljung–Box test (Q statistic 4.910 and P-value 0.996) normalized BIC and MAPE values. In a recent study in
suggested that there were no significant autocorrelation Brazil, SARIMA (2,1,3) (1,1,1)12 model offered best fit for
between residuals at different lag times and the residuals the dengue incidence data.8 However, in a previous study
were white noise. This was further corroborated by by Luz et al.,10 for monitoring dengue incidence in Rio
plotting the ACF and PACF of the residuals [Figure 2]. de Janeiro, Brazil, no seasonal differencing was reported
and SARIMA (2,0,0) (1,0,0)12 model was deemed best fit.
Moreover, the same model was also returned by the Choudhury et al.,9 reported SARIMA (1,0,0) (1,1,1)12 as
expert modeler. Having tested its validity, the prediction the most suitable model for forecasting dengue incidence
model was used to forecast incidence of DF/DHF cases in
the upcoming season in 2011. Figure 3 shows the month- Table 1: Normalized bayesian information criteria (BIC), mean
wise trends of DF/DHF over the past 10 years and for absolute percentage error (MAPE) and stationary R-squared
2011. The cases showed a similar seasonality, with a peak values of SARIMA models
in the month of October similar to previous years with an Models MAPE Normalized Stationary
BIC R-squared
estimated 546 cases (95% CI 311–781). The momentum
SARIMA (0,0,1) (0,1,1)12 263.361 9.426 0.543
in dengue would begin in August 2011, peak in October, SARIMA (0,0,1) (0,1,0)12 299.842 9.809 0.330
and then wane off toward December [Figure 3]. SARIMA (1,0,0) (1,1,1)12 245.526 9.590 0.516
SARIMA (1,0,1) (1,1,0)12 198.894 9.727 0.445
SARIMA (1,0,0) (1,1,0)12 160.090 9.761 0.394

a b c d
Figure 1: Autocorrelation function (ACF) and Partial autocorrelation function (PACF)
of data and transformed series Footnote: (a) shows time series data without any
difference; (b) shows transformed series with seasonal difference (1, period 12);
(c) shows transformed series with nonseasonal difference (1); (d) shows transformed Figure 2: Autocorrelation function (ACF) and partial autocorrelation function (PACF)
series with both seasonal (1, period 12) and nonseasonal difference (1) of residuals

Indian Journal of Public Health, Volume 56, Issue 4, October-December, 2012

[Downloaded free from http://www.ijph.in on Tuesday, June 29, 2021, IP: 183.171.126.85]
284 Bhatnagar S, et al.: Forecasting Incidence of Dengue
Figure 3: Observed and forecast incident cases of dengue fever/dengue
hemorrhagic fever (DF/DHF), rajasthan, 2001–2011
in Dhaka, Bangladesh. Separate studies undertaken to
forecast DF/DHF incidence in northern, southern, and
north-eastern Thailand have yielded SARIMA (2,0,1)
(0,2,0)12, ARIMA (1,0,1), and SARIMA (2,1,0) (0,1,1)12
models as most suitable.11-13
Our findings corroborated that DF/DHF cases followed
a seasonal pattern during the past decade 2001–10. The
model revealed that there would be again a seasonal spurt
in these cases with a peak in October 2011. This is also
consistent with the data from previous years with regard
to the timing of the peak.
However, the predictions may not be credible for
forecasting the number of dengue cases in epidemic
years, as it could be a consequence of a lack of immunity
in a population exposed for the first time to a given
dengue viral serotype.8
More importantly, meteorological factors such as
temperature, humidity, and rainfall have considerable
impact on dengue transmission, and climate variables
introduced into models can increase their predictive power.18
The forecasting models are based on reported cases,
which represent the severe cases of DHF/DSS admitted
to the hospitals and who have been laboratory-confirmed.
ARIMA modeling is a useful tool for interpreting
surveillance data and forecast of the cases to help guide
timely prevention and control measures. In addition, the
usefulness of forecasting expected numbers of infectious
disease may lie in providing decision-makers a clearer
idea of the variability to be expected among future
observations.19
Indian Journal of Public Health, Volume 56, Issue 4, October-December, 2012
Further research is recommended to evaluate the
effectiveness of integrating the forecasting model into
the existing disease control program in terms of its impact
in reducing the disease occurrence.
References
1. World Health Organization. Dengue and dengue
haemorrhagic fever. Factsheet No 117, revised May 2008.
Geneva:World Health Organization; 2008.Available from:
http://www.who.int/mediacentre/factsheets/fs117/en/. [Last
accessed on 2011 Apr 8].
2. World Health Organization. Situation update of dengue
in the SEA Region, Geneva: World Health Organization;
2010. Available from:http://www.searo.who.int/LinkFiles/
Dengue_Dengue_update_SEA_2010.pdf. [Last accessed
on 2011 Apr 8].
3. World Health Organization and the Special Programme
for Research and Training in Tropical Diseases (TDR).
Dengue: Guidelines for diagnosis, treatment, prevention
and control- New edition. Geneva: World Health
Organization; 2009. Available from: http://whqlibdoc.who.int/
publications/2009/9789241547871_eng.pdf. [Last accessed
on 2011 Apr 8].
4. Soebiyanto RP, Adimi F, Kiang RK. Modeling and predicting
seasonal influenza transmission in warm regions using
climatological parameters. PLoS One 2010;5:e9450.
5. Wangdi K, Singhasivanon P, Silawan T, Lawpoolsri S,
White NJ, Kaewkungwal J. Development of temporal
modelling for forecasting and prediction of malaria infections
using time-series and ARIMAX analyses: A case study in
endemic districts of Bhutan. Malar J 2010;9:251.
6. Loha E, Lindtjørn B. Model variations in predicting incidence
of Plasmodium falciparum malaria using 1998-2007
morbidity and meteorological data from south Ethiopia.
Malar J 2010;9:166.
7. Tian L, Bi Y, Ho SC, Liu W, Liang S, Goggins WB, et al.
One-year delayed effect of fog on malaria transmission: A
time-series analysis in the rain forest area of Mengla County,
south-west China. Malar J 2008;7:110.
8. Martinez EZ, Silva EA. Predicting the number of cases
of dengue infection in RibeirãoPreto, São Paulo State,
Brazil, using a SARIMA model. Cad Saude Publica 2011;
27:1809-18.
9. Choudhury MA, Banu S, Islam MA. Forecasting dengue
incidence in Dhaka, Bangladesh: A time series analysis.
Dengue Bull 2008;32:29-37.
10. Luz PM, Mendes BV, Codeço CT, Struchiner CJ, Galvani AP.
Time series analysis of dengue incidence in Rio de Janeiro,
Brazil. Am J Trop Med Hyg 2008;79:933-9.
11. Silawan T, Singhasivanon P, Kaewkungwal J, Nimmanitya S,
Suwonkerd W. Temporal patterns and forecast of dengue
infection in Northeastern Thailand. Southeast Asian J Trop
Med Public Health 2008;39:90-8.
12. Wongkoon S, Pollar M, Jaroensutasinee M, Jaroensutasinee
K. Predicting DHF incidence in Northern Thailand using time
[Downloaded free from http://www.ijph.in on Tuesday, June 29, 2021, IP: 183.171.126.85]

Bhatnagar S, et al.: Forecasting Incidence of Dengue 285

series analysis technique. International Journal of Biological
and Life Sciences 2008;4:117-121.
13. Promprou S, Jaroensutasinee M, Jaroensutasinee K.
Forecasting dengue haemorrhagic fever cases in southern
Thailand using ARIMA models. Dengue Bull 2006;30:99-106.
14. Peter JD. Time Series: A biostatistical introduction. Oxford
Statistical Science Series-5. 1990.
15. Box GE, Jenkins GM. Time series analysis: Forecasting and
control.San Francisco: Holden-Day; 1976.
16. Helfenstein U. The use of transfer function models,
intervention analysis and related time series methods in
epidemiology. Int J Epidemiol 1991;20:808-15.
17. Nobre FF, Monteiro AB, Telles PR, Williamson GD. Dynamic
linear model and SARIMA: A comparison of their forecasting
performance in epidemiology. Stat Med 2001;20:3051-69.
18. World Health Organization. Using climate to predict
infectious disease outbreaks: A review. Geneva: World Health
Organization; 2004.
19. Allard R. Use of time-series analysis in infectious disease
surveillance. Bull World Health Organ 1998;76:327-33.
Cite this article as: Bhatnagar S, Lal V, Gupta SD, Gupta OP. Forecasting
incidence of dengue in Rajasthan, using time series analyses. Indian J Public
Health 2012;56:281-5.
Source of Support: Nil. Conflict of Interest: No.

Author Help: Online Submission of the Manuscripts

Articles can be submitted online from http://www.journalonweb.com. For online submission articles should be prepared in two files (first page
file and article file). Images should be submitted separately.
1) First Page File:
Prepare the title page, covering letter, acknowledgement, etc., using a word processor program. All information which can reveal your
identity should be here. Use text/rtf/doc/pdf files. Do not zip the files.
2) Article file:
The main text of the article, beginning from Abstract till References (including tables) should be in this file. Do not include any information
(such as acknowledgement, your names in page headers, etc.) in this file. Use text/rtf/doc/pdf files. Do not zip the files. Limit the file size to
1 MB. Do not incorporate images in the file. If file size is large, graphs can be submitted as images separately without incorporating them
in the article file to reduce the size of the file.
3) Images:
Submit good quality color images. Each image should be less than 4096 kb (4 MB) in size. Size of the image can be reduced by decreasing
the actual height and width of the images (keep up to about 6 inches and up to about 1200 pixels) or by reducing the quality of image. JPEG
is the most suitable file format. The image quality should be good enough to judge the scientific value of the image. Always retain a good
quality, high resolution image for print purpose. This high resolution image should be sent to the editorial office at the time of sending a
revised article.
4) Legends:
Legends for the figures/images should be included at the end of the article file.

Indian Journal of Public Health, Volume 56, Issue 4, October-December, 2012