Analysis of clinical malaria disease patterns and trends in Vietnam 2009-2015

Abstract

Background: Viet Nam has made tremendous progress towards reducing mortality and morbidity associated with malaria in recent years. Despite the success in malaria control, there has been a recent increase in cases in some provinces. In order to understand the changing malaria dynamics in Viet Nam and measure progress towards elimination, the aim of this study was to describe and quantify spatial and temporal trends of malaria by species at district level across the country.

Methods: Malaria case reports at the Viet Nam National Institute of Malariology, Parasitology, and Entomology were reviewed for the period of January 2009 to December 2015. The population of each district was obtained from the Population and Housing Census-2009. A multivariate (insecticide-treated mosquito nets [ITN], indoor residual spraying [IRS], maximum temperature), zero-inflated, Poisson regression model was developed with spatial and spatiotemporal random effects modelled using a conditional autoregressive prior structure, and with posterior parameters estimated using Bayesian Markov chain Monte Carlo simulation with Gibbs sampling. Covariates included in the models were coverage of intervention (ITN and IRS) and maximum temperature.

Results: There was a total of 57,713 Plasmodium falciparum and 32,386 Plasmodium vivax cases during the study period. The ratio of P. falciparum to P. vivax decreased from 4.3 (81.0% P. falciparum; 11,121 cases) in 2009 to 0.8 (45.0% P. falciparum; 3325 cases) in 2015. Coverage of ITN was associated with decreased P. falciparum incidence, with a 1.1% (95% credible interval [CrI] 0.009%, 1.2%) decrease in incidence for 1% increase in the ITN coverage, but this was not the case for P. vivax, nor was it the case for IRS coverage. Maximum temperature was associated with increased incidence of both species, with a 4% (95% CrI 3.5%, 4.3%) and 1.6% (95% CrI 0.9%, 2.0%) increase in P. falciparum and P. vivax incidence for a temperature increase of 1 °C, respectively. Temporal trends of P. falciparum and P. vivax incidence were significantly higher than the national average in Central and Central-Southern districts.

Conclusion: Interventions (ITN distribution) and environmental factors (increased temperature) were associated with incidence of P. falciparum and P. vivax during the study period. The factors reviewed were not exhaustive, however the data suggest distribution of resources can be targeted to areas and times of increased malaria transmission. Additionally, changing distribution of the two predominant malaria species in Viet Nam will require different programmatic approaches for control and elimination.

Keywords: Bayesian; Malaria elimination; Modelling; Spatial; Viet Nam.

Fig. 1

Map of Vietnam with administrative divisions and neighbouring countries

Fig. 2

Malaria incidence with proportion of population covered by two different interventions, 2009–2015. ITN: insecticide-treated mosquito nets; IRS: insecticide residual spraying; PF: Plasmodium falciparum; PV: Plasmodium vivax

Fig. 3

Maps showing the proportion of population protected by insecticide treated nets in districts of Viet Nam from 2009 to 2013

Fig. 4

Maps showing the proportion of population protected by indoor residual spraying in districts of Viet Nam from 2009 to 2013

Fig. 5

Raw standardised morbidity ratios of a Plasmodium falciparum and b Plasmodium vivax by districts in Viet Nam from 2009 to 2014

Fig. 6

Decomposed monthly Plasmodium falciparum incidence per 10,000 population during the study period, 2009–2015. The top layer shows the original time series. The other layers show the decomposed components, denoting the seasonal component, long-term trend component and remainder component, respectively

Fig. 7

Decomposed monthly Plasmodium vivax incidence per 10,000 population during the study period, 2005–2015. The top layer shows the original time series. The other layers show the decomposed components, denoting the seasonal component, long-term trend component and remainder component, respectively

Fig. 8

Spatial distribution of the posterior means of structured and unstructured random effects for Plasmodium falciparum in Model III

Fig. 9

Spatial distribution of the posterior means of structured and unstructured random effects for Plasmodium vivax in Model III

Fig. 10

Trend analysis of a Plasmodium falciparum and b Plasmodium vivax during the study period 2009–2014