After more than 10 years of absence in Korea, Plasmodium vivax malaria re-emerged in 1993 near the Korean demilitarized zone. Since then it has been a major public health concern in the Republic of Korea. Although the mosquitoes from North Korea were suspected to be the main cause of the re-emerging infection, the subject is a matter of controversy, because the prevalence of malaria in North Korea was reported only in 1997–1998 [1]. While it was initially suggested that the incidence rate will decrease soon [2], it has not decreased rapidly and there was a fluctuation of incidence rate even after 10 years owing to the fluctuation of the malarial incidence in North Korea [3].
In this issue Noh and co-workers have estimated the annual pattern of malarial incidence by age, gender, and year in Korea using the data from the last 12 years (2001–2011) by adopting the hierarchical generalized linear model (HGLM) [4]. Spatial and temporal correlations have been estimated and the best model has been selected out of nine models. Results were presented as diseases map according to age and gender. In the model, spatial and temporal correlations were also considered for the best estimation. Results of estimated incidence were represented as diseases map [5]. To draw a diseasesmap by estimating the incidence rate fromsmall regions, empirical Bayesian and hierarchical Bayesian estimation methods have been used. Bayesian approach requires assumption of previous parameters, but HGLM approach presents hierarchical likelihood so that there are no sensitivity problems of parameter estimation that may occur in the case of wrong assumption of previous parameters [5]. The advantage of usingHGLMinstead of Bayesian approaches is to avoid complicated calculation process such as Gibbs sampling procedure for parameter estimation. The authors have developed R package for estimation [6].
The analysis well addresses the following question: Does the malarial incidence of one region in a year correlates with that of adjacent regions and the previous year? The study also clearly specifies the characteristics of malarial incidence by age and gender. Finally, the spatial and temporal correlations of malarial incidence are presented using the HGLM model.
The results of the study suggest that it is possible to predict malarial transmission according to time and geographic location, and it has also provided an analytical basis to define strategy and identify countermeasures for preventing malaria in Korea.
References
- 1.Gho W-G. Reemergence of malaria in Korea. J Korean Med Assoc. 2007 Nov;50(11):959–66. [Google Scholar]
- 2.Lee JS, Lee WJ, Cho SH, et al. Outbreak of vivax malaria in areas adjacent to the demilitarized zone, South Korea, 1998. Am J Trop Med Hyg. 2002 Jan;66(1):13–7. doi: 10.4269/ajtmh.2002.66.13. [DOI] [PubMed] [Google Scholar]
- 3.Park JW, Jun G, Yeom JS. Plasmodium vivax malaria: status in the Republic of Korea following reemergence. Korean J Parasitol. 2009 Oct;47(Suppl.):S39–50. doi: 10.3347/kjp.2009.47.S.S39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Noh M, Lee Y, Oh S, et al. Spatial and temporal distribution of Plasmodium vivax malaria in Korea estimated with hierarchical generalized linear model. Osong Public Health Res Perspect. 2012 Dec;3(4):192–8. doi: 10.1016/j.phrp.2012.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lee Y, Jang M, Lee W. Prediction interval for diseasemapping using hierarchical likelihood. Comput Stat. 2011 Mar;26(1):159–79. [Google Scholar]
- 6.Noh M, Lee Y. dhglm: Double Hierarchical Generalized Linear Models. R package version 1.0 (Internet) The R Foundation for Statistical Computing; Vienna: 2012. [(Accessed 01.11.2012)]. [Google Scholar]