Sains Ma1aysiana 26(3&4): 25-30 (1997)                                                                                    Pengajian Kuantitatif/

                                                                                                                                                                Quantitative Studies

On an Application of Sequential Bayesian Techniques for

Estimating the Fatality Rate of Dengue in Malaysia

Kamarulzaman Bin Ibrahim & Abdul Aziz Bin Jermain

Department of Statistics

Faculty of Mathematical Sciences

Universiti Kebangsaan Malaysia

43600 UKM Bangi Selangor D.E. MALAYSIA

ABSTRACT

Dengue is one of the main factors of mortality of inhabitants in the region of South East Asia. Malaysia is one of the countries which is facing a high incidence of dengue, particularly in the 70's and early 80's. The Ministry of Health has taken various measures in order to reduce the dengue epidemic. These include educating people about dengue and conducting research such as investigation of factors that influence the epidemic of dengue. In this study, a sequential Bayesian approach is applied to data of the proportion of death due to dengue over the period from 1982 to 1992. In the sequential Bayesian approach, the data for the year 1982 becomes the prior information for the 1983 data and so on. The data for the different periods are combined in a chronological manner until the final posterior distribution of the proportion of death due to dengue is obtained. It is found that the overall proportion is 0.59% and its standard deviation is 0.00002%.

ABSTRAK

Denggi adalah satu daripada faktor utama kematian bagi penduduk di rantau Asia Tenggara. Malaysia pula merupakan satu daripada negara yang sedang mengalami kadar insiden denggi yang tinggi, khususnya dalam tahun 70-an dan pada awal 80-an. Kementerian Kesihatan telah mengambil pelbagai langkah untuk mengurangkan wabak denggi. Ini termasuk memberikan pendidikan tentang denggi dan membuat kajian tentang faktor-faktor yang mempengaruhinya. Dalam kajian ini kaedah Bayesan jujukan digunakan terhadap data perkadaran yang mati akibat denggi dalam tempoh 1982 hingga 1992. Dalam kaedah ini, data tahun 1982 digunakan sebagai maklumat prior untuk data tahun 1983 dan seterusnya. Data dari tahun yang berlainan digabungkan secara kronologi sehingga diperoleh taburan posterior yang terakhir bagi perkadaran yang mati akibat denggi. Didapati bahawa perkadaran keseluruhan ialah 0.59% dan sisihan piawainya 0.00002%.

RUJUKAN/REFERENCES

Azizah Aris. 1996. Pencaman faktor-faktor risiko yang mempengaruhi jangkitan denggi: Kajian kes di Daerah Jinjang-Kepong, Selangor. MSc Thesis. Universiti Kebangsaan Malaysia

Kamarulzaman Ibrahim & Metcalfe, A. V. 1993. Bayesian overview for evaluation of mini-roundabouts as a road safety measure. Journal of the Royal Statistical Society (series) 42(5): 525-540.

Lee, E.T. & Cheong, W. H. 1987. A preliminary aedes aegypti larva survey in the suburbs of Kuala Lumpur city. Tropical Medicine 4: 111-118.

Press, S. J. 1989. Bayesian Statistics: Principles, Models. and Applications. New York: Wiley.

Ministry of Health. 1982-1992. Rancangan Kawalan Penyakit Bawaan Vektor. Kuala Lumpur: Head office RKPBV.