Forecasting Tuberculosis (Mortality) in Thailand Using Multivariate Linear Regression

Abstract

Tuberculosis (TB) is a major cause of morbidity, mortality, and disability worldwide. It is still a public health problem in Thailand. The objective of the study was to model and forecast TB mortality in Thailand using death certificate reports. A retrospective analysis of the TB mortality rate was conducted. Data were obtained from the national vital registration database for the 10-year period from 2000 to 2009, provided by the Ministry of the Interior and coded as cause-of-death using ICD-10 by the Ministry of Public Health. Multivariate linear regression was used for modeling and forecasting age-specific TB mortality rates in Thailand. Gender differences existed in TB mortality in Thailand with higher deaths occurring in males. TB mortality increased with increasing age for each sex and was also higher in the Central and Northern provinces. The trends of TB mortality decreased in most age groups and remained stable in others during ten-year period (2000 to 2009). The model forecast that the TB mortality will not increase over the 6-year period, and will actually decrease in most of age group and region for both sexes. The multivariate linear regression model can be used as a simpler method for forecasting TB morality. These findings provide information on forecasting for health authorities to help establish effective prevention programs in specific areas and groups where the TB mortality is relatively high.