UNDERSTANDING TUBERCULOSIS MORTALITY FLUCTUATION IN THE PHILIPPINES BY ARTIFICIAL INTELLIGENCE

Abstract

The Philippines is one of the countries with high tuberculosis (TB) mortality across the globe. This study investigates the trends of the TB mortality and morbidity incidence from 1998 to 2010 based on the surveillance data of the Department of Health. From the dataset, symbolic regression was used to forecast yearly morbidity cases of TB. Using the same analysis, a significant non-linear correlation between TB morbidity and TB mortality was revealed. Analyzing the equation, the trend component represented by the first two terms shows that the morbidity cases increases like a quadratic function. The correction factor, however, tells us that there are fluctuations around the trend curve because of the presence of the cosine terms. Following the equation generated through symbolic regression, in the next 15 years (2011–2025), there is an increasing number of TB morbidity cases considering the present TB Control Programs of the country. Mortality cases follow the same increasing trend but with evident fluctuations over the years. Therefore, without the presence of government and non-government-initiated TB prevention and control programs, there would be an increase in both TB morbidity and mortality cases in the Philippines. Conversely, the existence of such interventions will slow down the rise of TB mortality and morbidity. Full implementation of anti-TB programs is to be mandated to and must be adhered by all public and private healthcare institutions to help realize the vision of a TB-free Philippines.