Summary and Implications for Future Study

Abstract

HIV is a scourge; more than 2.5 million of the infected populations die every year. For control of non-curable diseases, there is a need for implementation of new innovative public health programs. Till date, HIV infection continues to have a significant impact, especially on children. In the year 2012, approximately 2.10 million children were living with HIV worldwide. In Indian context national adult HIV prevalence was 0.27% in 2011. Of these, women constituted 39% of all people living with HIV (PLHIVs), while children less than 15 years of age constituted 7% of all infections. As of March 2015, 0.10 million HIV-positive children had been registered under the antiretroviral therapy (ART) program, and 38,579 are receiving free ART. There has been a significant scale-up of HIV counseling and testing, prevention of parent-to-child transmission (PPTCT), and ART services across the country over the last 5 years. It is a great economic burden on National GDP and public health. AIDS also poses a threat to economic growth in many countries already in distress. The World Bank estimates an annual slowing in growth of income per capita by an average of 0.6% per country in ten worst affected countries. The negative impact of AIDS on households, productive enterprises, and countries seems to be partly from the high cost of treatment, which diverts resources from productive investments. Modelling is an essential tool for understanding disease progression and its relationships with social, biological, and environmental mechanisms that influence the spread of HIV/AIDS. Due to paucity of research work on modeling, the policymakers are unable to implement the new innovative programs at national level. The entitled book HIV Transmission: Statistical Modelling is an outcome of a series of analyses carried out by developing and applying mathematical and statistical models on a body of real-life data compiled in Karnataka State AIDS Control Societies to know the rate of HIV transmission from mother to child in Bangalore city. Several researchers developed mathematical and statistical models to analyze and test the HIV disease progression rate of mother-to-child transmission (MTCT). However, there are practical issues on which mathematical and statistical models are developed and applied and should be addressed to arrive at reasonable analytical reasoning on HIV transmission from mother to child. In that context, examination of practical issues related to building of mathematical modeling of HIV MTCT large data sets for reduction of high-dimensional HIV data sets is necessary. The methodology attempted in the present thesis is at three levels. First is to build the model on real-life HIV mother-to-child transmission large data sets to know the disease epidemic and endemic equilibrium with varied parameters. The validity of the developed model is examined through the estimation of proportion of babies who got HIV infected through defined risk factors. Second, two common practical issues, viz., the statistical criteria in terms of reduction of large high-dimensional HIV data sets on the basis of theoretical model and intervention with numerical simulation for the estimation of defined risk parameters of HIV MTCT, are considered. Third, sometimes it may not be feasible to build the desired and efficient models for the estimation of HIV mother-to-child transmission. In such a situation, if the researcher or statistician estimates the progression rate effectively by the way of building or developing mathematical models through different analytical procedures, it will be useful for nurturing HIV epidemic at population level. There are several data sets generated from HIV clients, viz., drug reaction, HIV TB co-infection, HIV-associated neurological disorders, ophthalmological complications, etc., for describing clinical complications. HIV itself is a life-threatening deadly disease, and many real-life data sets are generated from the lifelong treatment and their follow-up records. Over the past era, statisticians, researchers, and policymakers have put a great amount of efforts in developing and comparing suitable mathematical and statistical models on HIV data sets. These fitted models must be general and broad enough to allow for building theoretical and predictive models. The consistency of findings of the fitted models suggests that the quality of estimation is good, taking into consideration of the practical, analytical, and statistical issues and the efficient quality of estimation of parameters with varied socioeconomic, behavioral, and risk factors.