Developing a Predictive Information System for Determining the Prognosis of HIV and Tuberculosis Co-Infection in Incarcerated Individuals

Abstract

The research describes the development of the forecasting information system that can predict the outcome of a disease in inmates with HIVassociated tuberculosis. The aim of the research is to develop an additional high-precision diagnostic criterion giving the possibility of the timely correction of diagnostic, treatment and organizational measures for patients with HIV and tuberculosis co-infection that helps to improve the quality of medical care. The research material is based on the data from clinical cases of patients with HIV and tuberculosis co-infection who were hospitalized in a tuberculosis hospital providing specialized medical care to inmates from 2012 to 2018. (367 people). The study methodology was developed in several stages with the use of the methods of system analysis and mathematical modeling (logical-statistical method of optimally reliable partitions, the method of analysis of hierarchies, artificial neural network, methods of statistical grouping). The result of a complex multi-stage research makes possible the development of the prognostic index forecasting outcome of HIV-associated tuberculosis in prisoners. For the automated calculation of the developed index a software package was created. Interpretation of the received data allows timely correcting of diagnostic and treatment tactics in order to improve the quality of medical care and reduce the hospital mortality rate. The developed information system does not require complex and expensive diagnostic measures, it is easy to use and can be suggested to use as a screening method.