What is it about?
This study proposes a complete methodology of predicting whether a tuberculosis patient under treatment is likely to complete the treatment course or leave the treatment unfinished.
Featured Image
Why is it important?
Tuberculosis is the deadliest contagious disease and its treatment is long and complicated. As a result, some patients leave their treatment unfinished and can cause a higher risk of TB penetration in healthy population. If we can predict beforehand whether a patient will complete the treatment, we can pursue more aggressively and ensure that the patient does not default (by using incentives, counseling, etc.)
Perspectives
I'm hopeful that this paper will help public health teams running TB programs to devise a cost-effective way to reduce the rate of patients who are likely to lose treatment follow-up. Since the experiments are open source, anyone with primary knowledge of data tools and R language can reuse my model -- which offers high accuracy -- and quickly turn into a workable piece of software for real case studies.
Owais Hussain
Pakistan Air Force Karachi Institute of Economics and Technology
Read the Original
This page is a summary of: Predicting treatment outcome of drug-susceptible tuberculosis patients using machine-learning models, Informatics for Health and Social Care, February 2018, Taylor & Francis,
DOI: 10.1080/17538157.2018.1433676.
You can read the full text:
Contributors
The following have contributed to this page
