Grant: CIHR Fellowship

Tuberculosis (TB) is a bacteria that infected more than 10 million people in 2017. Some TB patients have bacteria that are not killed by our standard anti-TB drugs; these bacteria are referred to as drug-resistant. If patients receive the wrong drugs, they are more likely to die. The number of patients who die during treatment is an important indicator of how well a TB treatment program is working so it is critical that we report this number accurately. However, sometimes up to 20% of patients will stop coming to see the doctor during treatment because they moved or did not want to keep taking the drugs because of side effects etc. I will investigate how to accurately report the number of TB patients who died by correcting for patients who stopped coming to the doctor. I will compare two different correction methods using simulated data and then apply the best method to drug-resistant TB patient data from the Philippines. This project will help researchers and TB programs accurately estimate the number of TB patients who died so they can better measure TB quality of care.