Tampi Successfully Defends Dissertation – Center for Health Decision Science

Radhika Tampi, doctoral student in Health Policy concentrating in Decision Science, successfully defended her dissertation, “Simulation of Policy Levers Across the Tuberculosis Care Cascade.”

Tuberculosis (TB) causes substantial mortality and morbidity, especially among children and adults with comorbid diabetes and multidrug-resistant/rifampicin-resistant TB (MDR/RR-TB). Diagnosis in children is challenging because symptoms often overlap with those of other childhood illnesses, and young children often have difficulty producing sputum samples for testing. Tampi developed a microsimulation model of childhood TB which she used in her first two dissertation chapters to simulate how failures across the TB care cascade affect outcomes in children.

In Chapter 1, she quantified the mortality and morbidity consequences of delays in diagnosis and treatment among children younger than ten years. She found that delayed diagnosis increased mortality and morbidity across all age groups, with the greatest harms concentrated among children under two and children living with HIV. Improved diagnostic sensitivity led to meaningful gains in survival, particularly among the youngest children, but these gains were substantially reduced when treatment initiation was delayed. Similarly, immediate treatment only partially mitigated the harms of imperfect diagnostic sensitivity.

In Chapter 2, Tampi used the same model to evaluate trade-offs between health system access and diagnostic accuracy. She modeled access to care, access to chest x-rays, molecular rapid diagnostics, and the sensitivity and specificity of treatment decision algorithms. She found that access to care was the most important of these factors. Access to chest x-ray reduced the likelihood that children without TB would be inappropriately treated for TB.

In Chapter 3, Tampi used causal inference to explore whether the risk of unfavorable treatment outcomes in individuals with MDR/RR-TB and diabetes differed by baseline glycemic control in adults. She found that, relative to no diabetes, uncontrolled diabetes at baseline was associated with a 43% higher risk of unfavorable outcomes, whereas controlled diabetes was not. When she simulated hypothetical interventions in which all individuals remained engaged in treatment, the risk of unfavorable treatment outcomes was attenuated among those with uncontrolled diabetes at baseline, suggesting that disengagement from treatment may explain part, but not all, of the excess risk.

Learn more: Read the publication by Tampi and colleagues, 1,7-Malaria Reactive Community-Based Testing and Response (1,7-mRCTR) Approach in Tanzania: A Cost-Effectiveness Analysis