Evaluating the impact of cash transfers on tuberculosis (ExaCT TB): a stepped wedge cluster randomised controlled trial

Abstract

Background Mitigating financial barriers to tuberculosis (TB) diagnosis and treatment is a core priority of the global TB agenda. We evaluated the impact of a cash transfer intervention on completion of TB testing and treatment initiation in Uganda. Methods We conducted a pragmatic complete stepped wedge randomised trial of a one-time unconditional cash transfer at 10 health centres between September 2019 and March 2020. People referred for sputum-based TB testing were enrolled to receive UGX 20 000 (∼USD 5.39) upon sputum submission. The primary outcome was the number initiating treatment for micro-bacteriologically confirmed TB within 2 weeks of initial evaluation. The primary analysis included cluster-level intent-to-treat and per-protocol analyses using negative binomial regression. Results 4288 people were eligible. The number diagnosed with TB initiating treatment was higher in the intervention period versus the pre-intervention period (adjusted rate ratio (aRR)=1.34) with a 95% CI of 0.62–2.91 (p=0.46), indicating a wide range of plausible true intervention effects. More were referred for TB testing (aRR=2.60, 95% CI 1.86–3.62; p<0.001) and completed TB testing (aRR=3.22, 95% CI 1.37–7.60; p=0.007) per National Guidelines. Results were similar but attenuated in per-protocol analyses. Surveys revealed that while the cash transfer supported testing completion, it was insufficient to address long-term underlying social/economic barriers. Interpretation While it is uncertain whether a single unconditional cash transfer increased the number of people diagnosed and treated for TB, it did support higher completion of diagnostic evaluation in a programmatic setting. A one-time cash transfer may offset some but not all of the social/economic barriers to improving TB diagnosis outcomes.