Setting: Community-based population of homeless adults living in San Francisco, California.
Objective: To compare the effect of cash and non-cash incentives on 1) adherence to treatment for latent tuberculosis infection, and 2) length of time needed to look for participants who missed their dose of medications.
Design: Prospective, randomized clinical trial comparing a 5 dollar cash or a 5 dollar non-cash incentive. All participants received directly observed preventive therapy and standardized follow-up per a predetermined protocol. Completion rates and amount of time needed to follow up participants was measured.
Results: Of the 119 participants, 102 (86%) completed therapy. There was no difference between the cash and non-cash arms. Completion was significantly higher among males (OR 5.65, 95%CI 1.36-23.40, P = 0.02) and persons in stable housing at study entry (OR 4.86, 95%CI 1.32-17.94, P = 0.02). No substance use or mental health measures were associated with completion. Participants in the cash arm needed significantly less follow-up to complete therapy compared to the non-cash arm (P = 0.03). In multivariate analysis, non-cash incentive, use of crack cocaine, and no prior preventive therapy were associated with more follow-up time.
Conclusion: Simple, low cost incentives can be used to improve adherence to TB preventive therapy in indigent adults.