Back

A Multi-Country Evaluation of Patient Preferences for Tuberculosis Diagnostics: A Discrete Choice Experiment to Inform WHO Target Product Profiles

Shah, K.; Nalugwa, T.; Marcelo, D. B.; Nakawunde, R.; Trinh, T.; Shankar, D.; Van Luong, D.; Theron, G.; Worodria, W.; del Mar Castro, M.; Nahid, P.; Denkinger, C.; Cattamanchi, A.; Yu, C.; Kerkhoff, A.

2024-06-22 public and global health
10.1101/2024.06.19.24309124 medRxiv
Show abstract

BackgroundRecognizing new and improved diagnostics as a key step to reducing the global tuberculosis (TB) burden, the World Health Organization (WHO) publishes target product profiles (TPPs) to guide development of novel TB diagnostics; however, TPPs have not included the preferences of people undergoing TB testing. Understanding their preferences is crucial for optimizing the acceptability and uptake of TB diagnostic services. Design/methodsWe conducted a discrete choice experiment (DCE) among adults with presumptive or microbiologically confirmed TB in 5 countries (Philippines, Vietnam, South Africa, Uganda, and India). The DCE evaluated preferences for 5 TB diagnostic test attributes (sample type, accuracy, cost, location, time to result) with 3-4 levels per attribute. We estimated mean preference weights for attribute levels using Hierarchical Bayesian models and conducted willingness-to-trade analyses for preferred test features. ResultsAmong 1,033 participants (median age 43 years, Interquartile Range: 30-55), 52% were female, 10% had HIV, 14% had diabetes, 36% had current or prior TB), the most strongly valued test features were free cost, high diagnostic accuracy, and rapid (within 15 minutes) result availability. Among less valued attributes, facility-based testing (location) and sputum (sample type) were most preferred. The relative importance of attributes and their levels differed by country. In willingness-to-trade simulations, participants in each country were willing to trade 10-20% lower accuracy for the availability of rapid tests but were unwilling to trade accuracy for alternative sample types (e.g., tongue-swab, urine, blood) or testing locations (e.g., home- or community-based); participants in India and Vietnam were willing to pay $3.20 and $4.00 USD, respectively, for tests with rapid results. ConclusionsPersons accessing TB services in five high-burden countries universally value free cost, high diagnostic accuracy, and rapid result availability. However, many were willing to accept lower test accuracy and some were willing to incur small out-of-pocket costs to access rapid tests. To align with preferences of TB-affected communities, future TPP updates and new test development should prioritize rapid results while maintaining high accuracy and minimizing costs.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.