Tuberculosis is a major public health problem in Nigeria, and the country is 4th among the 22 countries with the highest TB burden worldwide. The World Health Organization estimates incident and prevalence rates for all forms of TB of 118 and 171 cases per 100,000 of population, respectively (WHO Tuberculosis Report 2012).
For the past several months the Company has been working with a team from Abuja, Nigeria and the Liverpool School of Tropical Medicine (Liverpool, United Kingdom). Study protocols and ethics committee approvals have been accepted for the evaluation of TBDx™ over a six (6) month period of time, in conjunction with a larger TB Reach study. The National and the State TB Programme coordinators for Nigeria are co-applicants in this project.
- Evaluate the performance of the TBDx™ computer vision technology against the diagnostic performance of routine smear microscopy, using culture as the ’gold standard’.
- Measure the economic impact for each positive case detected, using TBDx™ as the front-end screening solution with all positive cases provided to Xpert for confirmation and drug susceptibility testing.
Assessment of the TBDx™
This assessment will use a cross sectional design. The evaluation of performance will be followed by a description of its implementation within the hospitals of Abuja. The evaluation of performance will comprise assessing the TBDx™ platform by comparing results obtained with the platform against culture (the ‘gold standard’). It is expected that using a high sensitivity algorithm, the TBDx™ platform will result in a much higher sensitivity than manual smear microscopy, but lower specificity, thus requiring a confirmatory test. Using a lower sensitivity algorithm, it is expected that TBDx™ will produce similar performance to smear microscopy, but under a fully automated process. As these performance characteristics could be useful in different settings and can be selected a priori, the study will assess both a high sensitivity algorithm (which also has an expected low specificity) and a low sensitivity algorithm which is the expected sensitivity when conducting manual smear microscopy.
Each patient will be asked to provide three (3) sputum samples, collected using the spot-spot-morning scheme, as illustrated in the figure below. One sputum sample will be cultured in liquid and solid media. Two sputum samples will be used to prepare fluorescence smears, preferably stained using an automated staining machine, if available, and one of the two smears will be tested using Xpert. The two fluorescence smears will be read and graded manually and one will be read with TBDx, grading 300 fields per smear.
Reorganization of Laboratories to Support TBDx™ Scale-Up
The TBDx™ platform will be considered to be of sufficient value for deployment in the health system if the combined performance of TBDx™ (at a high sensitivity setting) and Xpert has:
- A higher accuracy than the expected 86% accuracy of diagnosis when using manual LED-FM smear-microscopy alone
- The specificity of the approach is not lower than 90% (in other words, the difference is not more than a 5% different than the expected LEDF-FM specificity of 95%)
If these premises are met, the project will conduct a descriptive assessment to document the modifications needed for the reorganization of the TB diagnostic services in selected district hospitals of Abuja FCT Metropolitan Area Council (AMAC).
It is expected that this reorganization would be resilient for testing large numbers of samples (up to 150-200/day), and the number of hospitals participating in the project will increase progressively until reaching 80-90% capacity or when staff feels they have reached full capacity within 8-hour shifts.
Once the study results are known and officials reach the conclusion that the combinations of TBDx™ and GeneXpert would be beneficial to the Nigerian health system, the Government of Nigeria believes they will have a very keen interest in expanding the deployment of TBDx™ in Nigeria.