In its most thorough external evaluation yet, TBDx™ was the beneficiary of a highly skilled and professional staff from the National Institute for Communicable Diseases (NICD) in Johannesburg, who dedicated considerable human and material resources to execute a well-planned test protocol. During the technology evaluation, TBDx™ processed 1,249 patient slide specimens and acquired approximately 375,000 digital images from April 15 – May 2, 2013.
Though TBDx™ has completed its processing and classification of the cases, and the microscopists have completed their reading of all the patient slides, the study protocol requires that performance be compared against culture, the ‘gold standard’. It will be another 4-6 weeks before the culture results are known for each of these cases. Additionally, all cases classified as positive by TBDx™ are to be confirmed by PCR (polymerase chain reaction) amplification testing to determine the benefit of using TBDx™ as a screening technology. This confirmation will be completed by the end of this week.
Quantitative analysis cannot be developed until culture results are available. However, during the evaluation, the NICD – Center for Tuberculosis (CTB) staff analyzed 300 processed cases as a preliminary analysis. These preliminary findings were encouraging:
- TBDx™ classification of positive cases matched very closely with positive classifications as determined by the microscopists. Among the positive cases that were reviewed, TBDx™ and the microscopists matched 46 of 48 cases. Of the two unmatched cases, one will require culture to make the final determination, and the other involved a case that should have been excluded from the study.
- TBDx™ was classifying cases as negative that the microscopists were classifying as negative. TBDx™ and the microscopists matched on 160 negative cases.
The most interesting part of the analysis, and one that everyone is anxious to better understand, when culture results become available, is the Scanty cases. These cases have a very low count of TB bacteria present in the sputum smear. These are the hardest cases to identify, and the cases most often missed by routine smear microscopy. Most importantly, these are the cases that the lab would hope a ‘screening’ technology could detect and that could be confirmed by another test such as PCR or culture.
The World Health Organization (WHO) considers a case to be Scanty when 1-9 bacilli are found in 100 fields of view analyzed in a sputum smear. The question that arises from these cases, and the data from the culture results will help to determine the best lab practice to consider when using the TBDx™ technology to identify Scanty cases, is when to classify a TBDx™ Scanty 1-2 case as Negative? If most of the Scanty cases are a very low count, such as Scanty 1-2, will the culture results show that it is cost beneficial to run a second test on these cases, or will it show that the cost is not worth the few cases that are detected in this category. Answers to these questions will become clearer when all of the data has been amassed and analyzed as per the study protocol.
In our next post we will present some of the ‘Lessons Learned’ over the 21- day test period. There were several. Among them is the quality of smear that results from using ‘target’ slides. We have learned that using circumscribed slides with a circle drawn in the middle, where the sputa is placed, produces high quality stained slides and well-focused digital images.
Before ending this post, we would like to extend our sincere appreciations to the NICD, and recognizing several individuals within the CTB who provided leadership and support, ensuring a well-executed test of our technology:
Dr. Nazir Ismail, Director of the CTB
Dr. Shaheed Vally Omar, Medical Scientist with the CTB
Dr. Chikwe Ihekweazu, who leads the Epidemiology Division of CTB
Minty Van Der Meulen, Microbiologist
George Ngconjana, Microbiologist
Zacharia Mabena, Lab Technician