Most of the dominant HIV testing kits used in Kenya and several other African countries have failed crucial thresholds set by the World Health Organisation.
Of the eight most widely used HIV Rapid Diagnostic Tests (RDTs), seven of which are pre-qualified by the WHO, only one met the recommended threshold in a recent evaluation.
The evaluation followed what WHO and partners say are increasing cases of HIV misdiagnosis, reaching up to 10.5 per cent in some African countries.
About 1.3 million Kenyans are estimated to be living with HIV.
Blood samples were collected from HIV clinics in Kenya, Guinea, Uganda, Cameroon, and the Democratic Republic of the Congo and tested with the eight kits at a WHO collaborating laboratory in Antwerp, Belgium.
The evaluation, the final report shows, was collaborated by the ministries of Health of the five African countries and the international charity group, Médecins sans Frontières (MSF) of Netherlands.
“Individual kits performed more poorly than in WHO evaluations, with only one test (STAT-PAK) meeting the recommended thresholds,” says the report published last month in the Journal of the International AIDS Society.
For example, one kit is shown to have returned a total of 438 false positives from the 2,785 samples collected from August 2011 to January 2015.
“Indeed, there were cases of misdiagnosis. Quite some actually!” study leader Cara S. Kosack of MSF told The Standard last week.
MSF runs several HIV programmes in Africa.
The report says HIV misdiagnosis has been a problem in some of the MSF programmes in sub-Saharan Africa, hence the evaluation.
The final report has interesting findings. For example while WHO has consistently blamed misdiagnosis on human error, it now emerges that the kits may be largely to blame.
The authors are categorical that the misdiagnosis has little to do with human factors normally associated with poor Africa such as unqualified staff, poor storage of samples and kits, or other methodological factors.
By performing the evaluation in a centralised and highly advanced laboratory and finding results similar to those back in Africa, the authors say it showed an inherent problem with the kits.
The evaluation identified three specific problems: results were found to differ by gender, geographical origin of the sample, and type of clinic the specimen had come from.