Kits for the diagnosis of infectious mononucleosis compared with the Paul-Bunnell test.

We compared the results obtained with six different test kits for infectious mononucleosis with those obtained with the Paul-Bunnell test. The investigation was carried out in one laboratory using 149 selected pools of patient sera. Each pool was tested three times with the Paul-Bunnell test and once with each kit. The results obtained with the kits were grouped according to the titre found with the Paul-Bunnell test. The percentage of positive results within each group was calculated for each kit. The Paul-Bunnell titre, which would have classified 50% of the specimens as positive, was estimated for each kit and this was designated the 50% cut off value. In general, there was good agreement. However, false positive test results were found rather frequently with one kit (19%) and the 50% cut off values differed. One kit showed a 50% cut off value at about 8, another at about 16, and the rest at between 16 and 32. We suggest the introduction of improved internal quality control combined with external quality assessment.