Background:
There is a need for a testing algorithm to confirm HSV-2 serology in the absence of western blot testing. Additionally the effect of HIV serostatus on HSV-2 serology remains problematic.

Objective:
To determine if HIV-1 serostatus affects HSV-2 results, and to develop an algorithm with a secondary test that approaches HSV-2 WB results on sera from Rakai, Uganda.

Method:
A population based survey of 460 samples (149 HIV+, 311 HIV-) from Rakai Uganda were tested by WB, Focus ELISA, Kalon ELISA, and BioKit Rapid test for HSV-2. We also compared combinations of these three assays to determine which best approaches WB results.

Result:
Compared to WB, Focus, Kalon and Rapid tests had sensitivities of 99%, 93% and 96% with specificities of 50%, 91% and 56% respectively. When segregated by HIV serostatus, the HIV positive samples had sensitivities of 100%, 94% and 97% with specificities of 38%, 80%, and 41% respectively. The HIV negative samples had sensitivities of 98%, 93%, and 95% with specificities of 63%, 92%, and 57% respectively. To maximize accuracy, all initial Focus test with index values between 1.0 and 4.0 were confirmed by Rapid and Kalon tests. Those with a strong Rapid result or a Kalon index value above 1.2 were considered positive. This algorithm provided a sensitivity of 97%, specificity of 85%, with a PPV of 86% and an NPV of 96%.

Conclusion:
HIV serostatus affects the specificity of HSV-2 testing, but an algorithm can be used to approach WB results.

Implications:
A testing algorithm can be used in settings where WB is not available.