Performance of a low-cost, rapid point-of-care CD4 test in resource-constrained settings

In high-burden countries, HIV is a significant cause of maternal death, accounting for nine percent of deaths of pregnant women in sub-Saharan Africa (42.5 percent in South Africa) and contributing to poor progress towards reducing maternal mortality in this region.

HIV infection during pregnancy also contributes to poor perinatal outcomes including stillbirth, pre-term birth and low birth weight.

Delayed initiation of antiretroviral interventions contributes significantly to HIV infection of infants and HIV-related maternal deaths.

Assessing CD4 count at first antenatal visit for HIV-infected pregnant women using rapid point-of-care (POC) CD4 testing would allow for same day initiation of antiretroviral intervention either for the mother’s own health or for the prevention of mother-to-child transmission which would substantially reduce maternal and newborn deaths.

A simple and accurate instrument-free, low-cost POC CD4 test that is applicable in poor, hard-to-reach settings could sustainably increase uptake of antiretroviral intervention and reduce loss to follow-up associated with centralised laboratory-based testing.

Objectives:

• To prospectively assess the test validity (misclassification, sensitivity, specificity, negative predictive value, and positive predictive value) in a field setting of the newly developed Burnet POC CD4 as performed by nurse-midwives compared to the reference method (a laboratory-based flow cytometry assay which is considered to be the gold standard).
• To compare and assess the feasibility and acceptability of using the Burnet POC CD4 test during first antenatal care visit.

June 2014 - June 2017

Activities:

The project was implemented in a high throughput reference hospital in South Africa (Rahima Moosa Mother and Child Hospital), and Coast Provincial General Hospital in Mombasa, Kenya. Pregnant women over 18 years of age, attending antenatal services for the first time, and testing HIV-infected and confirmed positive using two serial HIV rapid tests were invited to participate in the study.

The study was divided into two phases:

• Phase 1: Recruiting 100 women across the two sites for Beta testing of the POC device.
• Phase 2: Field validation of the CD4 test through enrolling 175 women across the two sites. Women were recruited over a 4-6 month period. A total of 275 consenting women were asked to provide a 5mL vacutainer EDTA tube of whole venous blood (for laboratory-based Burnet POC CD4 and flow cytometry assay). Women in phase 2 of the study were also be asked to provide finger-prick blood samples for the POC CD4 tests.

Health care workers and laboratory personnel were trained to perform and assess CD4 cell count in the following ways:

• Burnet POC CD4 from finger-prick at antenatal clinic (trained nurse-midwives) – Phase 2 only
• Burnet POC CD4 from venous blood at laboratory (trained laboratory technologist).
• Laboratory-based flow cytometry assay (gold standard methodology; e.g. Becton Dickinson FACSCount) from venous blood (trained laboratory technologist).

The primary comparison for this study was to determine the test validity of the Burnet POC CD4 against the reference standard of flow cytometry. Test validity was measured as the proportion of correctly identified women with counts ≤350 cells/µL (sensitivity) and the proportion that correctly identified women with counts >350 cells/µL (specificity). In addition, the positive predictive value (proportion of CD4 tests that read “treat” that are for women with actual CD4 counts 350 cells/µL) was also measured