Comparison of visual and ultraviolet light inspection versus DNA/protein biomarkers to assess product adherence with vaginal microbicide applicators.

BACKGROUND: Objective biomarkers of product use and protocol compliance are urgently needed. We compared the sensitivity and specificity of DNA and protein-based biomarkers, obtained from used vaginal gel applicators, to visual inspection of those applicators under ambient light (visual inspection of returned applicator [VIRA]) and ultraviolet light (UVL). METHODS: Forty women inserted hydroxyethylcellulose placebo gel vaginal applicators under direct observation. Applicators were evaluated by VIRA, UVL, and DNA/protein-based methods at 2 time points: within 7 days of the visit and after storing applicators for approximately 30 days. Semen biomarkers were assayed from vaginal swabs and returned applicators. RESULTS: The overall sensitivity and specificity of DNA and protein-based biomarkers in determining vaginal insertion versus sham handling of returned applicators were 98.3% and 100%, respectively, at both 7- and 30-day evaluations. The overall sensitivity and specificity of VIRA at 7 and 30 days after collection were significantly lower than those of DNA and protein-based biomarkers. Ultraviolet light inspection also had significantly lower overall sensitivity and overall specificity compared with DNA and protein biomarkers. The sensitivity of DNA and protein-based biomarkers for detecting insertion of wiped applicators was 95%, whereas the sensitivity of VIRA (range of 24%-28%) and UVL inspection (range, 38%-84%) was low for this subset. It was feasible to obtain semen biomarkers from vaginal swabs and returned used applicators. CONCLUSIONS: DNA and protein-based biomarkers offer significantly higher sensitivity and specificity compared with VIRA and UVL assessment. The accuracy of these objective biomarkers is maintained despite storage of returned products for approximately 30 days and under conditions potentially modeling field use.