Background: An objective and accurate method that measures adherence to vaginal microbicide gel regimens during clinical trials could provide more accurate estimates of microbicide efficacy, aid in targeting adherence promotion resources, and enable objective assessment of adherence promotion strategies.
Methods: We evaluated 4 methods to assess whether or not gel applicators had been vaginally inserted. At the study site, 50 women inserted hydroxyethylcellulose universal placebo gel through a polypropylene vaginal applicator and handled, but did not insert a second "sham-inserted" applicator. Applicators were discarded into a container capped with a medical event monitor system (MEMS) that recorded the time and date of opening. Fifteen additional participants did likewise at 2 study site visits, and administered gel on 6 intervening days at home. Applicators were scored as inserted, or not, by direct inspection under ambient light, ultraviolet (UV) light, staining with Alcian blue, and microscopic detection of vaginal cells stained with iodine.
Results: Mean sensitivity/specificity of 2 readings each by 3 test readers for UV, Alcian blue, ambient light, and iodine methods were 84/83, 79/83, 76/63, and 65/80%, respectively. Sensitivity of all methods was significantly higher in applicators inserted after one or more prior insertions of gel, with the highest sensitivity (95%) obtained with UV. MEMS caps accurately recorded applicator disposal time.
Conclusions: The modest accuracy of all 4 methods for applicator insertions without prior gel applications may limit their accuracy in monitoring coital regimens. However, for daily dosing regimens, MEMS monitoring and UV inspection should provide a rapid, reliable, and quantitative assessment of adherence.