Table 2

Impact of adherence and non-endpoint HIV infections on the effectiveness of tenofovir gel in HIV prevention in the CAPRISA 004 tenofovir gel trial.

No. of HIV infections/women yearsHIV incidence
TenofovirPlaceboNTenofovir gel (95% CI)Placebo gel (95% CI)Incidence rate ratioEffectiveness95% CIP value
Overall effectiveness of tenofovir gel
   HIV endpoints38/680.660/660.78895.6 (4.0, 7.7)9.1 (6.9, 11.7)0.6139%6, 600.017
Site-specific effectiveness
   Rural25/484.742/461.26115.2 (3.3, 7.6)9.1 (6.6, 12.3)0.5743%5, 670.023
   Urban13/195.918/199.52786.6 (3.5, 11.3)9.0 (5.3, 14.3)0.7426%-59, 670.380
HIV endpoints by levels of adherence*
   High adherers
      (>80% gel adherence)
11/259.225/269.43364.2 (2.1, 7.6)9.3 (6.0, 13.7)0.4654%4, 800.025
   Intermediate adherers
      (50–80% adherence)
10/159.810/99.71816.3 (3.0, 11.5)10.0 (4.8, 18.4)0.6238%-67, 770.343
   Low adherers
      (<50% gel adherence)
16/258.525/290.63676.2 (3.5, 10.1)8.6 (5.6, 12.7)0.7228%-40, 640.303
   Sensitivity analyses
   HIV endpoints plus
HIV infection not meeting
   protocol definition
39/680.660/660.78895.7 (4.1, 7.8)9.1 (6.9, 11.7)0.6337%4, 590.023
   HIV endpoints
      plus ineligibly enrolled
40/720.163/698.610755.6 (4.0, 7.6)9.0 (6.9, 11.5)0.6238%7, 600.015
   HIV endpoints plus women
      with post-trial infection
39/680.664/660.78895.7 (4.1, 7.8)9.7 (7.5, 12.4)0.5941%11, 610.015
   Per protocol analysis†32/589.253/575.48895.4 (3.7, 7.7)9.2 (6.9, 12.0)0.5941%7, 630.017
   All HIV infections‡43/720.176/698.810856.0 (4.3, 8.0)10.9 (8.6, 13.6)0.5545%19, 630.003
   Adjusted analysis§3860889Hazard ratio = 0.6337%6, 580.025

*Adherence could not be calculated for the five women who reported no sex during their follow-up in the study.

†Excludes all visits after 3 months’ interruption of drug supply.

‡All HIV infections = protocol-defined HIV endpoints (n = 98 women) + HIV infection not meeting protocol definition (n = 1 woman who did not have 2 RNA PCR results) + HIV infections among ineligibly enrolled women (n = 5 women) + posttrial HIV infections (n = 5 women) + window period HIV infections in eligible women (n = 8 women) + window period HIV infections in ineligibly enrolled women (n = 2 women).

§Adjusted for the following baseline covariates: age, site, parity, number of sexual partners (past 30 days), presence of STI, anal sex, contraceptive method, HSV-2 antibody status, and condom use; HSV-2 status is indeterminate in four women and missing in five women.