Table 1:          Hypothetical Outcomes, Relative Risks and Likely Regulatory Response to a Randomized Control Trial Comparing Microbicidal Product with Placebo and to “Condom Only” Arm*§

                                                  Two-arm           Two-arm           Three arm        

vs placebo        vs Condom       Placebo            Condom            Both controls     Possible

                                            Only         Only¹                Only²                FDA Action            Interpretation

A          RR <     1          RR <     1          Approve            Approve            Approve            Microb is effective

B          RR >     1          RR >     1          Disapprove        Disapprove        Disapprove        Microb increases risk

C          RR =     1          RR =     1          Disapprove        Disapprove        Disapprove        Microb has no impact on risk; no worse than condom only    

D          RR <     1          RR =     1          Approve            Disapprove        ??                    Microb better than placebo, but no better than condom only

E          RR <     1          RR >     1          Approve            Disapprove        ??                    Microb better than placebo, but less effective than condom only

F          RR >     1          RR =     1          Disapprove        Disapprove        Disapprove        Microb worse than placebo; equals condom only; Placebo protective?

G          RR >     1          RR <     1          Disapprove        Approve            ??                    Microb worse than placebo; both better than “condom only”

H          RR =     1          RR <     1          Disapprove        Approve            ??                    Microb equivalent to placebo; both better than condom only

I           RR =     1          RR >     1          Disapprove        Disapprove        Disapporve        Microb equivalent to placebo; both worse than condom only

 

 

* HIV incidence compared in experimental microbicide arm vs control arm: “”RR<1.0 denotes increased risk in the control arm as compared with the experimental arm; “RR> 1.0” denotes decreased risk in the control arm compared with the experimental arm; and “RR = 1.0” denotes approximately equal risks of infection in the experimental arm

 

§ These interpretations assume a) that true levels of condom use do not vary across trial arms, and (b) that self-reports of condom use reflect reality.  With a necessarily unblinded “condom-only” arm, both true levels and self-reports of their use might well vary systematically across trial arms, rendering these comparisons uninterruptible.

 

1 Likely response of regulators to the scenario presented, if they were reviewing a two-armed trial of experimental product and a placebo control

2 Likely response of regulators to the scenario presented, if they were reviewing a two-armed trial of experimental product and a “condom-only” control  

3 Likely response of regulators to the scenario presented, if they were reviewing a three-armed trial with both a placebo and “condom only” control arm

 

Source:  Adapted from Stein Z, Myer L, and Susser M. 2003. The design of prophylactic trials for HIV: The case of microbicides. Epidemiology.14(1):80-84.