Application of the case-crossover design to reduce unmeasured confounding in studies of condom effectiveness.

Application of the case-crossover design to reduce unmeasured confounding in studies of condom effectiveness.