Commentary on Graham et al.: Biochemical verification of abstinence in remotely conducted smoking cessation trials should not be a universal design requirement for rigor.