Gravar-mail: Investigating the Impact of Selection Bias in Dose-Response Analyses of Preventive Interventions