Brief Report: Negative Controls to Detect Selection Bias and Measurement Bias in Epidemiologic Studies