Estimating the Number of Cases of a Health Outcome Attributable to Risk Factors Using Complex Survey Data

• Activity Number: 688
• Type: Contributed
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Lin Tian*
• Companies: CDC
• Keywords: Health outcome ; risk factor ; survey ; population attributable fraction ; case

Abstract

Estimating the number of cases of a health outcome attributable to risk factors is important in judging public health priorities. However, this method is infrequently used in research due to the inability to readily compute an estimate using popular statistical packages, especially for national complex survey data. The goal was to estimate the number of cases of attention deficit hyperactivity disorder (ADHD) attributable to low birth weight and preterm delivery using hypothetical complex survey data. First, component population attributable fractions (PAFs) were estimated with the stratum specific relative risk ratios from the multivariable logistic regression models and the corresponding proportions of exposed cases in the strata; confidence intervals around the PAF estimates were calculated using the Bonferroni inequality. Second, the number of cases attributable to each stratum was calculated by multiplying the PAF by the weighted total number of ADHD cases. Finally, the Monte Carlo simulation method was used to calculate the interval estimates. SAS-callable Sudaan macro codes created by the author to carry out the calculations will also be presented.