273116 Aggregate Data Analysis for Public Health

Wednesday, October 31, 2012 : 9:30 AM - 9:50 AM

James Lightwood, PhD, MA , Clinical Pharmacy, University of California, San Francisco, San Francisco, CA
  
Analysis of multiple time series of aggregate population based statistics can be useful for understanding determinants of population health and for forecasting.

Most current estimates of health expenditure attributable to smoking in large populations are calculated using simulation models that use individual level cross sectional data from national health and healthcare expenditure surveys. Other estimates use time series analysis of natural experiments on local populations. Both these methods have drawbacks in modeling large populations and forecasting; particularly the need for modeling and associated assumptions to generalize the results to a larger target population.

Estimates based on simple statistical models and aggregate data for large populations can be useful supplements to other types of analysis. This session will discuss the development and use of this type of analysis with a focus on tobacco control and state and regional health expenditures attributable to smoking behavior.

The statistical methods will be introduced and illustrated with application to tobacco control in the United States. The strengths and weaknesses of the approach will be compared to other types of analysis, such as simulation models based on survey data using individual respondents as the unit of analysis. Issues such as omitted variables, time series properties of the data, avoiding ecological fallacy and forecasting will be discussed.

Learning Objectives:
Following the presentation, audience members will be able to describe the strengths and weaknesses of different approaches to estimating the costs associated with tobacco use at the population level, and learn the basic elements of evaluating a study using time series methods and aggregate data.

Keywords: Economic Analysis, Tobacco Policy

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I am an applied econometrician whose PhD dissertation has focus on applied nonstationary time series analysis; have published papers on topic in PLoS Med and Soc Sci Med; past consultant with WHO on economics on tobacco control with three published review articles and monograph chapters.
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.

Back to: 5101.1: Statistical Methods in Policy Research