Statistics in Medicine
Volume 27, Issue 20 pp. 3925-3926
Preface
Free Access

Welcome and opening remarks 2007 CDC-ATSDR symposium on statistical methods

James W. Stephens PhD, 

James W. Stephens PhD

Associate Director for Science, Centers for Disease Control and Prevention, Mail-Stop D-50, 1600 Clifton Road, NE, Atlanta, GA 30333, U.S.A.

Search for more papers by this author
First published: 17 April 2008
https://doi.org/10.1002/sim.3277

I am delighted to welcome you to the Eleventh Biennial CDC-ATSDR symposium on statistical methods. This symposium is one the largest of our symposia, with more than 200 individuals registered for the individual sessions and more than 130 registrants for the two short courses on Bayesian estimation and geographical information systems. I want to acknowledge and thank the CDC Statistical Advisory Group and the symposium organizing committees, and give special thanks to our co-sponsors the American Statistical Association and Emory University-Rollins School of Public Health for their efforts in making this symposium the best attended and one of the most exciting and relevant symposia to public health.

I would like to comment on the title of this year's symposium, ‘Analyzing and Mapping Health Inequities for Impacting Policies to Eliminate Disparities’. This title reflects not only the focus on a high-priority field for public health, but also demonstrates the increasing recognition of how the field of statistics contributes to the entire range of activities that comprise public health. During this symposium you will hear about both traditional uses of statistics and the development of new methodologies. However, this program goes far beyond what you might expect from the ‘Symposium on Statistical Methods’ title, and covers a wide array of practical applications of statistics in areas such as

  • racial and ethnic disparities,

  • strategies for hard to reach populations,

  • and access and availability of healthcare.

Looking through the program I note that the sessions provide a stimulating combination of presentations on methodology combined with practical applications, surely an indication of the increasing focus by statisticians on applying their methods in ways that have a direct effect on public health.

I would like to take this opportunity to highlight two broad areas of public health of specific interest to the CDC Office of the Chief Science Officer where statisticians and the statistical sciences can and must play a crucial role. The first example is the challenge that public health institutions and professionals face in communicating our science and recommendations in a manner that instills confidence and ensures that our recommendations will be trusted, accepted, and implemented. A little over a year ago a workgroup of the CDC Excellence in Science Committee was formed to develop recommendations for evidence-based standards for CDC recommendations and guidelines. This committee—the Quality of Evidence workgroup—produced a report with eight recommendations for improving CDC's approach to guidelines, including:

  • a recommendation for CDC to develop more consistent evidentiary standards as appropriate for the various types of guidelines,

  • a recommendation for CDC to establish ongoing mechanisms to evaluate the use and impact of guidelines,

  • a call for an increase in transparency in guidelines development so that people know how and why guidelines are developed.

To increase transparency, the report calls for all CDC guidelines to provide a methods section that discusses:
  • (a)

    How the guideline adds to or differs from what was available previously.

  • (b)

    Who was involved in the guideline's production and how.

  • (c)

    What information base was considered?

    • (i)

      The rationale for considering this evidence base.

    • (ii)

      What potential information bases were not consulted.

  • (d)

    How the information considered supports or relates to the conclusions or recommendations.

The recommendations for transparency are simple and easy to implement, but could have profound implications for public if fully adopted.

A second current challenge for public health is the need to provide clear, credible recommendations even when the science is uncertain. The public health community may be better at communicating uncertainty than many other areas of science and health, but we still face significant challenges in this arena, for example, in communicating the results from a newly completed study that has large uncertainties. And it is a challenge that is increasing due to emerging issues such as pandemic influenza where we must make recommendations even when we have limited data. This is not primarily a methodological challenge, but a challenge in how we communicate our work and how we translate and present our findings to the public and to policymakers. Those of you in the statistical sciences have an important role in helping us meet this challenge.

In closing I would like to highlight the importance of this symposium in bringing together statisticians from all levels of public health research and practice, from within the government, academia, and the private sector, to discuss ways in which the statistical sciences can better help CDC and public health accomplish its mission. On behalf of CDC and ATSDR, I wish you a successful symposium, and I look forward to seeing the impact of this work on public health in the coming years.

    The full text of this article hosted at iucr.org is unavailable due to technical difficulties.