2015 APHA Annual Meeting & Expo

Home Events & Meetings Annual Meeting Schedule & Program

Recorded Annual Meeting Presentations (RAMP) are now available for purchase.
Only $90 for presenters, $140 for registered attendees and $200 for non-attendees.

Online Program

325934
Bayesian Small-area Estimates of Self-reported Kidney Disease Prevalence in the United States by County


Wednesday, November 4, 2015 : 11:10 a.m. - 11:30 a.m.

Sai Dharmarajan, Department of Biostatistics, University of Michigan, Ann Arbor, MI
Jennifer Bragg-Gresham, PhD, School of Public Health, University of Michigan, Ann Arbor, MI
Hal Morgenstern, PhD, Departments of Epidemiology, Environmental Health Sciences, and Urology, University of Michigan, Ann Arbor, MI
Brenda W. Gillespie Gillespie, PhD, Center for Statistical Consultation and Research, University of Michigan, Ann Arbor, MI
Yi Li, PhD, Department of Biostatistics, University of Michigan School of Public Health, Ann Arbor, MI
Neil Powe, MD, MPH, MBA, Department of General Internal Medicine and Center for Vulnerable Populations at San Francisco General Hospital, University of California San Francisco, San Francisco, CA
Delphine Tuot, MD, Division of Nephrology at San Francisco General Hospital, University of California, San Francisco, San Francisco, CA
Deborah Rolka, MS, Division of Diabetes Translation, Centers for Disease Control and Prevention, Atlanta, GA
Sharon Saydah, Ph.D., Division of Diabetes Translation, Centers for Disease Control and Prevention, Atlanta, GA
Rajiv Saran, MD, MS, School of Medicine, Department of Nephrology, University of Michigan, Ann Arbor, MI
Audio (mp3) recording Recorded Presentation (mp4)
Objective: The estimated prevalence of self-reported kidney disease (SR-KD) in the US, as obtained from national survey data, is about 2.5%.  However, national surveys are not designed to provide estimates for small regions. As part of the national Chronic Kidney Disease surveillance system, we sought to estimate the SR-KD prevalence in US counties using 2011 and 2012 data from the Behavioral Risk Factor Surveillance Study (BRFSS) to understand geographic variation.

Methods: We estimated county-level prevalence of SR-KD using a Bayesian multi-level disease mapping model with spatially correlated random effects at the county level. Our model produced stable age-specific estimates for each year that were post-stratified using Census 2013 data to obtain an overall estimate of SR-KD prevalence for each county in that year.

Results: Estimated county-level prevalence of SR-KD in contiguous US counties ranged from 1.1 to 8.1 % (99th percentile: 4.4 %) in 2011 and from 1.2 to 11.3 % (99th percentile: 5.1 %) in 2012. Estimated SR-KD was lower in the north-east, and higher in the southern and western regions of the country. For counties with 500 or more respondents, our model-based estimates of SR-KD prevalence agreed with the corresponding BRFSS direct sample-based estimates, yielding a relatively small root mean squared error of 0.42 % for 2011 and 0.31 % for 2012.     

Conclusions: We believe this is the first attempt to estimate SR-KD prevalence at the county level in the US. Our approach yields estimates with reasonable statistical precision for small counties and compares well with direct sample-based estimates for large counties.

Learning Areas:

Biostatistics, economics
Epidemiology
Public health or related public policy

Learning Objectives:
Describe the geographical variation, by county, of self-reported kidney disease in the United Sates. Explain a statistical method to estimate prevalence of chronic disease in smaller regions using publicly available national survey data.

Keyword(s): Surveillance, Epidemiology

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I am a Biostatistics PhD candidate at the University of Michigan. As a graduate student research assistant at University of Michigan - Kidney Epidemiology and Cost Center I have been involved in research projects undertaken by the National Chronic Kidney Disease Surveillance System team here and learnt a lot about the epidemiology of kidney disease. My training in Biostatistics has helped me contribute to the team's surveillance efforts by undertaking the current project.
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: 5086.0: Applications of Spatial Data Analysis