Online Program

Exploring the co-occurrence patterns of multiple chronic diseases in the U.S. adult population using association rule discovery

Monday, November 4, 2013

Yoon-Ho Seol, PhD, Department of Clinical and Digital Health Sciences, Augusta University, Augusta, GA
Genny Carrillo-Zuniga, MD, MPH, ScD, School of Rural Public Health South Texas Center, Texas A&M, Health Science Center, McAllen, TX
Anita Bhat, MD, PhD, Department of Clinical and Digital Health Sciences, Augusta University, Augusta, GA
Miguel A. Zuniga, MD, DrPH, Department of Health Policy and Management, Texas A&M Health Science Center, McAllen, TX
There is an increasing prevalence of adults with multiple chronic conditions and the effective management of these conditions is complicated by the presence of significant comorbidities. The primary objective of this study was to explore the association patterns of multiple chronic conditions among U.S. adults and to examine to what extent these patterns vary by demographic characteristics and other risk factors.

We analyzed data from the nationally representative 2010 National Health Interview Survey (NHIS) using association rule mining. Based on the criteria of multiple chronic conditions used by CDC (Centers for Disease Control and Prevention), we selected nine chronic conditions including hypertension, heart disease, and diabetes using the responses of relevant survey items (e.g., Have you ever been told by a doctor or other health professional that you had hypertension?). We first applied association rule mining on these nine conditions to describe the overall prevalence and co-occurrence patterns of the conditions. We also compared each condition by limiting the number of the conditions that must appear on the antecedent part and the consequent part of the rules. To further explore how co-occurrence patterns are influenced by demographic and other characteristics, we examined several factors including age, gender, race, ethnicity, income, education, insurance, health status, residential area. We performed association analysis inside each of subgroups of the population by these factors and examine the differences in frequent association patterns using their support and confidence measures. Additionally, national estimates of corresponding subpopulations were derived using these association rules. The results of this study identified useful co-occurrence patterns of the chronic conditions and distinct subpopulations that were shaped by demographic and other risk factors. Developing a comprehensive understanding of these findings would help plan and implement successful public health programs and policies that are more responsive to the needs of U.S. adults.

Learning Areas:

Assessment of individual and community needs for health education
Communication and informatics
Public health or related research

Learning Objectives:
Identify the co-occurrence patterns of multiple chronic diseases among U.S. adults. Assess the influence of demographic and other risk factors on the co-occurrence patterns. Discuss the opportunities and challenges of performing association rule mining on national health survey data

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I contributed to the design, data collection and analysis, and interpretation of this study.
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: 3406.0: HIIT Poster Session 2