Defense Date


Document Type


Degree Name

Doctor of Philosophy


Health Related Sciences

First Advisor

Dolores G. Clement


Managing chronic conditions is seen as the public health challenge of the 21st century. The number of Americans with chronic conditions is expected to rise to 157 million by 2020. Diabetes prevalence and costs contribute to the growing problem. Diabetes was the seventh leading cause of death in 2006. Nationally, the cost of diabetes is expected to be $138 billion in 2020. Diabetes leads to multiple and significant complications. The desired outcomes of management of chronic conditions are improvement in clinical status, avoidance of complications, prevention of co-morbid conditions and avoidance of the costs associated with complications. In the 1990s, disease management programs were implemented in an attempt to effectively manage chronic conditions. The primary approach in these programs focuses on individual-level interventions. Despite these efforts, poor outcomes exist. As a result, other approaches to diabetes management are being explored. This study examines a system-level approach to diabetes management versus an individual level one. The system level approach is based on full implementation of the Chronic Care Model, framed in Social Ecology Theory. This retrospective, non-experimental study explores changes in adherence to select diabetes screening guidelines based on the approach to adopted by two health plans. Analyses were conducted on adherence to LDL, A1c, retinopathy and nephropathy screening 2 ½ years after program implementation. In addition, logistic regression analyses were conducted on the predictive impact of approach to chronic care management in relation to changes in adherence. Other variables known to impact health behaviors were factored into the analysis. There were two main findings of the study. The first is that for each of the screenings, there was a statistically significant difference between participants in the two plans. Comparisons of changes in adherence by approach from before implementation to after implementation resulted in a small number of subjects in some cells which can lead to accepting the null hypothesis when it is false. The second is that approach to management was associated with changes in adherence to three of the four screenings. The logistic models, however, account for less than 23% of the variance in adherence, a moderate effect size.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

August 2010