We are very excited to roll out the data from our first-ever data-driven journalism article! We surveyed hundreds of people from our database who are living with or caring for someone with diabetes to identify “the habits of a great A1c.”
This article serves to describe exactly how the survey was designed and analyzed and addresses some important limitations in interpreting the outcomes.
The Diabetes Daily Thrivable Insights Panel
Thrivable Insights is a diabetes research panel that is currently made up of about 20,000 people living in the United States. With this panel, we can utilize survey-based, phone-based, and other virtual research methods to gather important insights into various aspects of living (and thriving) with diabetes. You can learn more about this research effort and even join the panel by visiting this site.
Survey: Habits of a Great A1c
We queried hundreds of our Thrivable Insights panel members to learn more about what habits are common to those who achieve the most optimal A1c levels. We asked detailed questions about various aspects of management, including meal planning and patterns (e.g., carbohydrate intake habits), insulin use strategies, exercise habits, technology use, healthcare provider visits, socioemotional factors, and much more! Next, we evaluated the responses as they related to the patients’ A1c levels.
After the survey was completed, we segmented the data to examine separately the information reported by type 1 diabetes patients and caregivers vs. type 2 diabetes patients and caregivers. We examined the differences between specific habits of survey participants who reported an A1c of <6.5% vs. those who reported an A1c of >8.0%. Specifically, we analyzed the differences in the proportion of individuals who reported a particular habit using a comparison of proportions calculator to identify the statistically significant differences between the groups.
Study Strengths and Limitations
The strengths of this study include the robust sample sizes across all analyses, as well as a rigorous survey design and data analysis processes.
The main limitation of the study is that all the data (including the A1c levels) are self-reported. Also, while the specific statistical analysis performed can be used to infer an association between a particular habit and A1c level, it does not serve to correlate a particular habit with any outcome.
Finally, while we analyzed patients with type 2 and type 1 diabetes separately, we did not account for differences in patient age, sex, diabetes duration, or socioeconomic status in this initial set of analyses. Notably, approximately 75% of the respondents were female. We look forward to examining the data further in the near future to learn about what effects these demographical differences may have on the results!
This month, we will begin to roll out the data to show the reported differences between the two A1c groups in patients with type 1 and type 2 diabetes. Stay tuned to learn what specific management habits (and other factors) were common to those who reported achieving optimal A1c levels!
Post Views: 209