Presentation
Effects of Trend Information and Visualization Types on Interpretation of Health Data
SessionPoster Session 2
DescriptionThe widespread adoption of electronic health records (EHRs) has transformed the way patients interact with medical information. Access to health data through such digital technology redefines patients as more active participants in their healthcare interactions. As digital technology and access continue to expand, new knowledge and skills will be required to navigate the demands and complexities of health systems, such as confusing medical language or ineffective data visualizations. The primary goal of this research is to better understand how individuals utilize and interpret health data to inform how it can be effectively displayed.
The interface and visualizations used in EHRs create challenges for those interpreting the information. Health results are typically presented in a standard table format, often with unfamiliar units and labels (Morrow et al., 2019; Zikmund-Fisher et al., 2014; Zhang et al., 2020). Health management often requires not only understanding one result in isolation but also tracking and evaluating changes in health data over time. While previous work has examined how different format types affect the interpretation of single data points, it is unclear whether they are equally effective when presenting time series data. Clinical records are a rich source of longitudinal data, which helps reveal health trends over time, track progress, and inform risk predictions (West et al., 2015). Time-series data is typically plotted as a line graph (ScienceDirect, 2001). Given this convention, we aim to investigate how individuals interpret time-series data presented in line graphs and other presentation types to determine which best support understanding and action.
Overall health is described and understood as being multidimensional. It is influenced by, but not limited to, genetic, social, and environmental factors (National Academy of Medicine, 2001; Durch et al., 1997). Understanding the relationship between health and behavior is vital because many diseases are preventable with the adoption of healthy behaviors. Heart disease is the leading cause of death in the United States across men, women, and most ethnic groups, and high cholesterol is a key risk factor (CDC, 2025). Borderline cholesterol results are ambiguous; based on findings from a previous study, approximately half of the participants considered borderline results to be within an acceptable range, while the other half did not (Busciglio et al., 2025). Many people may not act when results are borderline, missing an opportunity for prevention. The current study restricts health results to borderline levels of risk to better understand how people interpret ambiguous yet critical health results. We aimed to gain an understanding of how people interpret trend data, whether they identify patterns across multiple results, and how these interpretations influence their perceived health status and motivation to pursue preventive measures.
We investigated how format type (table, number line, graph) and trend information (none, increasing, decreasing, stable) affect clarity, confidence, perceived risk, and behavioral intention. In an online study, 52 participants evaluated hypothetical cholesterol results. The survey item capturing format clarity asked about how clear/readable participants considered each format type. The second survey item provided data about confidence in interpretation, asking participants to rate their level of confidence in understanding the meaning of the results. Perceived risk ratings asked about the urgency of the results. The final survey item, behavioral intention, asked the participants about their likelihood to take action to reduce their cholesterol.
Our findings suggest that clarity was driven more by trend type than format type, with stable trends being rated as less clear than the increasing and decreasing trend types (all ps < .01). Table format types received higher clarity and confidence ratings than number lines in the stable trend condition (p < .05). Line graphs are a typical convention used to represent time-series data. Contrary to our predictions, line graphs offered no clear advantage over tables or number lines. We also observed higher reported confidence in understanding when interpreting increasing trend types than decreasing (p < .05), and lowest confidence in interpreting stable trends than increasing or decreasing (all ps < .01).
Perceived risk and behavioral intention were the highest for the increasing trend type (all ps < .001). The no trend and stable conditions did not differ from each other, but were both significantly lower than the increasing trend. Perceived risk and behavioral intention were significantly lower for the decreasing trend type than all other conditions (all ps < .05).
Overall, our findings indicate that adding trend information affects how people respond to health data. Responses vary when presented with improving or worsening health results, suggesting that viewers recognize changes across datapoints and consider them in their judgments. When results are ambiguous or borderline, presenting patients with additional data points may help encourage them to take informed action. Presenting data effectively can aid in the detection of disease patterns and help make the data more easily interpretable, ultimately promoting self-management and better patient outcomes (Abudiyab & Alanazi, 2022).
The interface and visualizations used in EHRs create challenges for those interpreting the information. Health results are typically presented in a standard table format, often with unfamiliar units and labels (Morrow et al., 2019; Zikmund-Fisher et al., 2014; Zhang et al., 2020). Health management often requires not only understanding one result in isolation but also tracking and evaluating changes in health data over time. While previous work has examined how different format types affect the interpretation of single data points, it is unclear whether they are equally effective when presenting time series data. Clinical records are a rich source of longitudinal data, which helps reveal health trends over time, track progress, and inform risk predictions (West et al., 2015). Time-series data is typically plotted as a line graph (ScienceDirect, 2001). Given this convention, we aim to investigate how individuals interpret time-series data presented in line graphs and other presentation types to determine which best support understanding and action.
Overall health is described and understood as being multidimensional. It is influenced by, but not limited to, genetic, social, and environmental factors (National Academy of Medicine, 2001; Durch et al., 1997). Understanding the relationship between health and behavior is vital because many diseases are preventable with the adoption of healthy behaviors. Heart disease is the leading cause of death in the United States across men, women, and most ethnic groups, and high cholesterol is a key risk factor (CDC, 2025). Borderline cholesterol results are ambiguous; based on findings from a previous study, approximately half of the participants considered borderline results to be within an acceptable range, while the other half did not (Busciglio et al., 2025). Many people may not act when results are borderline, missing an opportunity for prevention. The current study restricts health results to borderline levels of risk to better understand how people interpret ambiguous yet critical health results. We aimed to gain an understanding of how people interpret trend data, whether they identify patterns across multiple results, and how these interpretations influence their perceived health status and motivation to pursue preventive measures.
We investigated how format type (table, number line, graph) and trend information (none, increasing, decreasing, stable) affect clarity, confidence, perceived risk, and behavioral intention. In an online study, 52 participants evaluated hypothetical cholesterol results. The survey item capturing format clarity asked about how clear/readable participants considered each format type. The second survey item provided data about confidence in interpretation, asking participants to rate their level of confidence in understanding the meaning of the results. Perceived risk ratings asked about the urgency of the results. The final survey item, behavioral intention, asked the participants about their likelihood to take action to reduce their cholesterol.
Our findings suggest that clarity was driven more by trend type than format type, with stable trends being rated as less clear than the increasing and decreasing trend types (all ps < .01). Table format types received higher clarity and confidence ratings than number lines in the stable trend condition (p < .05). Line graphs are a typical convention used to represent time-series data. Contrary to our predictions, line graphs offered no clear advantage over tables or number lines. We also observed higher reported confidence in understanding when interpreting increasing trend types than decreasing (p < .05), and lowest confidence in interpreting stable trends than increasing or decreasing (all ps < .01).
Perceived risk and behavioral intention were the highest for the increasing trend type (all ps < .001). The no trend and stable conditions did not differ from each other, but were both significantly lower than the increasing trend. Perceived risk and behavioral intention were significantly lower for the decreasing trend type than all other conditions (all ps < .05).
Overall, our findings indicate that adding trend information affects how people respond to health data. Responses vary when presented with improving or worsening health results, suggesting that viewers recognize changes across datapoints and consider them in their judgments. When results are ambiguous or borderline, presenting patients with additional data points may help encourage them to take informed action. Presenting data effectively can aid in the detection of disease patterns and help make the data more easily interpretable, ultimately promoting self-management and better patient outcomes (Abudiyab & Alanazi, 2022).
Event Type
Poster Presentation
TimeTuesday, March 244:45pm - 6:15pm EDT
LocationRhinelander Gallery
Digital Health