HOW DO PATIENTS INTERPRET VARIABILITY AND SLOPE IN BLOOD PRESSURE DATA VISUALIZATIONS?

Purpose: Although there are multiple drugs that effectively treat hypertension, uncontrolled hypertension is a significant health problem in the US (Go et al., 2013; Roger et al., 2012). One reason hypertension remains uncontrolled is that we lack adequate methods for collecting and displaying patient blood pressure (BP) data, which can be used in decisions about hypertension treatment. As part of a larger project developing data visualizations to support shared decision making about hypertension treatment, we conducted a series of studies to better understand how patients interpret data variations inherent in graphs of BP.

Method: Across two studies, participants (Internet sample of patients with hypertension) viewed a series of graphs depicting BP data for fictitious patients. In Study 1 (N=51), we systematically varied systolic BP mean (130, 145, 160) and standard deviation (5, 15, 25) in the graphs. In Study 2 (N=50), we varied systolic BP mean and slope (increasing or decreasing). For each graph, participants rated perception of BP control, need for medication change, and subjective risk for heart attack and stroke using a within-subjects design. They also estimated the proportion of systolic BP values out of range.

Results: In Study 1, there was a significant mean by standard deviation interaction on all outcome measures, p<.05. For example, while perceived need for medication change increased with mean BP values, greater variability also significantly increased the perceived need to change medication, particularly when mean BP values were within the normal range; see Figure 1. In Study 2, there was a significant mean by slope interaction, p<.05. BP values increasing over time had a significant impact on perceived need for medication change; however, increasing trends had the greatest impact when BP mean was normal.

Conclusions: The use of data visualization in EHRs has the potential to transform clinical encounters. While the technology to develop these tools is available, little is known about how these data displays will influence patient decision making. These studies have demonstrated that trends and variability have a greater influence on perception of BP control than mean values. These findings are important for the development of data visualizations used for shared decision making in primary care, which must direct attention to clinical meaningful information (mean BP, NOT BP variability) (Hansen et al., 2010).