A Personalized Dashboard to Educate Veterans at Risk of Stroke
Mahesh Merchant PhD. PhD MSc
Richard L. Roudebush VA Medical Center, Indianapolis, IN
Funding Period: June 2010 - September 2011
Personalized risk communication methods have been used to educate patients who are at high risk of developing a cardiovascular event. These methods include personalized counseling, printed brochures and individualized graphs describing the status of the risk factors. Very few studies have provided an interactive tool for patients to use to compare the relative change in risk that might result from modifying their risk factors.
The goals of this pilot study were to: (a) develop a prototype of a personalized dynamic dashboard to educate veterans and help them make informed decisions about modifying their vascular risk factors (hypertension, diabetes mellitus, hyperlipidemia, and tobacco use) to reduce the probability of a vascular event; (b) test the usability of this tool from the patient's perspective; and (c) provide preliminary data regarding the effectiveness of this tool in terms of changes in lifestyle or risk factor control compared with nurse-education.
This was a three-arm randomized controlled study. Veterans from the panel of one experienced general internist were randomly assigned to one of three groups: (1) dashboard with nurse-education, (2) nurse-education alone, (3) usual care. To be eligible, patients must have had at least two of the following: body mass index >30 kg/m2; current tobacco use; systolic blood pressure >140 mm Hg or diastolic >90 mm Hg; LDL-cholesterol >130 mg/dL, or hemoglobin Hb1Ac >8%. We developed the dashboard in our Human Computer Interface/Information Technology laboratory. Patients in the dashboard group received two in-person counseling sessions regarding methods for controlling risk factors, and were asked to use the dashboard tool to enter values for their own risk factors and to observe the relative effects of varying their risk factor control on their overall vascular risk. The nurse-education consisted of two in-person counseling sessions regarding methods for controlling risk factors. Both intervention sessions lasted approximately 30 minutes; therefore, the majority of the time was used to interact with the tool in the dashboard group whereas the entire time was used for personalized risk factor discussion in the nurse-education group. For the usual care group, data were retrospectively collected about the primary care visits and laboratory tests which took place during the same period as the visits for intervention patients.
A total of 77 patients were enrolled (25 dashboard with nurse-education, 27 nurse-education alone, and 25 usual care); 24 dashboard with nurse-education and 25 nurse-education alone patients completed the study with one withdrawal in the dashboard with nurse-education group and 2 withdrawals in the nurse-education group. The three groups were statistically similar in terms of baseline characteristics and risk factor control at baseline.
Patients found the tool easy to use: median score at baseline of 6.3 (range 4.5-7.0, inter-quartile range 1.2) and at the 6-month visit of 6.4 (range 5.0-7.0, inter-quartile range 1.3) on a 7-point Likert scale. Patients were also very satisfied with the tool: median score at baseline of 5.8 (range 3.0-7.0, inter-quartile range 1.5) and at the 6-month visit of 6.3 (range 3.8-7.0, inter-quartile range 1.5) on a 7-point Likert scale.
The dashboard group and the nurse-education alone group were similar in terms of motivating patients to take some action to modify their vascular risk factors (e.g., making a lifestyle change or discussing the risk factor with their physician), their level of confidence that they could do something to improve their risk, and their assessment of how important it was to make such a change. No clinically or statistically significant differences in HbA1c, BMI or hypertensive medication use were observed over the 6 month study period, but the two intervention groups both demonstrated a reduction in blood pressure whereas blood pressure increased in the usual care group: mean change from baseline to 6-months in systolic blood pressure for dashboard was -0.8 mm Hg (+/-standard deviation, 18.1); nurse-education, -2.3 (+/-11.3); and usual care +8.4 (+/-19.1) (p-value for the dashboard versus usual care, p=0.13, for nurse-education versus usual care, p=0.06). Similarly, diastolic blood pressure improved in the two intervention groups: dashboard, -2.0 mm Hg (+/- 8.1); nurse-education, -3.4 (+/-7.3); versus usual care +3.5 (+/-13.2) (p-value for the dashboard versus usual care, p=0.13, for nurse-education versus usual care, p=0.04). Qualitative data suggested that dashboard patients were more likely to report that smoking was their most important vascular risk factor whereas nurse-education patients were more likely to report that hypertension was their most important vascular risk factor.
Management of risk factors for patients at high risk of vascular events is a challenge faced by patients as well as caregivers. The VA system has been engaged in a movement toward providing patient-centric primary care and health maintenance. Given these promising preliminary findings, it appears as if this electronic dashboard might be an effective tool that clinicians can deploy when seeking to work with veterans to engage in vascular risk factor modification.
None at this time.
DRA: Aging, Older Veterans' Health and Care, Other Conditions
Keywords: Disparities, Patient-centered Care, Quality assurance, improvement, QUERI Implementation, Risk adjustment, Stroke
MeSH Terms: none