The Effectiveness of Personalize Stroke Risk Communication
Hayden B. Bosworth PhD
Durham VA Medical Center, Durham, NC
Funding Period: March 2009 - September 2009
In 2005, over 17,000 patients were treated for stroke within the VA with a cost of almost $315 million. Prevention of stroke through reduction of established risk factors is an essential part of the VA Stroke QUERI strategic plan for the VA. In spite of this, in the Veterans Affairs, only 13% of patients with known CVD achieve target BP and cholesterol control. Combining risk factors into a composite measure of risk offers a better global assessment of individual risk and is recommended by the American Heart Association and American Stroke Association for prioritizing interventions. This practice is rarely done in routine clinical practice and its use as a tool to motivate patient behavior has not been tested. Current evidence from VA patients suggests that patients with hypertension do not adequately translate their risk factors into an accurate estimation of stroke risk. Improving the accuracy of stroke risk perceptions may be particularly important in motivating risk reduction in patients.
The objectives of this study are to: 1.) Assess the impact of personalized stroke risk communication to patients at risk for stroke on patient knowledge, beliefs, and preferences for risk reduction behaviors. 2.) Evaluate the impact of personalized risk communication on medication adherence and blood pressure. 3.) Explore the feasibility and obtain sample size estimates for a larger, investigator initiative research (IIR) application testing this tool.
A two-group randomized controlled trial testing a personalized risk communication intervention compared to an education-only control group was conducted. Eighty-nine patients were randomized and followed for 3months. Both groups received written and verbal patient education on stroke risk factors and prevention. Patients in the intervention arm also received personalized risk communication based on the Framingham stroke and coronary heart disease risk scores. A verbal and graphic presentation of their personal risk, risk relative to an age matched cohort, and their optimal or target risk based on optimal risk factor modification was presented. Outcomes measured immediately following the intervention and at 3months included: risk perception and worry; risk factor knowledge; decision preference and conflict; medication adherence; health behaviors; and blood pressure.
This group had very low understanding of numeric information, high perceived risk for stroke or heart attack, and a high proportion of medication non-adherence. Both the personalized and standard risk information improved patients' knowledge of vascular risk factors immediately after the information was presented, but this increase in knowledge was not sustained at 3months. Patients who received personalized risk communication were more likely to choose personal behavioral changes, such as diet or exercise, over medication management when compared to patients in the standard education group. In addition, patients in the personalized group had less decision conflict over this choice with an estimated effect size of 0.67. Patients who received personalized communication had greater reduction in perceived risk and worry compared to the standard education group, however this difference was small (estimated effect size 0.12). There was no evidence that personalized risk communication had any impact on medication adherence, exercise, or blood pressure at 3months. Self-reported medication adherence did not change appreciably in either group at 3months with non-adherence reported as 54% in the personalized group and 51% in the standard group.
This study suggests that personalized stroke and CHD risk communication influences patients' preferences for risk reduction and decision conflict. However, personalized risk communication alone had little to no lasting impact on patient knowledge, health behaviors, or blood pressure. Personalized risk communication may be an important component of multi-faceted self-management programs and support informed medical decision making and further research in this setting is warranted.
None at this time.
MeSH Terms: none