HSR&D Citation Abstract
Search | Search by Center | Search by Source | Keywords in Title
Can Framingham Calculators Identify Persons at High-Risk for Stroke in a Large Healthcare Organization?
Cheng E, Williams LS, Vassar S, Ekundayo J, Bravata DM. Can Framingham Calculators Identify Persons at High-Risk for Stroke in a Large Healthcare Organization? Poster session presented at: American Heart Association / American Stroke Association International Stroke Conference; 2010 Feb 23; San Antonio, TX.
Typically, stroke prevention programs enroll persons who have already suffered a stroke. If persons at high-risk for a first time stroke could be identified in advance, they could be enrolled in prevention programs before a stroke occurs. However, the currently available calculators for predicting first-time stroke were developed in cohort studies where there is a concerted effort to ensure complete data collection, not on entire populations where data collection may be incomplete.
To determine whether the Framingham stroke calculator, derived from administrative data, can identify persons who will develop a stroke in the following year.
Our sample consisted of all 380 persons who were hospitalized for a first-time ischemic stroke in 2008 at 5 VA Medical Centers (VAMCs) in the Southwestern United States and a control group of 12,390 randomly selected persons receiving care within these VAMCs. For all persons, we obtained VA demographic, diagnostic, and physiologic administrative data available as of the end of 2007 to implement several variations of the Framingham stroke risk calculator. We calculated c-statistics in order to compare the predictive validity of this tool in this setting compared with the originally published studies on the Framingham cohort.
Persons with a 2008 stroke had a higher risk profile than controls on several components of the general cardiovascular Framingham calculator: older age, more likely to be male, higher systolic blood pressure, higher total cholesterol, and more likely to have diabetes (all p_0.05), resulting in a greater overall Framingham score (p_0.0001) among persons with a 2008 stroke. The c-statistic for the Framingham calculator was 0.60 for identifying persons who would develop a stroke in 2008. After excluding the 27% of cases and 20% of controls who did not have any 2007 blood pressure or cholesterol measurements, the c-statistic improved to 0.67. However, these figures were still below the c statistics of 0.77 to 0.84 reported in the original Framingham cohort studies.
Discussion: The VA administrative databases contain the information required to implement a stroke risk calculator on a widespread population. The calculator was able to identify persons at high risk for stroke, though as expected, its discriminating properties in a population setting are less than in the original cohort studies. Further work to better define administrative diagnostic and physiologic variables and to determine an appropriate threshold c statistic for population screening may enhance the performance and use of a stroke risk calculator. This research has received full or partial funding support from the American Heart Association, National Center.