Introduction/Objective:
Recent studies show that targeted interventions on lifestyle factors such as weight management and diet can be successful in reducing ischemic stroke (ISC) and transient ischemic attack (TIA) rates. The objective of this study was to examine which subpopulations of patients at risk for secondary stroke presenting to a hospital with an ISC or TIA were more likely to receive interventions in a multi-hospital health system.
Methods:
Data from 26 hospitals participating in a multi-state healthcare system stroke registry, from January 2009 to December 2015, were used. Patients admitted with a diagnosis of ISC or TIA discharged to home were included. Patients on comfort care or those not discharged home were excluded. At-risk groups included patients with Body Mass Index (BMI) ≥ 25 and those prescribed medication for hypertension or high cholesterol in-hospital. Risk-related interventions included educational material given to patients during admission. Mixed effects logistic regression models with backward elimination were used to identify significant predictors of receiving the intervention from the following variables: year of discharge, age, insurance (private, Medicare, other/self-pay), BMI grouping, ambulation status, length of stay, stroke severity, and medical histories of family stroke, previous stroke or TIA, atrial fibrillation, coronary artery disease, heart failure, dyslipidemia, hypertension, and drug/alcohol abuse.
Results:
A total of 19,661 patients met the inclusion criteria. Of the 8,334 patients with a BMI ≥ 25, 57% (n=4,717) received weight management intervention. Of the 9,676 prescribed medication for hypertension, 55% (n=5,348) received information on antihypertensive diet. Of the 10,999 patients prescribed medication to lower cholesterol, 64% (n=7,088) received cholesterol lowering diet information. From 2009 to 2015, interventions increased for patients with a BMI ≥ 25 (40% to 66%), prescribed medication for hypertension (37% to 53%), and prescribed medication to lower cholesterol (39% to 67%). The mixed effects logistic regression models showed that all risk groups were significantly less likely to receive intervention if they had lower BMIs, were unable to ambulate versus able to ambulate alone, and had no family history of stroke. For those on medication for cholesterol, patients with Medicare were significantly less likely to receive the intervention compared to those on private insurance or other payment types (AOR=0.78, p<0.001).
Conclusions:
This large patient cohort demonstrates there are improvement opportunities for in-hospital secondary stroke prevention efforts. Over time, prevention efforts have increased for at-risk patients, but many are still not receiving it. The disparity in intervention rates suggest that a more targeted strategy to educate at-risk populations may need to be developed.