The purpose of this study is to evaluate the implementation of a locally tailored secondary stroke prevention program for Veterans with stroke using existing tools and programming. A stroke prevention program is a systematic delivery of tools, support materials, and programming for stroke survivors and for the clinical providers to systemically deliver to patients after an acute stroke event who at risk for a secondary stroke. It is also includes an organization process to ensure that those at risk are delivered appropriate prevention services. The program also is design to help patients self-manage their recovery from stroke.
Stroke affects at least 15,000 veterans each year, and this number will likely increase as the veteran population ages. According to the AHA, the prevalence of stroke is expected to double by 2020 with the increased proportion of older adults nationwide. Our preliminary QUERI work indicates that stroke risk factors are often undermanaged in VHS.
This proposed study of a stroke risk factor management program may benefit the VHS in several ways. First, it offers VHS a systematic program for reduction in stroke risk factors leading to better health for our veterans and a reduction in inpatient and outpatient rehabilitation and home health services for these events. Second, the VSPP takes into account the varied resources and services offered in VAMCs across the nation, allowing the program to be tailored both to a given facility and to the individual veteran's needs and readiness to change. Importantly, the program could allow all VA facilities to offer guideline-concurrent stroke risk reduction programs and therefore increase compliance with VA/DoD, AHA, and JCAHO stroke care guidelines and improve their quality of stroke care.
Objective 1: To conduct a formative evaluation of existing practices and to understand needs of both clinical providers and veteran patients with strokes and their families to help manage secondary stroke prevention and recovery.
Objective 2: To evaluate a locally tailored veteran secondary stroke prevention program at two VA medical facilities in comparison to patients with stroke at two matched VAMC control sites.
Comparison(s): Other regionally matched facilities to compare rate of compliance with the various measures during the course if the study at the intervention site.
Ages Eligible for Study: 18 Years and above, Genders Eligible for Study: Both
Veterans 18 years or older hospitalized with stroke or TIA at Indianapolis VAMC and Houston VAMC; willing to participate, access to telephone; speaks and understands English; no severe cognitive impairments; life expectancy of at last 6 mos; willingness to follow-up in VA outpatient care.
Severe aphasia or cognitive impairment; active alcohol or substance abuse; cannot or unwilling to participate; does not speak or understand English;life expectancy less than 6 mos; no access to telephone; no VA outpatient follow-up
We have incorporated the local preferences of VA providers involved with stroke care and designed a secondary stroke prevention program that incorporates both system and patient focused components including risk factor assessment, stroke discharge checklists, secondary stroke prevention Rx pad for providers, stroke/TIA self-management, stroke support group, and the American Stroke Association Sharegiver (Peer) Program.
VSPP Provider Based Intervention: Based on the results of Aim 1, our provider based intervention included posting secondary stroke prevention best practices for stroke discharge at the inpatient neurology workstations at both sites. We provided inservice presentations to clinical providers of stroke care on lifestyle counseling and motivational interviewing. We created and disseminated to the clinical providers Secondary Stroke Risk Factor Prescription pads tailored to each site listing the respective programs offered to veteran patients along with the contact information.
VSPP Patient Based Intervention: The two sites locally tailored the patient self-management portion of the intervention. In the Indianapolis program, the participants received 6 biweekly telephone sessions to deliver the stroke self-management program (intervention) or a placebo telephone call program (control) that mimicked the intervention schedule. Sessions targeted primary outcomes of stroke self-management, and secondary outcomes of self-efficacy. The self-management program followed a standardized manual (control call simply asked how the patient was doing), and interviews were conducted at baseline during hospital visit or within four weeks of discharge), 3 and 6 months with a booster call at 4.5 months. The Houston program was designed to deliver to its program over two sessions: one was delivered during the inpatient stay and the second during the outpatient neurology clinic visit. Similarly, they incorporated goal setting. Outcomes were collected on the same scheduled as Indianapolis.
To address veteran preferences, both sites implemented the American Stroke Association Sharegiver Program (peer support program). Veteran stroke survivors volunteered and were trained to deliver a peer support program to enrolled veteran patients who were hospitalized with stroke recently and randomized to the intervention.
We have completed enrollment at both sites (Indianapolis and Houston). As described in Aim 2, we recruited veterans with acute stroke or TIA and randomizing participants to receive either the stroke prevention program or the control group. At close of enrollment on December 31, 2011, we had enrolled a total of 189 participants (65 from Indianapolis and 124 from Houston). Fifteen subjects consented but refused baseline.
The final sample included 174 veteran patients with an acute stroke/TIA who were consented, completed a baseline interview, and were randomized to receive either a VSPP (n=87) or a control program (n=87) stratified by stroke/TIA and by receipt of inpatient rehabilitation to control for stroke severity. Among this sample, 126 (72.4%) completed a 3 month interview and 123 (70.7%) completed a 6 month interview.
Aim 2 Provider-Based Outcomes. We determine medication management for stroke prevention treatment at the time of stroke discharge and during the 6 months following the acute stroke event as a dichotomous yes/no variable using methodology developed in other VA studies and guided by a standard list of medications advocated in the stroke risk factor management program. The key stroke risk factors of interest were: 1) hypertension, 2) diabetes, 3) hypercholesterolemia. All of these outcomes used the affected subgroup as the denominator and those on treatment as the numerator. In all randomized patients, we determined medication compliance based on previously published methods in veterans evaluating the medication possession ratio comparing 6 months prior to acute stroke to 6 months after the stroke. We used PBM data to determine adherent stroke risk reduction treatment prescribed post-stroke. We examined CPRS records to see if risk factor screening was done during the stroke hospitalization or in the subsequent six months and lifestyle counseling/referral made to appropriate specialists and existing facility programs in physical activity, diet, weight management, smoking cessation during the study period. Data was scored dichotomously as yes/no.
Aim 2.Patient-Centered Outcomes: We assessed Stroke health-related quality of life with the SS-QOL. This 35-item instrument assessed 12 domains relevant to stroke patients including energy, mobility, work, upper extremity function, ADLs, family roles, social roles, vision, language, thinking, mood, and personality.
We have completed provider interviews including 19 in Houston and 26 in Indianapolis. The majority of providers endorsed the idea that they have a role in secondary stroke risk factor management (81% in Indianapolis and 100% in Houston). Many providers worked with others in the VA facility or referred patients to other services or programs to assist in risk factor management (Indianapolis, 52% and Houston, 68%). Main themes that emerged from the interviews included the need for the following programs and strategies: stroke support group; peer program, provider stroke prevention education, standardized provider check list or discharge check list; materials for patient education and administration support.
We have completed 6 focus groups comprised of 21 veterans with either stroke or TIA. The groups were audiotaped, transcribed, and coded for emerging themes. Several key themes emerged from the data including: patient readiness to change after stroke event, social influences on patient risk factor modification, multiple risk factor management, and patient needs and ideal programs for stroke patients. In addition, participants were asked to evaluate existing stroke risk factor programs including the ASA Peer Program, Houston STOP risk factor clinic, and Myhealthevet website.
We have completed 4 focus groups of enrolled veterans who have completed the study or have served as a trained volunteer as a stroke peer visitor as part of the intervention. We are currently qualitatively coding the data.
Aim 2 Provider Outcomes
To evaluate the delivery of provider based secondary stroke prevention practices, we extracted 617 medical records which met our criteria from 2008 and 2009 from four VAMCs using a standardized abstraction form. Of those, we completed full abstractions for 392 medical records. The remaining 237 records were deemed not valid. Prior to the VistaWeb extraction, we established excellent inter-rater reliability among the trained medical chart abstractors (79.1%, almost perfect, Kappa > 0.81-1.00 and 58% of ICCs were equal to or greater than 0.80) indicating most of the total variation came from differences between subjects and not raters.
Aim 2. Provider Best Practices. Within the 4 sites, providers were adherent to best practices across the cohorts for medication prescriptions and lab tests ordered to manage stroke secondary risk factors at discharge. The most significant changes in provider practices occurred at the two VAMCs preparing for stroke center certifications (Houston and Miami VAMCs). The Joint Commission performance indicator of providing stroke education at discharge on specific stroke risk factors significantly increased. The rates in Indianapolis and West Haven decreased or remained stable over time. For smoking cessation counseling, the rates increased within all 4 sites although not statistically significant. Moreover, the rates of documented successful smoking cessation or reduction in smoking increased for 3 of the 4 VAMCS although not statistically significant. For blood pressure management during the 6 months after discharge for stroke, all 4 VAMCs documented nearly 100% of the stroke survivors had a blood pressure at goal in 2009.
Aim 2a. Patient Compliance with Provider Prescriptions. Medication possession ratios based upon patient medication refills were calculated to evaluate medication compliance for the randomized subjects using PBM data pre (6 months prior to the acute stroke/TIA event) and post (6 months after the event). Based upon the literature standard of 80% compliance, we used this rate to dichotomize and model the data using logistics regression with terms for site, strata (stroke, stroke with rehabilitation, TIA), randomization group, time and the group by time interaction. Compliance was evaluated for three clinical areas: diabetes, hypercholesterolemia, and hypertension. A random subject effect was included to incorporate the correlation of repeated measurements into the model.
In the intervention group, the odds of compliance to diabetes meds post stroke were significantly larger than the odds of compliance prior to the stroke (odds ratio=3.45 (1.08, 10.96) p <0.04). In other words, in the intervention group the increase in the odds of compliance was statistically significant. The control group saw a decrease in the likelihood of compliance from pre to post stroke hospitalization (odds ratio =0.51 (0.10, 2.70), p<.42). The comparison of the change in logs odds in the intervention group to the change in log odds in the control group was marginally significant (odds ratio=6.74 (0.88, 51.31), p=0.06) which indicates an intervention effect on compliance of DM drug. Both the intervention (odds ratio=5.98 (2.81, 12.76), p<.0001) and control groups (odds ratio=3.83 (1.83, 8.01), p<.0004) had significant increases in the odds of compliance to statin medications; however, the comparison of changes in log odds of compliance between these two groups showed that the increases were not significantly different. For compliance to hypertension medications, the intervention group showed significantly greater odds of compliance post intervention than pre intervention (odds ratio=3.68 (1.81, 7.48), p<.0004). The control group showed no difference in compliance rates between these two time points. The comparison of the increase in log odds between the intervention and control groups was not statistically significant (odds ratio=2.34 (0.86, 6.40), p<.10).
Aim 2 Patient Outcomes: Stroke Specific Quality of Life (SSQoL)
Baseline SSQoL overall and subscales scores were compared between study groups by 2 sample t-test. Scores at 3 and 6 months were compared between groups by using a mixed model with terms for site, strata, baseline score, group, month and the group by month interaction. A random subject effect was used to correlate observations from the same subject. Ran subgroup analyses with subgroups defined as NIH Stroke Scale (NIHSS)<5 (n=138) (less stroke severity) and 5<=NIHSS<=13 (n=36) (i.e., stroke severity). A Satterthwaite approximation was used for the degrees of freedom for the test of group differences at each month.
There were no significant differences between the intervention and control groups at 3 and 6 months for the SSQoL overall score or any of the subscales scores. From the subgroup analysis based upon stroke severity scores (NIHSS), at 3 months, mean SSQoL self care was significantly higher for the intervention group than the control group amongst patients with NIHSS 5 (greater stroke severity). At 6 months, there was no difference in mean SSQoL self care between intervention and control. At 6 months, mean SSQoL work was significantly higher for the control group than the intervention group amongst patients with NIHSS<5 (less stroke severity). In addition, there was a trend in self-reporting more perceived energy (p<.08) and better overall stroke specific, quality of life (p<.10) at three months compared to control group participants.
We are currently conducting secondary analyses on patient self-reported self-management behaviors, well-being and functioning (e.g, pain, depression, fear of movement), and knowledge of stroke risk factors and stroke symptoms.
The nurse education template for hospitalized patients has been revised for stroke patients as part of this study and the template was adapted for the documentation of nurse education for hospitalized patients with other medical conditions (e.g., diabetes) at the Roudebush VAMC.
The methods used for this implementation study was used as an example in materials for the national VA training program for Implementation Research, July 19-20, 2010, Denver, CO.
A systematic secondary stroke prevention program for veterans with stroke/TIA may impact stroke self care and overall stroke specific quality of life for veteran patients with stroke/TIA, and improve medication compliance for veteran patients with diabetes who experience an acute stroke.
This study demonstrates that it is feasible to develop and deliver a locally tailored intervention for secondary stroke prevention which incorporates both patients' and providers' preferences. Specifically, we promoted best practices by providers and the adoption of secondary stroke risk prevention by veteran patients with an acute stroke/TIA. We observed over a threefold increase in diabetes medication compliance among stroke survivors who were diabetic and assigned to the intervention group and we observed a significant decrease in diabetic medication compliance among those assigned to the control group. For hypertensive stroke survivors assigned to the intervention, we observed a significant increase in medication compliance after the intervention while those assigned to the control group remained the same. For statins, we observed significant increases within both groups among stroke survivors with elevated cholesterol. This overall increase may have reflected a recent stain performance measure that was implemented in VHA.
The addition of the provider practice comparisons at the control sites (Miami and West Haven) allowed us to examine temporal trends in VHA during the study period. We observed both Houston and Miami sites significantly increase their documentation of stroke education delivered to hospitalized patients with stroke at discharge across time. Both sites were working on stroke center certifications during the study and had initiated internal stroke quality improvement programs.
Finally, we report trends in better stroke specific quality of life at 3 months after the intervention among stroke survivors with a greater stroke severity based upon the NIH Stroke scale. This suggests that a patient focused intervention that targets patients experiencing a more severe stroke in terms of functioning may potentially impact health related quality of life.
This was a bundled multi-level intervention and therefore, it was difficult to attribute the effect of each component on the outcomes.
- Damush TM, Myers L, Anderson JA, Yu Z, Ofner S, Nicholas G, Kimmel B, Schmid AA, Kent T, Williams LS. The effect of a locally adapted, secondary stroke risk factor self-management program on medication adherence among veterans with stroke/TIA. Translational behavioral medicine. 2016 Sep 1; 6(3):457-68.
- Schmid AA, Andersen J, Kent T, Williams LS, Damush TM. Using intervention mapping to develop and adapt a secondary stroke prevention program in Veterans Health Administration medical centers. Implementation science : IS. 2010 Dec 15; 5:97.
- Damush TM, Jackson GL, Powers BJ, Bosworth HB, Cheng E, Anderson J, Guihan M, LaVela S, Rajan S, Plue L. Implementing evidence-based patient self-management programs in the Veterans Health Administration: perspectives on delivery system design considerations. Journal of general internal medicine. 2010 Jan 1; 25 Suppl 1:68-71.
- Damush TM, Anderson JA, Yu Z, Ofner S, Myers L, Schmid AA, Williams LS. Implementation of a Secondary Stroke Prevention Program: Effect on Stroke Specific Quality of Life. Poster session presented at: Society of Behavioral Medicine Annual Meeting and Scientific Sessions; 2013 Mar 22; San Francisco, CA.
- Damush TM, Myers L, Anderson JA, Yu Z, Ofner S, Schmid AA, Williams LS. Implementation of a Secondary Stroke Prevention Program: Effect on Medication Adherence. Poster session presented at: Society of Behavioral Medicine Annual Meeting and Scientific Sessions; 2013 Mar 22; San Francisco, CA.
- Damush TM, Anderson JA, Yu Z, Ofner S, Myers L, Schmid AA, Williams LS. Secondary Stroke Prevention Program: Effect on Stroke Specific Quality of Life. Poster session presented at: American Heart Association / American Stroke Association International Stroke Conference; 2013 Feb 6; Honolulu, HI.
- Damush TM, Myers L, Anderson JA, Yu Z, Ofner S, Schmid AA, Williams LS. Implementation of a Secondary Stroke Prevention Program: Effect on Medication Adherence. Poster session presented at: American Heart Association / American Stroke Association International Stroke Conference; 2013 Feb 6; Honolulu, HI.
- Van Puymbroeck M, Schmid AA, Carter S, Koester W, Docherty C. Modified Rehabilitation has Long-Term Impact on Walking for Individuals with TIA. Poster session presented at: American Heart Association / American Stroke Association International Stroke Conference; 2013 Feb 6; Honolulu, HI.
- Damush TM, Atkins D, Jackson GL. Implementation Science Research Opportunities in the Veterans Health Administration. Panel presentation at the Annual Meeting of Society of Behavioral Medicine. Paper presented at: Society of Behavioral Medicine Annual Meeting and Scientific Sessions; 2011 Apr 29; Washington, DC.
- Damush TM, Guccione A, Kerns R, Natarajan S. Career Opportunities in VHA. Paper presented at: Society of Behavioral Medicine Annual Meeting and Scientific Sessions; 2011 Apr 28; Washington, DC.
- Damush TM, Nicholas G, Kimmel B, Kent T, Williams LS. Implementation of the American Stroke Association Sharegiver Program (Peer Support). Poster session presented at: American Heart Association / American Stroke Association International Stroke Conference; 2011 Feb 9; Los Angeles, CA.
- Anderson JA, Willson P, Rittman M, Damush T. Chronic disease self-management: Establishing a framework for research, education, and practice. Poster session presented at: Sigma Theta Tau International Honor Society of Nursing International Nursing Research Congress; 2010 Jul 12; Orlando, FL.
- Damush TM, Williams LS, Schmid AA. Adapting Secondary Stroke Prevention Programs to Local Resources, Structures, and Front Line Inputs. Poster session presented at: National Institutes of Health Conference on the Science of Dissemination and Implementation: Methods and Measurement; 2010 Mar 15; Bethesda, MD.
- Damush TM, Williams LS, Schmid AA. Adapting Secondary Stroke Prevention Programs to Local Resources, Structures, and Front Line Inputs. Poster session presented at: VA QUERI National Meeting; 2009 Dec 10; Washington, DC.
- Damush T, Plue L, Schmid A, Kent TA, Anderson JA, Murphy C, Kimmel B. Adapting Secondary Stroke Prevention Programs to Local Resource Structures and Front Line Input. Paper presented at: American Heart Association / American Stroke Association International Stroke Conference; 2009 Feb 17; San Diego, CA.
- Damush T, Anderson JA, Bosworth H. Implementation of Patient Self-Management Programs among Veterans. Poster session presented at: VA QUERI National Meeting; 2008 Dec 10; Phoenix, AZ.