Harry Gregg Foundation

Background: The recommendations for dual antiplatelet (DAPT aspirin + clopidogrel) for secondary stroke prevention has evolved over time. Following the publication of CHANCE trial (07/2013), the AHA/ASA updated the DAPT recommendations from Class III harm (10/2010) for patient with noncardioembolic ischemic stroke, to Class IIb benefit ≥ risk (02/2014), and Class IIa benefit >> risk (03/2018) for a subgroup of patients with minor stroke (NIHSS≤3). Subsequent to the last guideline update, the POINT trial (05/2018) provided further support for the effectiveness of DAPT. Methods: We evaluated antiplatelet prescription patterns of 1,024,074 noncardioembolic ischemic stroke survivors (median age 65 years and 46% women) eligible for antiplatelet therapy (no contraindications) and discharged from the Get With The Guidelines-Stroke Hospitals between Q1 2011 and Q1 2019. Results: Baseline patient characteristics were similar within the four periods: pre-CHANCE (01/2011-07/2013), pre-2014 guideline update (08/2013-02/2014), pre-POINT/2018 guideline update (03/2014-05/2018), and post-POINT (06/2018-03/2019). Use of DAPT gradually increased from 16.7% in the pre-CHANCE period, to 19.4% pre-2014 guideline update, 23.3% pre-POINT/2018 guideline update, and 29.8% post-POINT period (p<0.001, Figure). Yet increase in DAPT use was observed over time for individuals with NIHSS≤3 (17.1%, 19.9%, 24.1%, and 31.4%, p<0.001) and those with NIHSS>3 (18.7%, 22.8%, 28.3%, and 28.3%, p<0.001). Conclusions: A sustained increase in DAPT use for secondary stroke prevention was observed after publication of pivotal trials and AHA guideline updates. While recommended for minor strokes or TIA only, such increase was also observed in ischemic stroke patients with NIHSS>3, where the risk-benefit ratio of DAPT remains to be established.

Introduction: Direct oral anticoagulants (DOACs) are increasingly used as alternatives to warfarin for secondary prevention in ischemic stroke patients with atrial fibrillation. Despite demonstrated efficacy in clinical trials, there are few real-world experiences of DOACs vs. warfarin in community practice. Methods: We analyzed ischemic stroke survivors with atrial fibrillation discharged from the Get With The Guidelines Stroke hospitals between 2011-2014 and linked to Medicare claims for longitudinal outcomes through 2015. A propensity score overlap weighting method was used to compare DOACs vs. warfarin. The primary outcomes were major adverse cardiovascular events (MACE) and home time, a patient-centered outcome reflecting desire of “being alive at home, without recurrent stroke, or being hospitalized for complications.” Results: Among 11,662 stroke survivors (median age 80), 4,041 (34.7%) were discharged on DOACs (dabigatran, rivaroxaban, or apixaban) and 7,621 on warfarin. Except for NIHSS (median 4 [IQR 1-9] vs. 5 [2-11]), baseline demographics, medical history, and clinical characteristics were similar between two cohorts. Compared with warfarin, patients discharged on DOACs were less likely to experience MACE (33.95% vs. 40.36% per year, adjusted hazard ratio 0.89, 99% CI 0.83-0.96) and had more days at home (mean 287 vs. 263 days during the first year post discharge, adjusted difference 15.6 days, 99% CI 9.0-22.1) ( Table ). Additionally, there were fewer deaths, all-cause readmissions, cardiovascular readmissions, hemorrhagic strokes, and bleeding hospitalizations in DOAC-treated patients, although no significant differences in fatal bleeding, ischemic stroke readmission, systemic embolism, pneumonia, or sepsis (two negative outcome controls) between the two cohorts. Conclusions: In ischemic stroke survivors with atrial fibrillation, DOACs were associated with improved long-term clinical outcomes compared with warfarin.

Background: Treatment with intravenous recombinant tissue plasminogen activator (rtPA) for acute ischemic stroke (AIS) in patients with a recent history of myocardial infarction (MI) within 3 months prior to AIS is controversial. Methods: We used probabilistic linkage of Medicare claim files with the Get With The Guidelines-Stroke registry to identify patients who were treated with rtPA for AIS between 2/2009 and 12/2015. Recent MI was determined from Medicare claims. Multivariable logistic regression models were used to evaluate the association of a recent history of MI with in-hospital clinical outcomes. Results: Among 40,396 patients age 65 years or older treated with rtPA for AIS from 1522 sites (mean age [SD] 81.0 [8.1] years; 58.0% women), 241 (0.6%) had recent MI within 3 months prior to AIS, of which 19.5% were STEMI. Patients with recent MI had more severe stroke than those without (median NIHSS [IQR]: 13.0 [7.0-20.0] vs. 11.0 [6.0-18.0]) and were more likely to be on antiplatelets prior to AIS (87.8% vs. 53.6%). The unadjusted rates of in-hospital mortality and any serious rtPA-related complications were 17.4% and 13.5% for those with recent MI, and 9.0% and 9.4% for those without, respectively. Hemopericardium was observed in 1 (0.4%) patient with recent MI. After risk adjustment, recent MI within 3 months was associated with a significant increased risk of mortality (adjusted odds ratio [OR] 1.61, 95%CI [1.11-2.34], P=0.012), whereas there were no statistically significant differences in any serious rtPA-related complications ( Table ). When stratified by the type of MI, recent ST-elevation MI was associated with higher risk of death and tPA complications, but the difference was not statistically significant in those with recent non ST-elevation MI. Conclusions: Among older ischemic stroke patients treated with rtPA, recent MI within 3 months was associated with an increased risk of mortality, and recent STEMI with mortality and serious tPA complications.

Background: Several cardiotoxic substances contribute to the development of heart failure (HF). The burden of comorbid substance use disorders (SUD) among patients with HF is under-characterized. Objectives: To describe the national burden of comorbid SUD (tobacco, alcohol, or drug use disorders) among hospitalized HF patients in the U.S. Methods: We used data from the 2014 National Inpatient Sample to calculate the proportion of hospitalizations for a primary HF admission with tobacco, alcohol, or drug use disorder diagnoses, accounting for demographic factors. Drug use disorder analysis was further sub-divided into specific illicit substance categories. Results: There were a total of 989,080 HF hospitalizations of which 35.3% (n=348,995) had a documented SUD. Tobacco use disorder (TUD) was most common (n= 327,220, 33.1%) followed by drug use disorder (DUD) (n=34,600, 3.5%) and alcohol use disorder (AUD) (n=34,285, 3.5%). Female sex was associated with less TUD (OR 0.59; 95% CI, 0.58-0.60), AUD (OR 0.23; 95% CI, 0.22-0.25) or DUD (OR 0.58; 95% CI 0.55-0.62). Tobacco, alcohol, cocaine, and opioid use disorders were highest among HF patients age 45 to 55, while cannabis and amphetamine use was highest in those <45 years. Native American race (versus White) was associated with increased risk of AUD (OR 1.67; 95% CI 1.27-2.20). Black race was associated with increased risk of AUD (OR 1.09; 95% CI 1.02-1.16) or DUD (OR 1.63; 95% CI 1.53-1.74). Medicaid insurance (versus Medicare) was associated with greater TUD (OR 1.27; 95% CI 1.23-1.32), AUD (OR 1.74; 95% CI 1.62-1.87), and DUD (OR 2.15; 95% CI 2.01-2.30). Decreasing quartiles of median household income were associated with increasing SUD. Conclusions: Comorbid SUD disproportionately affects certain HF populations, including men, younger age groups, lower SES patients, and race/ethnic minorities. Further research on interventions to improve prevention and treatment of SUD among hospitalized HF patients are needed given the high rates of SUD in this population. Systematically screening hospitalized HF patients for SUD may reveal opportunities for treatment and secondary prevention.

Background: Heart failure with reduced ejection fraction (HFrEF) has a significant impact on patients’ health status. A primary goal of treatment is to minimize patients’ symptoms, while maximizing their functional status and quality of life. The variability across outpatient practices, and the characteristics of those practices, in optimizing patients’ health status is unknown. Methods: In the multicenter CHAMP-HF registry, the association between practice characteristics and Kansas City Cardiomyopathy Questionnaire Overall Summary Scores (KCCQ-os) in 3,494 patients across 140 US centers was assessed with hierarchical logistic regression adjusting for 23 patient characteristics to predict excellent health status (KCCQ-os ≥75). A median odds ratio (MOR) was estimated to quantify the average relative difference a patient would exhibit excellent health status when being treated at one random practice versus another. Results: The mean (±standard deviation) KCCQ-os was 64.2±24, and 39% (n=1,380) had a KCCQ-os ≥75. In a fully-adjusted model, patients were more likely to have excellent health status if the practice was in a suburban (vs. urban) location (OR=1.47, p = 0.02), while those enrolled at a primary care (vs. cardiology) setting were less likely to have a KCCQ≥75 (OR=0.52, p = 0.02). The MOR across practices was 1.69 (95% CI=1.46, 1.90; p < 0.0001) indicating a median 69% higher odds of a statistically identical patient having excellent health status when treated at one random practice versus another. Conclusion: In this first large, contemporary assessment of health status in outpatients with HFrEF, we found marked variation by site in patients having excellent health status. Proposed efforts to use patient-reported outcomes as a quality measure may support improved attention to patients’ health status in clinical care.

Background: There is controversy over dietary protein’s effect on long-term chronic disease risks in overweight and obese patients with heart failure. The aim of this study was to compare the effects of two calorie-restricted diets differing in protein content on chronic disease risks (e.g., weight, adiposity, glycemic control, lipid profiles, and blood pressure). Methods: A total of 97 overweight and obese (mean body mass index [BMI], 37.0 ± 6.2 kg/m 2 ) patients with heart failure, aged 58.8 ± 9.7 years, 70% males, consumed two diets, each for 12 weeks, in a randomized controlled design. The diets were: (1) a high protein diet (30% protein, 40% carbohydrates, and 30% fat) and (2) a standard protein diet (15% protein, 55% carbohydrates, and 30% fat). Their effects on weight (BMI) visceral fat (i.e. waist circumference), glycemic control (i.e. glycosylated hemoglobin [HgbA1C]), lipid profiles (total cholesterol [TC], low-density lipoprotein [LDL], high-density lipoprotein [HDL], triglycerides [TG]), and blood pressure at the beginning and end of each dietary intervention were analyzed. Results: Both diets were equally effective in promoting weight loss and visceral fat loss and in improving TC, but the high protein diet decreased to a greater extent HgbA1C levels ( P < 0.001) when compared with the standard protein diet. The high protein diet improved significantly TG levels ( P < 0.001) and both systolic and diastolic blood pressures when compared with the standard protein diet ( P < 0.001 and P < 0.001, respectively, see figure). No differences were noted in LDL and HDL levels. Conclusion: Energy restricted diets facilitate weight loss and visceral fat loss and lower TC in overweight and obese patients with heart failure. However, the high protein diet promoted better improvements in TG levels, glycemic control, and blood pressure than the standard protein diet, and may be superior in reducing chronic disease risks and potentially slowing the progression of disease in this population.

Background: The benefits of intravenous tPA in acute ischemic stroke are time-dependent and hospitals participating in Target: Stroke have substantially improved the proportion of patients being treated with door-to-needle (DTN) times of ≤60 minutes. However, very few patients were treated with DTN times of ≤45 or ≤30 minutes during Target: Stroke Phase I (2010-2013). This study aims to assess whether there have been greater achievement of DTN times ≤45 and ≤30 minutes since the launch of Target: Stroke Phase II in Q2 2014. Methods: Target: Stroke Phase II identified and disseminated additional best practice strategies, provided updated clinical decision support tools, and set new hospital recognition goals. Rates of DTN times ≤45 and ≤30 minutes were compared prior to Target: Stroke (2003-2009), during Phase I (2010-2013), and during Phase II (Q2 2014-present) by weighted linear regression. Results: There were 131,166 intravenous tPA treated patients from 1397 GWTG-Stroke hospitals. Patient characteristics were similar during the study periods. Median DTN time declined from Phase I to Phase II: 66 minutes (IQR 51-87) to 54 minutes (IQR 41-71) and % of patients with DTN times ≤60 minutes increased from 42.0% to 62.7%, (P<0.0001). The % of patients with DTN times ≤45 minutes increased from Phase I to Phase II: 17.3% to 33.9%, absolute difference +16.6%, (P<0.0001). The % of patients with DTN times ≤30 minutes also increased from Phase I to Phase II: 4.1% to 10.3%, absolute difference +6.2%, (P<0.0001). The estimated annual rate of increase in patients with DTN times ≤45 minutes was 0.25% per year pre-Target Stroke, 3.5% per year during Phase I, and 7.7% per year during Phase II (P<0.0001) (Figure). Conclusions: There has been accelerating progress in achieving more rapid DTN times for tPA administration among GWTG-Stroke hospitals participating in Target: Stroke Phase II. Nevertheless, opportunities remain to further improve the timeliness of care and clinical outcomes.

Introduction: Timely reperfusion is performed less optimally in acute ischemic stroke (AIS) than in acute myocardial infarction (AMI). The degree to which hospital performance is correlated on emergent AMI and AIS care is unknown. Hypothesis: There would be a positive correlation between hospital performance on door-to-balloon time (D2B) for AMI and door-to-needle time (DTN) for AIS; and hospital performance on D2B would predict DTN even after controlling for patient and hospital differences. Methods: Prospective study of all hospitals participating in both Get With The Guidelines (GWTG)-Stroke and -Coronary Artery Disease from 2006-09 and treating ≥10 patients. We compared hospital-level DTN and D2B before and after risk adjustment using Spearman’s rank correlation coefficients and hierarchical linear regression modeling. We also correlated hospitals’ DTN and D2B data from 2013-14 using GWTG (DTN) and Hospital Compare (D2B). Results: There were 43 hospitals contributing data (1976 AIS and 59,823 AMI patients). Hospitals’ DTN times for AIS did not correlate with their D2B times for AMI (median DTN 85 min [IQR 77-99] vs. median D2B 72 min [IQR 62-81]; ρ=-0.09; p=0.55). There was no correlation between hospitals’ proportion of eligible patients treated within target time windows for AIS and AMI (median DTN<60 minutes: 21% [IQR 11-30]; median D2B<90 minutes: 68% [IQR 62-79]; ρ=-0.14; p=0.36). The lack of correlation between hospitals’ DTN and D2B times persisted after risk adjustment. From 2013-14, hospitals’ DTN performance in GWTG was not correlated with D2B performance in Hospital Compare (N=546 hospitals; see figure). Conclusions: We found no correlation between hospitals’ observed or risk-adjusted DTN and D2B times. Opportunities exist to improve hospitals’ performance of time-critical care processes for AIS and AMI in a coordinated rather than condition-specific manner.

Intro: Many patients are transferred to stroke centers for advanced stroke care, especially after IV tPA. We sought to determine differences in the baseline characteristics and outcomes between AIS cases presenting directly to stroke centers’ front doors vs. transfers-in from another regional acute care hospital. Methods: Using data from the national GWTG-Stroke registry, we analyzed 970,390 AIS cases (01/2010 - 03/14). Patients at hospitals with high transfer-in rates (>15%) were selected (284 hospitals, 303,739 patients). Due to large sample size, instead of p-values, standardized differences were reported. Multivariable model (MV) examined the association of transfer-in vs. front door with the primary and secondary outcomes, adjusting for patient and hospital characteristics including NIHSS. Results: High volume transfer-in hospitals admitted 31% of their patients via transfer. Transfer-in patients were younger, more often white and non-Hispanic. They had similar stroke risk factors except for hypertension and previous stroke/TIA which were less common. Transfer-in had worse initial NIHSS, more often had altered consciousness and language disturbance. Transfer-in patients had longer length of hospital stay, higher mRS at discharge, and were less often discharged home. In-hospital mortality was ∼ 3% higher in transfer-in as compared with front-door. Among tPA treated patients, sICH < 36hr was more common in transfer-in patients. On MV, transfer-in patients had overall worse outcomes as shown by the higher odds of in-hospital mortality, longer length of stay, and not able to ambulate independently at discharge (Table). Conclusion: Many hospitals receive high volumes of stroke patients via transfer. Because transfer-in patients have worse outcomes, these patients have the potential to negatively influence institutional outcomes rates. Transfer-in patients should be carefully accounted for in risk adjusted models of hospital outcomes.

Objectives: To assess the quality of ischemic stroke-related care based on American Heart Association (AHA) Get With The Guidelines (GWTG)-Stroke metrics for patients receiving dialysis compared to those with less severe or no CKD . Methods: The analysis is based on data from 232,236 Medicare-eligible fee-for-service patients admitted to 1581 GWTG-Stroke participating hospitals between January 2009 and December 2012. GFR was determined based on the MDRD study equation. Patients receiving dialysis were identified by ICD-9 diagnosis codes (V45.11 renal dialysis status, 585.6 end stage renal disease, V56.X encounter for dialysis and dialysis catheter care). Results: Compared to no CKD, dialysis patients received fewer guideline-based therapies (Table). Compared to other CKD stages, dialysis patients less often received VTE prophylaxis, timely treatment with IV-tPA, and defect free care. Treatment with early antithrombotics, discharge on antithrombotics, anticoagulation for atrial fibrillation, and statin use were lowest in patients with CKD stage 5. Conclusion: Among Medicare beneficiaries with acute ischemic stroke, guideline-adherent care is less frequent in patients on dialysis compared to those with less severe or no CKD. Additional work is needed to determine the reasons for these differences.