Search

Background: Patients with ischemic stroke have worse outcome if there is concomitant coronavirus 2019 (COVID-19) infection. We hypothesize that patients with intracerebral hemorrhage (ICH) during the pandemic era had worse outcomes than ICH patients during the pre-pandemic period, and that concomitant COVID-19 in ICH patients is associated with worse outcome. Methods: We analyzed data from Get With The Guidelines® Stroke. We implemented a two-stage design: first, we compared outcomes for ICH patients pre-pandemic (admitted between March 2019 and February 2020) with outcomes for ICH patients admitted during the pandemic (from March 2020 to February 2021); second, we compared outcomes of ICH patients with and without COVID-19 infection admitted during the pandemic (from March 2020 to February 2021). We used multivariable logistic regression adjusted by age, sex, race/ethnicity, and clinical covariates for in-hospital mortality and poor functional outcome using the modified Rankin Scale (mRS) at discharge (mRS >= 4) and a for length of stay (LOS). Results: The first stage included 62,743 pre-pandemic and 64,681 intra-pandemic ICH cases 64,681. ICH patients during the pandemic had worse outcomes, including higher modified rankin scale (mRS) (OR 1.10, CI 1.06-1.14) and higher mortality (OR 1.04, CI 1.01-1.07). The second stage included 60,091 COVID-19-negative and 1,326 COVID-19-positive ICH patients. Patients with concomitant COVID-19 infection had worse outcomes, including higher LOS (RR 1.32, CI 1.25-1.39), high mRS (OR 1.68, CI 1.40-2.01), and increased mortality (OR 1.50, CI 1.33-1.71). (Table). Conclusion: When comparing pre- and intra-pandemic cases of ICH, critically ill stroke patients admitted during the pandemic had worse outcomes. When comparing COVID-19 positive and negative ICH cases that happened during the pandemic, those with the infection had worse clinical outcomes. COVID-19 infection is an independent risk factor for poor outcomes in ICH.

Background: In 2015, endovascular therapy (EVT) for large vessel occlusions became standard of care for acute ischemic stroke. Lower utilization of IV alteplase has been reported in women, but whether sex differences in EVT use in the United States exists has not been established. Methods: We identified all acute ischemic stroke discharges from Get With The Guidelines–Stroke hospitals between 2012 and 2019 who were potentially eligible for EVT, based on National Institutes of Health Stroke Scale score ≥6 and arrival Results: Of 302 965 patients potentially eligible for EVT, 42 422 (14%) received EVT. Before 2015, EVT treatment rates were 5.3% in women and 6.6% in men. From 2015 to 2019, treatment rates increased in both sexes to 16.7% in women and 18.5% in men. The adjusted odds ratio for EVT in women compared with men was 0.93 (95% CI, 0.87–0.99) before 2015, and 0.98 (95% CI, 0.96–1.01) after 2015. There were no significant sex differences in outcomes except that after 2015, women were less able to ambulate at discharge (adjusted odds ratio, 0.95 [95% CI, 0.95–0.99]) and had lower in-hospital mortality (adjusted odds ratio, 0.93 [95% CI, 0.88–0.99]). Conclusions: EVT utilization has increased dramatically in both women and men since EVT approval in 2015. Following statistical adjustment, women were less likely to receive EVT initially, but after 2015, women were as likely as men to receive EVT. After EVT, women were more likely to be disabled at discharge but less likely to experience in-hospital death compared with men.

The Michigan Stroke Transitions Trial (MISTT) tested whether in-home social work case management (SWCM) or SWCM combined with access to a website providing stoke-related information improved outcomes relative to usual care for patients discharged home post-stroke and their caregivers. The aims of this secondary analysis are 1) to describe the actual support social work case managers (SWCM) provided to MISTT participants and 2) use select case studies to illustrate the relationship between SWCM and quantitative patient and caregiver outcomes. Data for the study were derived from SWCM case notes on 157 patients and their caregivers who received the MISTT intervention. Case notes were coded in two steps with a subset of cases coded by two researchers and reviewed for interrater reliability in each step. The first round of coding was guided by primary SWCM intervention goals. The second round of coding identified SWCM sub-themes within each primary goal. Key themes indicate SWCMs aided with understanding the post-hospitalization period, helped patients navigate a range of systems and services, identified needs and supported patient goals, provided psychosocial support, and centered support on stroke recovery and prevention. Case studies illustrate ways in which SWCM were key supports during the transition period, but that support does not cleanly align with quantitative findings from patient-reported outcomes. This study aligns with a growing body of work documenting the complexity of transitions of care and has implications for how we support patients and caregivers as they move from inpatient to outpatient care and measure outcomes.

Background: Cerebral infarction involving the insula has been associated with decreased survival following stroke. We hypothesized that infarct volume may reduce this association. Methods: The subjects were acute stroke patients who had consented to 2-year follow-up after stroke as part of the Michigan Acute Stroke Care Overview and Treatment Surveillance System registry. One hundred and eleven subjects exhibited areas of acute ischemic infarction on neuroimaging studies, 25 of whom had infarction involving the insula. Cox proportional hazard ratios (HR) were calculated to determine the association between mortality and acute infarction involving the insula, infarct volume, and other factors known to affect survival after stroke. Results: In unadjusted analysis, subjects with insula infarction had a nonsignificant twofold increase in 1-year mortality (HR = 2.1, 95% CI 0.6–7.0; p = 0.25). When adjusted for infarct volume, however, the HR for insula infarction was reduced to the null value (HR = 1.0, 95% CI 0.2–4.1; p = 1.00), indicating that the effect of insula infarction was entirely confounded by infarct volume. Conclusions: Insula infarction was associated with a nonsignificant twofold increase in mortality after stroke; however, this association was completely eliminated after adjusting for infarct volume. Infarct volume thus should be considered in future studies of insula infarction and mortality.

Introduction: Mortality after stroke is often hastened by forgoing life-sustaining measures. We sought to determine the patient and hospital characteristics associated with early comfort measures only (CMO) after ischemic stroke (IS), intracranial hemorrhage (ICH), and subarachnoid hemorrhage (SAH) in the Get With the Guidelines-Stroke registry. Methods: We identified patients with IS, ICH, or SAH between November 5, 2009 and September 30, 2013 who: 1) had known early CMO status; 2) were not transferred in from or out to another acute care facility; and 3) from sites with > 10 patients during the study period. Using multivariable logistic regression, we assessed patient and hospital factors associated with early CMO use by stroke type. Early CMO was defined as occurring on hospitals day 1 or 2. Results: Among 963,525 patients from 1,675 hospitals, 54,794 (5.6%) received early CMO (IS: 3.0%; ICH: 19.4%; SAH: 13.1%). There was a decrease in early CMO use (p Conclusions: Nationwide, early CMO is utilized in over 5% of stroke patients being more common in ICH and SAH than IS. Different patient and hospital characteristics influence CMO use among stroke types.

Introduction: Emergency dispatchers are usually the first contact point for stroke patients accessing the healthcare system. Dispatcher stroke recognition is associated with better paramedic stroke recognition and faster emergency medical services (EMS) response time, but stroke is often unrecognized during the initial call. We conducted a systematic review to quantify the accuracy of emergency dispatcher stroke recognition when employing stroke screening tools. Methods: We conducted a comprehensive search Medline, EMBASE, CINAHL, and Cochrane databases to identify studies of dispatcher stroke recognition accuracy using a stroke screening tools such as the Medical Priority Dispatch System (MDPS) or Face Arm Speech Time (FAST) mnemonic. Studies were included if they reported dispatcher sensitivity or positive predictive value (PVP) using hospital stroke/TIA diagnosis as the gold standard. Study quality was assessed using QUADAS 2. Two independent reviewers determined study eligibility, assessed quality, and abstracted data. Results: We identified 1450 abstracts; seven cohort studies met inclusion criteria. Stroke screening tools included MPDS (n=4), FAST (n=2), and a novel screening algorithm (n=1). The sensitivity of dispatcher stroke recognition (5 studies, range 41-83%) and PPV (7 studies, range 42% -68%) were variable. Specificity was high (3 studies, 95-99%). No screening tool demonstrated superiority. Included studies had moderate risk of bias and dramatically variable stroke prevalence (range 3-68%). Conclusion: Published reports of screening tools utilized by emergency dispatchers demonstrate suboptimal sensitivity and PPV. More research is needed to identify the causes of poor dispatcher stroke recognition and to improve stroke screening tools.

Background: Depression following an acute ischemic stroke (AIS) is associated with poor functional recovery, less social activity, and mortality. Patterns of health service use in the 90 days after an AIS have not been described by depression severity. Methods: Data from Get With The Guidelines-Stroke and a prospective cohort of AIS patients enrolled in the AVAIL (Adherence eValuation After Ischemic Stroke Longitudinal) study were linked with Medicare claims. AVAIL patients depressive symptoms were measured by the PHQ-8 90 days after hospital discharge by phone and categorized as no/minimal depression (0-4), mild (5-9), and moderate-severe (≥10). The number of transitions was defined as the number of acute and post-acute inpatient and outpatient services in the claims data within 90 days of the index admission. Home-time was the number of days alive and out of inpatient care. Differences in clinical characteristics and service use were evaluated between depression severity groups. Results: Of 538 AIS patients examined, 63% had no/minimal depression, 24% mild and 14% had moderate-severe depression 90-days post-stroke. Patients experienced up to 6 care transitions in 90-days; the 10 most common care patterns representing 80% of this sample (Figure). Across categories of increasing depression severity (no/minimal, mild, moderate-severe), use of services increased from no/minimal to those with depression (% with ≥1 service after discharge: 71%, 87%, 85%; p=0.002), and home time decreased (mean±SD: 82±16, 75±23, 77±23; p=0.006), although rehospitalization within 90 days was 19% for each group (p=0.984). Conclusion: This study demonstrates an association between post-stroke depression and worse outcomes following stroke. Research is needed to determine whether targeted care coordination interventions can improve outcomes among stroke patients with depression.

Background: While shorter door-to-needle (DTN) times with intravenous tPA is associated with better outcomes, prior studies have demonstrated that older patients are less likely to be treated within 60 minutes. Whether stroke quality improvement programs can impact care to a similar degree for older and younger patients has not been well studied. This study aims to assess the improvements in DTN times before and after the launch of Target: Stroke Phase I in 2010 among different age groups. Methods: Target: Stroke identified and disseminated 10 best practice strategies, provided clinical decision support tools, and set hospital recognition goals. Rates of DTN times ≤60 minutes and cumulative improvements pre- 2003-2009 were compared to post-Target Stroke 2010-2013 for patients age Results: There were 71,169 intravenous tPA treated patients (27,303 pre-; 43,866 post-Target Stroke) from 1030 GWTG-Stroke participating hospitals. Patients were median age 72 (IRQ 60-82). Overall, patients with DTN times ≤ 60 minutes increased from 26.5% (95% CI 26.0-27.1%) pre-intervention to 41.3% (95% CI 40.8-41.7%) post-intervention (P Conclusions: The implementation of Target: Stroke was associated with significant improvements in DTN times across all age groups, but differences by age, while narrowed, still persisted post-Target: Stroke. Despite overall progress, additional efforts will be needed to address these age related differences in timely stroke care.

Introduction: The concept of frailty as a state of decreased reserve and cumulative decline in multiple physiologic systems resulting in poor outcomes is well accepted in geriatrics. However, the impact of frailty on outcomes after ischemic stroke is less well understood. Our objectives were to identify a frailty score based on age-related deficits and to determine its association with stroke outcomes. Methods: Ischemic strokes within a biracial population in the Greater Cincinnati area in 2005 were identified. Research nurses extracted information from the medical charts including 35 age-related deficits that included comorbidities and admission lab values. A frailty score was identified using exploratory factor analysis to eliminate deficits with poor fit, followed by a two-parameter logistic latent trait model. Logistic regression was then used to evaluate association between frailty score (tertiles) and in-hospital mortality (death or discharged to hospice), dependent at discharge (mRS ≥ 2), discharge home and 1 year mortality. Models adjusted for age, race, sex, NIHSS, and pre-stroke mRS. Results: A total of 2,092 first event ischemic cases were identified in 2005. The median age was 72 years, 23% black, 56% female, median NIHSS 4 (IQR 2, 8). Of the 35 deficits, 15 loaded on the frailty factor: assisted living, aid to walk, dementia, depression, congestive heart failure, cardiac pacemaker, carotid artery disease, coronary artery disease, atrial fibrillation, impairment from prior stroke, hypertension, diabetes, elevated cholesterol, peripheral vascular disease, and abnormal hemoglobin. Higher frailty score was associated with increased in-hospital mortality, poor discharge mRS, and 1 year mortality (Table 1). Discussion: We created a frailty score based on 15 age-related deficits and showed strong association with poor outcomes. This frailty score could be useful for clinical care, as well as research, to identify age-related physiologic decline.

Background: Get With The Guidelines (GWTG)-Stroke is a national, hospital-based quality improvement program developed by the American Heart Association. While studies have shown a beneficial effect of hospital participation in GWTG-Stroke upon processes of care, whether there are associated improvements in clinical outcomes has not been previously investigated. Methods: From among all acute care US hospitals, we matched 366 hospitals that joined the GWTG-Stroke program between April 2004 and December 2007, with 366 hospitals that did not. Matching was based on ischemic stroke case volume, calendar year, baseline hospital post-stroke 1-year all-cause mortality rates, teaching status, and geographic region. Outcomes of all acute ischemic stroke (AIS) patients admitted to the study hospitals were abstracted from the CMS administrative claims database (65 years and older). Outcomes at matched hospitals were compared in the PRE-GWTG-Stroke period (-540 to -181 days before program launch), RUN-UP period (-180- to -1 day), EARLY period (0 to 180 days) and SUSTAINED period (181 to 540 days). Additional analysis was performed of the entire BEFORE (-540 to -1 days) and AFTER periods (0 to 540 days). The main analytical approach was stratified Cox proportional hazard modeling, with matched site ID at stratum. We adjusted for patient characteristics (age, gender, race, medical history) and hospital characteristics (rural vs. urban, # beds, annual IS discharges.) Results: The study analyzed 88,584 AIS admissions at the 366 GWTG-Stroke hospitals and 85,401 admissions at the 366 matched non-GWTG-Stroke hospitals. In adjusted analysis comparing BEFORE and AFTER periods, GWTG-Stroke hospitals achieved reduced 30 day mortality (30M - HR 0.911, p Conclusions: Hospitals joining the GWTG-Stroke quality improvement program between 2004-2008 achieved significantly greater improvement in stroke patient outcomes than matched hospitals not joining the program, with lower all-cause mortality at 30 days and 1 year and higher rates of discharge directly to home.

Characteristics and Outcomes of Stroke Patients Transferred to Hospitals Participating in the Michigan Coverdell Acute Stroke Registry, 2009-2011 Background: As stroke systems of care are evolving the number of stroke patients transferred between hospitals is increasing. Our objectives were to describe the characteristics of acute stroke patients who were transferred to hospitals participating in the Michigan Coverdell Acute Stroke Registry, and to determine the independent association between transfer status (TS) and in-hospital mortality (IHM) and in-hospital complications (IHC). Methods: From 2009-2011, 30934 acute ischemic (IS) and hemorrhagic stroke (HS) patients were admitted to 35 registry hospitals. Patients with an in-hospital stroke, TIA, or unknown arrival mode were excluded (N= 14,732). Independent factors associated with TS and predictors of IHC (defined as deep vein thrombosis, pneumonia, and/or UTI) were identified using multivariable logistic regression models. Results: The mean age of the 16202 admissions was 69.2 years, 51% were female, 68% were white, 83% had an IS, 7.4% died in-hospital, and 13.8% had an IHC. Overall, 19% (N= 3091) were transferred to a registry hospital. The transfer rate increased from 2009-2011 (16.9% -21.1%), and was higher in HS vs. IS patients (37.6% vs. 15.3%). Significant predictors of TS were year, age, gender, race, stroke type, pre-stroke ambulatory status, nursing home residence, and medical history of diabetes or prior stroke. Length of stay (LOS) was longer for transferred patients vs. non-transferred (7.9 vs. 5.3 days, p Conclusions: The frequency of hospital transfers increased markedly in this registry. Transferred patients experience higher rates of IHM, IHC and longer LOS. Further studies are needed to understand the relationship between TS and outcomes and the implication for improved clinical care and reducing poor outcomes in this higher risk group.

Introduction: A substantial evidence base supports the positive benefits of post-stroke rehabilitation. The objective of this study was to evaluate the impact of inpatient rehabilitation on 2-year healthcare costs in Ontario, Canada. Methods: Acute patient data was drawn from the 2004 and 2008 Ontario Stroke Audits for 3439 patients with ischemic stroke discharged alive from an acute care hospital. Patients were assigned a propensity score using 15 demographic and clinical variables. Patients admitted to inpatient rehabilitation (Rehab) were then matched one-to-one with a patient not admitted to inpatient rehabilitation (No-Rehab) using propensity score and modified Rankin Scale (mRS) score. The cohort was stratified by mRS into three groups (mRS 0-2,3,4-5) and comparisons were drawn between Rehab and No-Rehab patients for 2-year mortality and government-billed healthcare costs (hospital, home care, and drug benefit costs). Results: No difference in mortality was noted between Rehab and No-Rehab patients with discharge mRS scores of 0-2 (p=0.39), but Rehab patients cost an average of $33,056 more over the 2 years (p Conclusions: Findings suggest that disability level plays an important role in the cost-effectiveness of inpatient rehabilitation. Among patients with milder disability, inpatient rehabilitation is significantly more costly with no demonstrated benefit in 2-year mortality. However among patients with moderate to severe disability, inpatient rehabilitation significantly reduces mortality rates and may actually reduce overall healthcare spending up to 2-years post stroke.

Background: Multiple studies have shown that women have poorer quality of life (QOL) than men after stroke. We hypothesized that women would have lower QOL scores measured with the EuroQOL-5D (EQ5D) up to 1 year after ischemic stroke than men in the AVAIL cohort, a large, nationally representative, observational registry. Methods: After excluding proxy responses (n= 190), lost-to-follow-up (n= 62), and missing EQ5D scores (n=77), this analysis included 2,450 patients with ischemic stroke or TIA enrolled at 99 sites participating in the GWTG program. Baseline and 3 month sociodemographic and clinical data were compared by gender. EQ5D, a 5-item measure that assesses self-reported mobility, self care, usual activities, pain/discomfort, and anxiety/depression, was obtained at 3 and 12 months via telephone by centrally located and trained interviewers. A mixed effect multivariable model of EQ5D (primary outcome at record level) at 3 and 12 months was performed with gender and other relevant covariates as fixed effects and patient ID as the random effect. Results: Women were older than men (mean 65.8 versus 64.2 years). At three months, women were less likely to have household income meet needs (68.4% vs. 72.5%), to be married (47.7% vs. 69.3%), working (31.8% vs. 47.8%), or ambulating at discharge (66.9% vs. 71.7%) than men. Women were more likely to be living alone (30.1% vs 18.3%), and were discharged on more medications (mean 7.2 vs. 6.6; p=0.0006), but men were more likely to have > 4 risk factors (15.4% vs. 11.1%; p Conclusions: In this cohort, QOL was independently associated with socioeconomic status, disability, marital status, number of medications, and comorbidities, but not gender.

Introduction: Delays in post-stroke rehabilitation can negatively affect functional recovery and health-care costs. While clinical guidelines recommend that all stroke patients receive a standardized assessment during the acute hospitalization, the degree and determinants of acute assessment for rehabilitation (AAR) following ischemic stroke are unknown. Methods: We analyzed data from 1540 Get With The Guidelines-Stroke hospitals from 01/08/2008 to 03/31/2011. Patients who died in hospital, left AMA, or were transferred in from or out to another acute hospital were excluded. Univariate (chi-square or Wilcoxon as appropriate) and multivariable logistic regression analyses with GEE were used to identify factors independently associated with an AAR while accounting for within hospital clustering. Results: Among 616,982 ischemic stroke patients, 89.5% had an AAR documented. Those without AAR were more likely white, older, female, unable to ambulate prior to admission, from a chronic care facility, have Medicare health insurance and comorbid conditions. Also without an AAR were those with moderate-severe stroke (NIHSS≥6), unable to ambulate on day 2, and were not cared for in a stroke unit. Nine percent of patients discharged home without services were not assessed for rehabilitation. In multivariable analysis, many factors were independently associated with receiving an AAR; however, patients with the greatest odds (OR>1.2) were of Black race, without a history of carotid stenosis, ambulating independently prior to admission, had stroke symptoms outside of a healthcare facility, were treated at a Northeast hospital, in a stroke unit, had complications from thrombolytic therapy, and were ambulating on hospital day 2 ( Table ). Conclusion: Although 90% of ischemic stroke patients received an AAR, the results suggest important subpopulations were overlooked. Quality improvement efforts are needed to ensure that all stroke patients are assessed and referred for the appropriate level of rehabilitation care for their needs. Further research of the unexplained variation in AAR is warranted.

National guidelines recommend dysphagia screening (DS) before any oral intake in hospitalized stroke patients to reduce the risk of hospital-acquired pneumonia (HAP). We examined the relationship between DS and HAP in acute ischemic stroke patients in the Get With the Guidelines-Stroke (GWTG-S) program. Methods: Data from 1251 GWTG-S hospitals from 04/01/2003 to 03/01/2009 were analyzed. GWTG-S defines HAP as a clinical diagnosis of pneumonia requiring antibiotics. Use of a bedside, evidence-based swallow screen prior to any oral intake qualified as a DS. Univariate analyses (chi-square for categorical variables or Wilcoxon for continuous variables) and multivariate logistic regression analyses were performed to examine the relationship between DS and HAP, adjusting for patient and hospital characteristics Results: Among 365,726 ischemic stroke patients, 213097 (59.83%) underwent DS, and 25,166 (6.88%) developed HAP. When compared to patients without pneumonia ( Table 1 ), patients with HAP were older and more frequently had CAD/MI, diabetes, prior stroke/TIA, dyslipidemia, atrial fibrillation. They, underwent DS less often, and had increased length of stay, morbidity and in-hospital mortality. Among the subgroup who had NIHSS recorded (n=160,837, 44%), HAP patients had higher median NIHSS (13 vs. 5). Among patients with NIHSS Discussion: Our data suggests that dysphagia screening is associated with a lower likelihood of HAP, but screening rates remain low. Strategies that increase the rate of dysphagia screening among all stroke patients, even those with mild strokes, should be more broadly implemented. Prospective validation of these findings is warranted. Table 1. Unadjusted associations between patient and hospital characteristics and clinical outcomes Overall(% or value) HAP(% or value) No HAP(% or value) Study Population 365726 (100%) 25166(6.9%) 340560(93.1%) Age (years) Median (IQR) 73(61, 82) 77(66, 85) 73(61,82) Female 52.5 49.2 52.8 Dysphagia Screen Performed 59.8 54.8 60.2 In Hospital Death 5.7 18.1 4.8 P-values are