Abstract TMP66: Intravenous Thrombolysis is Associated with Lower Risk of Severe Disability and Death among SPAN-100 Positive Acute Ischemic Stroke Patients

Stroke ◽  
2013 ◽  
Vol 44 (suppl_1) ◽  
Author(s):  
Bruce Ovbiagele ◽  
Mat Reeves ◽  
S. C Johnston ◽  
Philip Bath ◽  
Gustavo Saposnik ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Treatment Effect ◽  
Lower Risk ◽  
Intravenous Thrombolysis ◽  
Severe Disability ◽  
Neurological Deficits ◽  
Stroke Severity ◽  
Low Probability

BACKGROUND: Clinicians are cautious about administering intravenous thrombolysis (tPA) to acute ischemic stroke (AIS) patients who are very elderly and/or have severe neurological deficits. The Stroke Prognostication using Age and NIHSS (SPAN) index combines age plus stroke severity (NIHSS) to create a binary measure (≥ 100 vs. < 100) to predict clinical outcome. We evaluated the effectiveness of tPA by SPAN-100 index status among a large sample of AIS patients. METHODS: Data on 7140 AIS participants in the Virtual International Stroke Trials Archive (VISTA) collaboration. Outcome measures included severe disability or death at 3 months (defined as modified Rankin Scale {mRS} 4-6) and death alone. Effect of tPA on outcomes was assessed using multivariable logistic regression adjusting for SPAN-100 status. RESULTS: Among all patients, 743 (10.5%) were SPAN-100 positive (≥ 100), and 2731 (38.2%) received tPA treatment. Of those treated with tPA, SPAN-100 positive patients were more likely to experience severe disability or death (73.2% vs. 36.3%; p<0.0001) or death alone (33.6% vs. 11.4%; p<0.0001) than SPAN-100 negative patients. However, among SPAN-100 positive patients, tPA was associated with a significantly lower risk of severe disability and death, and tPA had a significantly greater treatment effect among SPAN-100 positive vs. SPAN-100 negative patients (Table). Logistic regression analyses showed significant interactions between SPAN-100 status and tPA (mRS of 4-6 <0.001; death 0.029) confirming that tPA had a greater treatment effect among SPAN-100 positive vs. SPAN-100 negative patients, even after adjustment for age and NIHSS. CONCLUSIONS: Despite the low probability of a favorable outcome, tPA reduces the risk of severe disability and death among SPAN-100 positive AIS patients. SPAN-100 index can be readily used in emergency care settings to identify high risk AIS patients who may be less prone to catastrophic outcomes after tPA treatment.

Abstract WP110: Intravenous Thrombolysis in Acute Ischemic Stroke Patients With Large Vessel Occlusion and Mild Stroke Severity

Stroke ◽  
2020 ◽  
Vol 51 (Suppl_1) ◽  
Author(s):  
Georgios Tsivgoulis ◽  
Nitin Goyal ◽  
Aristeidis H Katsanos ◽  
Konark Malhotra ◽  
Michael T Frohler ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Intravenous Thrombolysis ◽  
Functional Independence ◽  
Large Vessel ◽  
Neurological Deficits ◽  
Stroke Severity ◽  
Large Vessel Occlusion ◽  
Vessel Occlusion ◽  
Safety Outcomes

Introduction: We investigated the effectiveness of intravenous thrombolysis (IVT) in acute ischemic stroke (AIS) patients with large vessel occlusion (LVO) and mild neurological deficits defined as National Institutes of Health Stroke Scale scores <6 points (mELVO). Methods: The primary efficacy outcome was three-month functional independence (FI; mRS-scores of 0-2) that was compared between patients with and without IVT treatment. Other efficacy outcomes of interest included three-month favorable functional outcome (FFO; mRS-scores of 0-1) and mRS-scores distribution at discharge and at 3 months. The safety outcomes comprised all-cause 3-month mortality, symptomatic intracranial hemorrhage (ICH), asymptomatic ICH and severe systemic bleeding. Results: We evaluated 336 AIS patients with mELVO (mean age: 63±15 years, 45% women). Patients treated with IVT (n=162) had higher FI (85.6% vs. 74.8%, p=0.027) with lower mRS scores at hospital discharge (p=0.034) compared to the rest. Three-month mRS-scores tended to be lower in the IVT group (p=0.069). No differences were detected in any of the safety outcomes including symptomatic ICH, asymptomatic ICH, severe systemic bleeding and 3-month mortality (p>0.1). IVT was associated with higher likelihood of 3-month FI (OR=2.19, 95%CI: 1.09-4.42), 3-month FFO (OR=1.99, 95%CI: 1.10-3.57) and functional improvement at discharge [cOR (per 1-point decrease in mRS-score)=2.94, 95%CI: 1.67-5.26] and at 3 months (cOR=1.72, 95%CI: 1.06-2.86) on multivariable logistic regression models adjusting for potential confounders including mechanical thrombectomy. Conclusion: IVT is independently associated with higher odds of improved discharge and three-month functional outcomes in AIS patients with mELVO. IVT does not increase the risk of systemic or intracranial bleeding.

scholarly journals Smoking does not alter treatment effect of intravenous thrombolysis in mild to moderate acute ischemic stroke – a Dutch String-of-Pearls Institute (PSI) Stoke Study

2020 ◽  
Author(s):  
Anna Kufner ◽  
Martin Ebinger ◽  
Gert Jan Luijckx ◽  
Matthias Endres ◽  
Bob Siegerink ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Treatment Effect ◽  
Symptom Onset ◽  
Interaction Analysis ◽  
Intravenous Thrombolysis ◽  
Stroke Severity ◽  
Adjusted Relative Risk ◽  
Stroke Registry ◽  
String Of Pearls

AbstractBackgroundThe smoking-thrombolysis paradox refers to a better outcome in smokers who suffer from acute ischemic stroke (AIS) following treatment with thrombolysis. However, studies on this subject have yielded contradictory results and an interaction analysis of exposure to smoking and thrombolysis in a large, multicenter database is lacking.MethodsConsecutive AIS patients admitted within 12 hours of symptom onset between 2009 and 2014 from the prospective, multicenter stroke registry (Dutch String-of-Pearls Stroke Study) were included for this analysis. We performed a generalized linear model for functional outcome three months post-stroke depending on risk of the exposure variables (smoking yes/no, thrombolysis yes/no). The following confounders were adjusted for: smoking, hypertension, atrial fibrillation, stroke severity, and stroke etiology.ResultsOut of 468 patients, 30.6% were smokers and median baseline NIHSS was 3 (interquartile range 1-6). Smoking alone had a crude and adjusted relative risk (RR) of 0.99 (95% CI 0.89-1.10) and 0.97 (95%CI 0.87-1.01) for good outcome (modified Rankin Score ≤2), respectively. A combination of exposure variables (smoking and thrombolysis) did not change the results significantly (crude RR 0.87 [95% CI 0.74-1.03], adjusted RR 1.1 [95%CI 0.93-1.33]). Smoking alone had an adjusted RR of 1.2 (95% CI 0.6-2.7) for recanalization following thrombolysis (N=88).ConclusionsIn patients with mild to moderate AIS admitted within 12 hours of symptom onset, smoking did not modify treatment effect of thrombolysis.

scholarly journals The anesthetic approach for endovascular recanalization therapy depends on the lesion site in acute ischemic stroke

2021 ◽  
Author(s):  
Kilian Fröhlich ◽  
Gabriela Siedler ◽  
Svenja Stoll ◽  
Kosmas Macha ◽  
Thomas M. Kinfe ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Middle Cerebral Artery ◽  
Acute Ischemic Stroke ◽  
Cerebral Artery ◽  
Posterior Circulation ◽  
Neurological Deficits ◽  
Stroke Severity ◽  
Lesion Volume ◽  
Middle Cerebral Artery Territory ◽  
Cerebral Lesion

Abstract Purpose Endovascular therapy (EVT) of large-vessel occlusion in acute ischemic stroke (AIS) may be performed in general anesthesia (GA) or conscious sedation (CS). We intended to determine the contribution of ischemic cerebral lesion sites on the physician’s decision between GA and CS using voxel-based lesion symptom mapping (VLSM). Methods In a prospective local database, we sought patients with documented AIS and EVT. Age, stroke severity, lesion volume, vigilance, and aphasia scores were compared between EVT patients with GA and CS. The ischemic lesions were analyzed on CT or MRI scans and transformed into stereotaxic space. We determined the lesion overlap and assessed whether GA or CS is associated with specific cerebral lesion sites using the voxel-wise Liebermeister test. Results One hundred seventy-nine patients with AIS and EVT were included in the analysis. The VLSM analysis yielded associations between GA and ischemic lesions in the left hemispheric middle cerebral artery territory and posterior circulation areas. Stroke severity and lesion volume were significantly higher in the GA group. The prevalence of aphasia and aphasia severity was significantly higher and parameters of vigilance lower in the GA group. Conclusions The VLSM analysis showed associations between GA and ischemic lesions in the left hemispheric middle cerebral artery territory and posterior circulation areas including the thalamus that are known to cause neurologic deficits, such as aphasia or compromised vigilance, in AIS-patients with EVT. Our data suggest that higher disability, clinical impairment due to neurological deficits like aphasia, or reduced alertness of affected patients may influence the physician’s decision on using GA in EVT.

Abstract 16967: Increased Blood Pressure Variability is Associated With Worse Outcome in Acute Ischemic Stroke

Circulation ◽  
2015 ◽  
Vol 132 (suppl_3) ◽  
Author(s):  
Adam de Havenon ◽  
Haimei Wang ◽  
Greg Stoddard ◽  
Lee Chung ◽  
Jennifer Majersik
Keyword(s):  
Blood Pressure ◽  
Logistic Regression ◽  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Blood Pressure Variability ◽  
Harmful Effect ◽  
Final Analysis ◽  
Ordinal Logistic Regression ◽  
Stroke Severity ◽  
The Mean

Background: Increased blood pressure variability (BPV) is detrimental in the weeks to months after ischemic stroke, but it has not been adequately studied in the acute phase. We hypothesized that increased BPV in acute ischemic stroke (AIS) patients would be associated with worse outcome. Methods: We retrospectively reviewed inpatients at our hospital between 2010-2014 with an ICD-9 code of AIS; 213 were confirmed to have AIS by a vascular neurologist. A modified Rankin Score (mRS) after discharge was available in 148/213, at a mean of 86 ± 60 days. In 45/213 the discharge mRS was either 0 or 6, in which case they were included in the final analysis. BPV was measured as the standard deviation (SD) of each patient’s systolic blood pressure readings during the first 24 hours and 5 days of hospitalization (9,844 total readings), or until discharge if discharged in <5 days (Figure 1). The SBP SD was further divided in quartiles. A multivariate ordinal logistic regression with the outcome of mRS, the primary predictor of quartiles of SBP SD, and baseline NIH stroke scale (NIHSS) to control for initial stroke severity. Results: Mean±SD age was 64.2 ± 16.3 years, NIHSS was 12.6 ± 7.9, and mRS was 2.7 ± 2.1. The mean SBP SDs for the first 24 hours and 5 days were 12.1 ± 6.2 mm Hg and 14.1 ± 4.9 mm Hg. In the ordinal logistic regression model, the quartiles of SBP SD for the first 24 hours and 5 days were positively associated with higher mRS (OR = 1.37, 95% CI 1.01 - 1.74, p = 0.009; OR = 1.30, 95% CI 1.03 - 1.63, p = 0.028). This effect became even more pronounced in patients with the highest quartile of variability (OR = 2.76, 95% CI 1.29 - 5.88, p = 0.009; OR = 2.10, 95% CI 1.01 - 4.36, p = 0.046). Conclusion: In our cohort of 193 patients with AIS, there was a significant association between increased systolic BPV and worse functional outcome, after controlling for initial stroke severity. This data suggests that increased BPV may have a harmful effect for AIS patients, which warrants a prospective observational study.

Nomogram to Predict Symptomatic Intracranial Hemorrhage after Intravenous Thrombolysis in Acute Ischemic Stroke in an Asian Population.

2021 ◽  
Vol 19 ◽  
Author(s):  
Xiaohua Xie ◽  
Jie Yang ◽  
Lijie Ren ◽  
Shiyu Hu ◽  
Wancheng Lian ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Intracranial Hemorrhage ◽  
Calibration Curve ◽  
Intravenous Thrombolysis ◽  
Asian Population ◽  
Individual Risk ◽  
Clinical Effects ◽  
Symptomatic Intracranial Hemorrhage

Background: Symptomatic intracranial hemorrhage (sICH) is a serious hemorrhagic complication after intravenous thrombolysis (IVT) in acute ischemic stroke (AIS) patients. Most existing predictive scoring systems were derived from Western countries Objective: To develop a nomogram to predict the possibility of sICH after IVT in an Asian population. Methods: This retrospective cohort study included AIS patients treated with recombinant tissue plasminogen activator (rt-PA) in a tertiary hospital in Shenzhen, China, from January 2014 to December 2020. The end point was sICH within 36 hours of IVT treatment. Multivariable logistic regression was used to identify risk factors of sICH, and a predictive nomogram was developed. Area under the curve of receiver operating characteristic curves (AUC), calibration curve, and decision curve analyses were performed. The nomogram was validated by bootstrap resampling Results: Data on a total of 462 patients were collected, of whom 20 patients (4.3%) developed sICH. In the multivariate logistic regression model, the National Institute of Health stroke scale scores (NIHSS) (odds ratio [OR], 1.14; 95% confidence interval [CI], 1.06–1.23, P < 0.001), onset to treatment time (OTT) (OR, 1.02; 95% CI, 1.01–1.03, P < 0.001), neutrophil to lymphocyte ratio (NLR) (OR, 1.22; 95% CI, 1.09–1.35, P < 0.001), and cardioembolism (OR, 3.74; 95% CI, 1.23–11.39, P = 0.020) were independent predictors for sICH and were used to construct a nomogram. Our nomogram exhibited favorable discrimination ability [AUC, 0.878; specificity, 87.35%; and sensitivity, 73.81%]. Bootstrapping for 500 repetitions was performed to further validate the nomogram. The AUC of the bootstrap model was 0.877 (95% CI: 0.823–0.922). The calibration curve exhibited good fit and calibration. The decision curve revealed good positive net benefits and clinical effects Conclusion: The nomogram consisted of the predictors NIHSS, OTT, NLR, and cardioembolism could be used as an auxiliary tool to predict the individual risk of sICH in Chinese AIS patients after IVT. Further external verification among more diverse patient populations is needed to demonstrate the accuracy of the model’s predictions.

Abstract T P307: Similar Rates of Early Cognitive Dysfunction after Intracerebral Hemorrhage and Acute Ischemic Stroke

Stroke ◽  
2015 ◽  
Vol 46 (suppl_1) ◽  
Author(s):  
M. Carter Denny ◽  
Suhas S Bajgur ◽  
Kim Y Vu ◽  
Rahul R Karamchandani ◽  
Amrou Sarraj ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Ischemic Stroke ◽  
Intracerebral Hemorrhage ◽  
Acute Ischemic Stroke ◽  
Cognitive Dysfunction ◽  
Hospital Setting ◽  
Stroke Severity ◽  
Regression Analyses ◽  
Post Stroke ◽  
Poor Outcome

Introduction: Post-stroke cognitive dysfunction (CD) affects at least 1/3 of acute ischemic stroke (AIS) patients when assessed at 3 months. Limited data exists on CD in intracerebral hemorrhage (ICH). The role of early, in-hospital cognitive screening using the brief Montreal Cognitive Assessment (mini MoCA) is being investigated at our center. Hypothesis: We assessed the rates of early CD in ICH and AIS and hypothesized that even minor deficits from these disorders causes significant CD. Methods: 1218 consecutive stroke patients admitted from 2/13 to 12/13 were reviewed; 610, 442 with AIS and 168 with ICH, with admission NIHSS and mini MoCAs were included in the final analyses. CD was defined as mini MoCA <9 (max 12). Poor outcome was defined as discharge mRS 4-6. Stroke severity was stratified by NIHSS score of 0-5, 6-10, 11-15, 16-20, 21-42 as in ECASS-I . Chi-squared tests and univariate logistic regression analyses were performed. Results: Baseline characteristics are shown in table 1. AIS and ICH groups were similar with regard to race, gender and stroke severity. ICH patients were younger, had longer stroke service lengths of stay and poorer outcomes than AIS patients (p=0.03, p<0.001, p<0.001). No difference was seen in rates of CD between AIS and ICH patients (60% vs. 57%, p=0.36, OR 1.2 (CI 0.8-1.7)). CD rates ranged from 36% for NIHSS 0-5 to 96% for 21-42 (figure 1). Older patients were twice as likely to have CD (p<0.001, OR 2.2 (CI 1.6 - 3.0)). Patients with CD had five times the odds of having a poor outcome compared to the cognitively intact (p<0.001, OR 5.2 (CI 3.4-7.7)). In univariate logistic regression analyses, age was a significant predictor of CD in AIS, but not in ICH (p= <0.001, p=0.06). Conclusion: Post-stroke CD is common across all severities and occurs at similar rates in AIS and ICH. More than 1/3 of patients with minor deficits (NIHSS 0-5) had CD in the acute hospital setting. Whether early CD is predictive of long term cognitive outcomes deserves further study.

Abstract TP21: Is Presence of Diabetes Mellitus a Key Prognostic Factor among Patients Receiving Endovascular Treatment for Acute Ischemic Stroke?

Stroke ◽  
2017 ◽  
Vol 48 (suppl_1) ◽  
Author(s):  
Seong-Joon Lee ◽  
Yang-Ha Hwang ◽  
Ji Man Hong ◽  
Jin Wook Choi ◽  
Dong-Hun Kang ◽  
...  
Keyword(s):  
Diabetes Mellitus ◽  
Logistic Regression ◽  
Regression Analysis ◽  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Logistic Regression Analysis ◽  
Multiple Logistic Regression Analysis ◽  
Univariate Analysis ◽  
Intravenous Thrombolysis ◽  
The Impact

Introduction: Given the recent positive endovascular therapy trials for acute ischemic stroke (AIS), this therapeutic strategy is now being increasingly incorporated into routine clinical practice. Identifying prognostic factors among AIS patients receiving endovascular revascularization treatments (ERT) in the real world could be important for clinicians and patients. While the impact of diabetes mellitus (DM) on IV thrombolytic outcomes after AIS has been extensively investigated, there is a paucity of data assessing effects of DM on ERT outcomes after AIS. We evaluated the impact of comorbid DM on ERT for AIS. Methods: From Jan 2011 to Feb 2016, patients with AIS who underwent ERT for cervicocephalic occlusions were consecutively enrolled into the Acute Stroke due to Intracranial Atherosclerotic occlusion and Neurointervention - Korean Retrospective (ASIAN KR) registry from 3 hospitals. Patients were excluded if onset to puncture time over 8 hours, in-hospital stroke, or unavailable 3-month mRS. DM was diagnosed if a patient had the history, or hemoglobin A1c on admission was over 6.5. Univariate analysis was performed to compare the characteristics between DM and non-DM population. Multiple logistic regression analysis was used to validate the effect of comorbid DM on 3 month outcomes. Results: Of 721 patients, 667 (93%) were finally included, with 233 DM patients and 434 non-DM patients. In the univariate analysis, comorbidity with hypertension (71.2% vs. 58.3%, p=0.001) and dyslipidemia (36.7% vs. 26.7%, p=0.012) were more frequent in the DM population. Periprocedural factors such as target vessels, intravenous thrombolysis, and final reperfusion grades did not differ. Good outcomes with mRS 0-2 were less frequent in the DM population (43.3% vs. 53.7%, p=0.011). In the logistic regression analysis adjusting age, male sex, initial NIHSS, premorbid mRS, hypertension history, atrial fibrillation, intravenous thrombolysis, onset to puncture time and successful reperfusion, DM was an independent predictor of poor outcomes (mRS 3-6; 1.933, 1.274-2.933, p=0.002). Conclusion: In patients receiving ERT for AIS due to cervicocephalic artery occlusions, the presence of DM as a comorbidity confers greater odds of a poor functional outcome.

scholarly journals Endovascular treatment of acute ischemic stroke: the end or just the beginning?

2014 ◽  
Vol 36 (1) ◽  
pp. E5 ◽  
Author(s):  
Maxim Mokin ◽  
Alexander A. Khalessi ◽  
J Mocco ◽  
Giuseppe Lanzino ◽  
Travis M. Dumont ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Endovascular Treatment ◽  
Acute Stroke ◽  
Acute Ischemic Stroke ◽  
Randomized Trials ◽  
Intravenous Thrombolysis ◽  
Lessons Learned ◽  
Neurological Deficits ◽  
Medical Community ◽  
Current Status

Various endovascular intraarterial approaches are available for treating patients with acute ischemic stroke who present with severe neurological deficits. Three recent randomized trials—Interventional Management of Stroke (IMS) III, Mechanical Retrieval and Recanalization of Stroke Clots Using Embolectomy (MR RESCUE), and Synthesis Expansion: A Randomized Controlled Trial on Intra-Arterial Versus Intravenous Thrombolysis in Acute Ischemic Stroke (SYNTHESIS Expansion)—evaluated the efficacy of endovascular treatment of acute ischemic stroke and, after failing to demonstrate any significant clinical benefit of endovascular therapies, raised concerns and questions in the medical community regarding the future of endovascular treatment for acute ischemic stroke. In this paper, the authors review the evolution of endovascular treatment strategies for the treatment of acute stroke and provide their interpretation of findings and potential limitations of the three recently published randomized trials. The authors discuss the advantage of stent-retriever technology over earlier endovascular approaches and review the current status and future directions of endovascular acute stroke studies based on lessons learned from previous trials.

scholarly journals Headache as a Predictive Factor of Severe Systolic Hypertension in Acute Ischemic Stroke

2003 ◽  
Vol 30 (3) ◽  
pp. 210-214 ◽  
Author(s):  
Yoon-Ho Hong ◽  
Yong-Seok Lee ◽  
Seong-Ho Park
Keyword(s):  
Logistic Regression ◽  
Ischemic Stroke ◽  
Cerebral Infarction ◽  
Acute Ischemic Stroke ◽  
Acute Cerebral Infarction ◽  
Stroke Severity ◽  
Plasma Catecholamine ◽  
Catecholamine Level ◽  
Stroke Subtypes ◽  
The Mean

ABSTRACT:Background:Elevation of blood pressure (BP) is common in acute cerebral infarction, with several studies reporting a high plasma catecholamine level or previous hypertension as a contributory factor. However, more comprehensive studies on associated clinical parameters are lacking. Our main aim in undertaking this study was to correlate clinical variables associated with a BPelevation in acute ischemic stroke.Methods:Consecutive patients who were admitted to the emergency room and diagnosed with an acute cerebral infarction within 24 hours after the onset of symptoms were investigated. A BP elevation was defined as a high systolic (³200mmHg) or diastolic (³110 mmHg) pressure. The mean systolic and diastolic BP were compared between the different stroke subtypes, lesion locations (carotid vs. vertebrobasilar), and hemispheric sides. The frequency of symptoms, risk factors, location of the infarct, stroke severity, vascular status and laboratory abnormalities were analyzed in order to build a regression model.Results:One hundred thirty-one patients were recruited (M:F=60:71, mean age 66±12 years) and an elevated BP was identified in 33 patients (25.2%). The mean systolic and diastolic BP did not differ significantly between the stroke subtypes, lesion locations, and hemispheric sides. According to univariate logistic regression, an elevated systolic BP correlated with headache (p=0.01) and underlying hypertension (p=0.02) while an elevated diastolic BP correlated with underlying hypertension (p=0.01). Multivariate logistic regression analysis revealed previous hypertension (OR 5.21, 95% CI 1.40-19.37) and headache (OR 4.09, 95% CI 1.44-11.66) to be independent predictors of an elevated systolic BP.Conclusions:Headache itself is closely associated with severe systolic BP elevation in acute ischemic stroke. Whether treatment of elevated BP improves headache and clinical outcome is not yet known, necessitating future controlled studies.

Sign in / Sign up

Export Citation Format

Share Document