FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2803351570> ?p ?o ?g. }

• W2803351570 endingPage "1556" @default.
• W2803351570 startingPage "1549" @default.
• W2803351570 abstract "HomeStrokeVol. 49, No. 6Patient-Reported Outcome Measures (PROMs) for Acute Stroke: Rationale, Methods and Future Directions Free AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissionsDownload Articles + Supplements ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toSupplementary MaterialsFree AccessReview ArticlePDF/EPUBPatient-Reported Outcome Measures (PROMs) for Acute Stroke: Rationale, Methods and Future Directions Mathew Reeves, PhD, Lynda Lisabeth, PhD, Linda Williams, MD, Irene Katzan, MD, Moira Kapral, MD, Anne Deutsch, PhD and Janet Prvu-Bettger, PhD Mathew ReevesMathew Reeves From the Department of Epidemiology, Michigan State University, East Lansing (M.R.) Search for more papers by this author , Lynda LisabethLynda Lisabeth School of Public Health, University of Michigan, Ann Arbor (L.L.) Search for more papers by this author , Linda WilliamsLinda Williams Roudebush VA Medical Center, Indianapolis, IN (L.W.) Search for more papers by this author , Irene KatzanIrene Katzan Neurological Institute, Cleveland Clinic, OH (I.K.) Search for more papers by this author , Moira KapralMoira Kapral Division of General Internal Medicine, Department of Medicine, University of Toronto, ON, Canada (M.K.) Search for more papers by this author , Anne DeutschAnne Deutsch Shirley Ryan AbilityLab, Northwestern University, Chicago, IL (A.D.) Search for more papers by this author and Janet Prvu-BettgerJanet Prvu-Bettger Duke Clinical Research Institute, Durham, NC (J.P.-B.). Search for more papers by this author Originally published1 Jun 2018https://doi.org/10.1161/STROKEAHA.117.018912Stroke. 2018;49:1549–1556Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2018: Previous Version 1 Traditionally, important clinical outcomes after stroke have included survival, stroke recurrence, and the need for long-term aftercare, as well as a diverse range of measures that quantify the direct and indirect impact of stroke on patient functioning. Instruments used to describe patient function include the modified Rankin Scale (mRS)1 and the Barthel Index (BI).2 The fact that the use of the mRS in stroke patients was first published in 1957,1 and the application of the BI to stroke patients was described as early as 1967,2 serves to illustrate that the stroke community has had a long history of measuring outcomes that are meaningful to patients. Although these traditional measures of stroke outcome can be considered inherently patient-centered, there are important distinctions in how these data are assessed and recorded. Of central interest to this report is the collection of health information directly from stroke patients themselves—so-called patient-reported outcome measures (PROMs). It is important to distinguish PROMs from clinician-reported outcomes, which involve the collection of data by clinicians after observation of the patient (which may also include clinical judgment), and observer-reported outcomes which involve data collected by a third party, for example, caregiver.3Over recent years, there has been an explosion of interest in PROMs across research, clinical care, and public health.4–6 The scope of what constitutes a PROM can vary between different organizations and reports, but they all typically include measures of functional status, well-being and health-related quality of life (HRQOL), and symptom burden7; some reports also include patient experiences of care (eg, satisfaction) and health behaviors.5 Regardless of the exact definition, the most essential feature of a PROM is the notion that it represents the status of the patient’s health that comes directly from the patient without interpretation by the clinician or anyone else.4 As such PROMs have the unique feature of describing health status from the viewpoint of the patient. This definition of PROMs holds great promise in terms of improving the quality and efficiency of healthcare.5 Although PROMs are not new to the stroke community, the wider implementation of PROMs into clinical stroke practice with the goals of improving patient outcomes and quality of care,4,5 promoting shared decision-making, and making systems-level comparisons to promote value in health care8 are new. Because this broader application of PROMs is so recent, evidence of the validity of this approach, for example, the degree to which a specific PROM reflects the quality of care, is still lacking.The goals of this article are to describe the rationale for the development and implementation of PROMs in stroke care and research. We include a brief overview of the methodological principles for developing PROMs, an inventory of currently available PROMs for use in stroke patients, a review of the practical issues related to data collection, and the application of stroke PROMs to clinical practice and performance measurement. Recommendations for future work to develop, test, and implement PROMs in stroke patients are also provided.Methodological Principles for Developing PROMsA detailed account of the development and psychometric testing of PROMs is beyond the scope of this article. Table 1 provides an overview of the essential attributes related to the underlying conceptual model, instrument development, and reliability, validity, and interpretation of PROMs. A more detailed description is provided in the online-only Data Supplement; the reader is also referred to excellent reviews9 and texts.10Table 1. Summary of Key Attributes, Definitions, and Recommendations Regarding Methodological Features of PROMsAttributeDefinitions, InterpretationComments, RecommendationsConceptual and measurement modelDescription of the underlying theoretical construct and its domains.Should describe target population for instrument.Involves defining how specific items of the instrument map to the domains of the constructReliability (reproducibility)Extent that repeated measurements yield the same results.Uses test-retest reliability designs.The lack of random error.Unidimensional scales: measured by r, K, or R (depending on design/data type).Measured by correlation coefficient (r), kappa (K), or intraclass correlation (R).Multi-item scales: Cronbach alpha (a measure of internal reliability).Correlations >0.70 are desirable.ValidityExtent that an instrument measures what is intendedRequires evaluation of several approaches (ie, construct, content, and criterion validity).Accuracy.Complicated by lack of gold standard for PROMs assessment.Lack of systematic error or bias.Sensitivity/specificity used if gold standard available. Content validityExtent an instrument includes all relevant dimensions without redundancyAssessed qualitatively from patients and cliniciansComprehensivenessAssessed quantitatively using factor analysis, internal reliability testing (Cronbach alpha), and item response theory.Face validity Construct validityDegree that instrument correlates with other known measures of the constructConvergent validity refers to the presence of correlations with domains that are part of the constructExtent to which measure behaves in a way consistent with theoretical construct.Divergent validity refers to the lack of correlations with domains that are not part of the construct Criterion validityDegree that instrument predicts a directly observable phenomenon related to the constructObservable phenomenon is external and distinct from domains that are part of the construct itselfResponsivenessAbility of instrument to reflect clinically meaningful changes in the patient’s healthAnchor-based methods rely on objective independent reports of change from patientsMID is the smallest change in score that is meaningful to patients and cliniciansDistribution-based methods rely on statistical definition of change (eg, >0.5 SD)Interpretation of scoresDegree that scores are easily understood and interpretableInvolves explaining minimum and maximum values and MIDAdapted from Reeve et al9 with permission. Copyright © 2013, Springer Nature. MID indicates minimal important difference; and PROMs, patient-reported outcome measures.Inventory of Currently Available PROM Instruments for Use in Stroke PatientsPROM instruments can be categorized in several different ways, the most important being whether the instrument is designed as a generic, disease-specific, or domain-specific measure.10 PROM instruments can also be classified as either multidimensional, when they measure multiple different constructs or domains, or unidimensional, when they measure a single domain (eg, pain, fatigue). By definition, generic PROMs are applicable to all patients, regardless of their illness or health status. Multidimensional generic PROMs, for example, SF-36 and EuroQOL, incorporate several different health domains or constructs including physical, mental, and social function. Although useful for comparing outcomes at a group level and in making population-level comparisons, generic measures are less capable of measuring individual-level change over time, and may not capture all the aspects of health that are important to stroke patients. For these reasons, multidimensional disease-specific measures have been developed for stroke patients (see QOL PROMs section). Table 2 lists the multidimensional generic and stroke-specific PROMs commonly used in stroke patients.Table 2. Inventory of Multidimensional Generic and Stroke-Specific PROMs Commonly Used in Stroke PatientsMeasurePurpose or FocusTime to AdministerDomains AssessedPhysicalCognitiveSocialRole*DepressionPsychologicalMental HealthSomaticVitality EnergyOtherGeneric EQ5D10Assess QALYs+general health5 minMobility, self-careXXAnxietyPain GHQ-2810Screen for psychological disorders5 minXXAnxietyXInsomnia MOS SF-3610Assess HRQOL not stroke-specific10 minXXXXXPainX Neuro-QOL11Assess QOL5 minMobility, UE/LE, ADL, self-care, B/BXXXXAnxiety, emotional behavior, affectXFatigueCommunication, stigma, sleep disturbance PROMIS 1012Assess general health10 minXThinkingXXXMood, emotionXPainFatigueStroke-specific SS-QOL13Assess HRQOL specific to stroke15 minMobility, UE, self-careThinkingXXMoodXVision, personality, language SIS (64 item)14Assess multiple dimensions poststroke15–20 minMobility, function, strength, ADL/IADLMemory, thinkingXXEmotionCommunication SA-SIP3015Stroke-specific SIP assesses QOL<30 minMobility, body care, movement, ambulationXXEmotional behaviorCommunication, alertness, behavior SATIS-Stroke16Satisfaction with ICF activities and participationN/RMobility self-careLearning, general tasksXXCommunicationHRQOL indicates health-related quality of life; ICF, International Classification of Functioning, Disability and Health; NeuroQOL, Quality of Life in Neurological Disorders; PROM, patient-reported outcome measures; PROMIS, Patient Reported Outcome Measurement Information System; QALY, Quality-adjusted life years; SIP, Sickness Impact Profile; SIS, stroke impact scale; and SS-QOL, Stroke-Specific Quality of Life.*Role includes instrumental activities of daily living.Although multidimensional stroke–specific measures have distinct advantages, their breadth can also compromise their ability to measure change over time. As a result, a broad array of unidimensional domain-specific measures that each assesses one specific aspect of a patient’s health has now emerged. Beyond their broad applicability, domain-specific measures have the advantages of more directly assessing the impact of diseases and treatments and being more sensitive to change. However, although unidimensional domain-specific measures are informative, they do not always correlate with overall health, and so they are often included in conjunction with other performance-based or clinician-observed outcome measures.Patient Reported Outcome Measurement Information System and Quality of Life in Neurological Disorders Measures for StrokeThe shift towards regarding PROMs as the gold standard approach for assessing health at the individual patient-level has led to the development, under the auspices of the National Institutes of Health, of the Patient Reported Outcome Measurement Information System (PROMIS),5 and Quality of Life in Neurological Disorders (NeuroQOL) groups.11 PROMIS and NeuroQoL consist of a constellation of both multidimensional and unidimensional domain-specific measures that are applicable across a wide range of diseases, conditions, and patient populations. NeuroQoL instruments were specifically developed for use in neurological populations including stroke. The large number of PROMIS/NeuroQOL measures that now exist offers the ability to measure a broad array of health domains including function, emotional health (eg, depression, anxiety), symptom burden (eg, fatigue, pain, sleep disturbance), self-efficacy, and social roles, across an even broader range of patient populations.A major feature of PROMIS/NeuroQOL is the methodological rigor used to develop and test the measures. This includes the use of item response theory which utilizes computer adaptive testing to measure self-reported health. With computer adaptive testing, a computerized algorithm maximizes information obtained about the patient’s status by administering questions to the patient based on their responses to previous questions until a prespecified SE is reached (typically this occurs within 5 questions). Responses are converted to summary T-scores that are calibrated to the general population (with mean, 50; SD=10).Potential benefits of using PROMIS/NeuroQOL instruments include more efficient data collection, improved measurement precision, reduced ceiling and floor effects (fewer patients scoring at the extreme ends of the distribution), improved interpretability (through standardized scores), and broad applicability across clinical patients and populations. Short-forms, consisting of static sets of questions, are also available for situations where computer adaptive testing is not feasible. For example, the 10-item PROMIS Global Health (PROMIS-10) is a generic instrument designed to measure subdomains of mental and physical health.12 The International Consortium for Health Outcomes Measurement has recommended the PROMIS-10 instrument for patients with stroke.8 The feasibility of collecting PROMIS/NeuroQoL data in clinical stroke practice has been demonstrated.17,18 However, additional studies on the use of these tools in representative stroke populations, as well as in clinical trials are needed to provide information on the range and variability of scores, change over time, and how to interpret them.19 As a way of illustrating the use of PROMIS measures the online-only Data Supplement includes a panel that compares 2 traditional outcome measures (mRS and BI) with 4 PROMIS-based PROMs (physical health, mental health, anxiety, and activities of daily living) using a hypothetical real-life case example.HRQOL PROMs for StrokeQOL is a multidimensional construct that describes the general well-being of people or societies and thus encompasses a wide array of underlying factors including physical health, mental health, family, social functioning, and environment. HRQOL broadly refers to the impact these factors have on physical and mental health and can be summarized as an individual’s or group’s perceived physical and mental health over time. Because these concepts include dimensions that are important to patients and affect their health, measurement of HRQOL in stroke patients has increased dramatically since the 1990s (Table 2). Like other PROMs, HRQOL measures may be generic or disease-specific. Generic HRQOL instruments are most useful for comparing HRQOL among stroke survivors to the general population or to persons with other health conditions. Disease-specific HRQOL measures are used when the disease is known to impact domains not assessed by generic HRQOL instruments, for example, vision or language effects of stroke. Disease-specific HRQOL measures are also needed when the items measuring a particular domain are known to demonstrate significant ceiling or floor effects.20,21 In stroke patients, generic HRQOL measures typically produce floor effects, where for example, the patient may have preserved function (for example, being able to walk slowly with assistance) which is not reflected by the score because the instrument lacks the specific items or response categories necessary to discriminate across lower levels of physical function (for example, the instrument may ask about ability to walk one block). Floor and ceiling effects lead to reduced precision and the inability to identify true differences. In some cases, the underlying psychometric characteristics of a generic instrument (ie, reliability, validity) are not reproducible in a disease-specific sample; this was demonstrated for the reduced MOS Physical and Mental Composite scores in stroke patients.20 Item response theory-derived instruments like the PROMIS and Neuro-QOL item banks address these limitations, although further testing in stroke populations is needed.Several multidimensional generic HRQOL measures, including EQ5D and SF-36, have been used to describe stroke outcomes (Table 2).10 However, most of these instruments were developed using classical test theory, which although having useful assumptions related to signal detection has the drawback of being sample-dependent. This latter characteristic makes it much more difficult to know how an instrument will perform in a population with different abilities or characteristics than the original sample.Stroke outcome studies have increasingly used disease-specific HRQOL instrument because of their improved content validity, increased precision, and better responsiveness compared with generic instruments. The 2 most commonly used multidimensional disease-specific instruments are the SS-QOL13 and the stroke impact scale (Table 2).14 Both instruments are multidimensional, were developed in collaboration with stroke survivors, have multiple validated language translations, and have been used in ischemic and hemorrhagic stroke. Because NeuroQOL includes item response theory-derived item banks designed to measure the broad health impacts of neurological conditions, they may also serve as acceptable disease-specific PROMs for stroke, although further evaluation is needed.Functional Status PROMs for StrokeFunctional status, broadly defined as physical, cognitive, and social functioning, are the most commonly measured domains for stroke patients. The World Health Organization International Classification of Functioning, Disability and Health framework22 suggests measurement of activity limitations and participation restrictions across an array of functional subdomains (including cognition, communication, mobility, self-care, interpersonal relationships, and social roles) that measure health and disability. The International Consortium for Health Outcomes Measurement stroke working group has a similar multidomain perspective of functional status and has recommended use of the PROMIS-10 along with additional questions to measure mobility, self-care, and communication problems.8 The mRS1 and the BI2 are 2 of the most commonly used measures of functional status—both capture the domains of mobility, self-care, and independence. Neither measure assesses cognitive or social function. Although commonly used as end points in clinical trials, the mRS and BI are so broad that additional instruments are often needed to provide more detailed data on the level of functioning and changes over time.An advantage of PROMs of functional status is the opportunity for clinicians and investigators to evaluate a specific domain in ways meaningful to the patient. Assessment of mobility through the measurement of walking speed is a good example; serial measurements afford the opportunity to assess recovery and monitor function over the episode of illness. However, depending on the research question and available data, several instruments may be necessary to fully examine changes in functional status over the course of stroke recovery.There can be disadvantages to using PROMs of functional status. PROMs that rely on self-appraisal of functioning are often questioned for their accuracy. Discrepancies between patient-reported functional status and clinician-observed or performance-based assessments have led investigators to collect a range of other measures including objective assessments of performance, although this can create challenges relative to the burden of data collection.Symptom-Based PROMs for StrokeSymptoms experienced by stroke patients are obviously best assessed through patient self-report. Most symptom-based instruments available for use in stroke patients were originally developed in other disease conditions, and many have not been adequately evaluated for use in stroke. Depression, anxiety, and fatigue are all symptoms recommended by clinical guidelines for assessment in stroke patients. We highlight several instruments that can be used to measure symptom-based PROMs in stroke patients, with a focus on their reliability and validity. We also note that PROMIS/NeuroQoL include several unidimensional domain-specific measures of symptoms relevant to stroke patients including pain, emotional distress (anger), stigma, and stress. However, experience with the use of these measures in stroke patients is currently limited.Poststroke depression is common and can hamper recovery after stroke. Routine screening for depressive symptoms in stroke survivors is recommended in several guidelines. According to a recent meta-analysis, the 3 most accurate screening tools for poststroke depression are the Center of Epidemiological Studies-Depression Scale, the Hamilton Depression Rating Scale, and the Patient Health Questionnaire-9.23 All have sensitivity values at or above 0.75 and specificity at or above 0.79. Given its shorter length, the Patient Health Questionnaire-9 may be more practical to use in a clinical setting.Although several instruments exist to measure anxiety in patients, there is no consensus on the optimal instrument to use to screen stroke patients. The 7-item anxiety subscale of the Hospital Anxiety and Depression Scale-A was identified in a recent meta-analysis as the most commonly used scale in studies of anxiety poststroke.24 It has high sensitivity (0.89) but lacks specificity (0.56).25 The Generalized Anxiety Disorder Scales 2 and 7 are frequently used across many conditions and have demonstrated good reliability, criterion validity, sensitivity, and specificity in primary care settings.26 However, they have yet to be formally evaluated in stroke patients.A recent scientific statement on poststroke fatigue highlights the importance of monitoring this often under-recognized symptom in stroke patients.27 The statement provides a concise summary of the performance of fatigue scales that have been used in stroke and suggests the use of the 10-item Fatigue Severity Scale, which is the most commonly used instrument. The Fatigue Severity Scale has good internal consistency and convergent validity28 for assessment of fatigue in general, although formal evaluation in stroke patients has yet to be done.Practical Issues Related to the Collection of PROM DataPROMs by definition are patient reported and thus require that patients are cognitively able to understand and answer questions and, if administered via interview, do not have hearing or other language dysfunction (ie, comprehension and expression deficits) or visual deficits. Population-based studies of stroke survivors suggest that between a quarter and a third of patients have deficits that could complicate or even prevent them from completing a PROM instrument.29,30 Importantly, there may be differences between those with and without these deficits with respect to age, sex, race-ethnicity, and health status.29,30 Also of importance is the potential for stroke-related neuropsychological deficits to adversely affect the ability of stroke survivors to accurately report PROM data.31 Thus a variety of stroke impairments can result in patients who cannot provide answers, or require assistance from another person, or need to use an alternative data collection method. Exclusion of such subjects can introduce selection bias.To minimize the risk of selection bias, surrogate, or proxy respondents are often used to report information on patient function and activity limitations. However, research suggests that proxies report greater levels of impairment than patients, and the value of proxy reports for more subjective measures, such as QOL and depressive symptoms, is not clear.32 In a review of patient-proxy reliability studies, moderate to high reliability between patients and proxies were noted for certain PROMs,32 but such comparisons are necessarily focused on patients without cognitive impairment or aphasia limiting the generalizability of findings. Also, the fact that stroke severity was the most consistent predictor of disagreement suggests that reliability would be lower for more severely affected patients.32 So although the use of proxies can reduce selection bias, they might introduce measurement error. Depending on the research question, researchers should consider the advantages and disadvantages of including proxy data and, if included, should conduct sensitivity analysis to understand the impact of including proxy data on study findings.Alternative methodological approaches should be considered to improve ascertainment of PROMs across the full spectrum of stroke patients. These include alternative data collection methods and use of statistical techniques to lessen selection bias. Alternative methods for collecting data may include computer-assisted survey instruments, pictorial versions of scales for aphasia patients, and observations of patient behavior versus self-report.33 Development of alternative approaches to collect PROMs is still in its infancy; more studies are needed to evaluate the influence of different modes of administration on reliability and usability. Statistical techniques to minimize selection bias include accounting for missing PROM data using statistical imputation methods, such as multiple imputation and inverse probability weighting.34,35 It is important to recognize that these statistical approaches typically assume data are missing at random—meaning that after accounting for variables that predict missingness, the missing data does not introduce selection bias into the study. However, in many clinical and research scenarios, missing PROM data are likely associated with the unobserved values of the outcome itself—meaning the data are missing not at random. Statistical methods that quantify the potential impact of missing not at random data do exist but require additional assumptions.34In summary, reliance on patient-reported information from stroke populations is challenging, and data collected from PROMs may introduce selection bias and measurement error. Additional methodologic work is needed to explore novel data collection methods and statistical approaches that can minimize these effects.Application of PROMs to Clinical Stroke CarePROMs can be useful in clinical practice if they provide timely information that assists in identifying health issues, improving shared decision-making, facilitating personalized care, or reporting on clinician and health system performance.36 There are, however, many practical challenges in implementing PROMs in clinical settings. Key operational barriers include limited time and infrastructure for data collection, and technological challenges integrating PROMs into electronic health records and generating results in a format and time that are useful to clinicians delivering care.37User’s guides for implementing PROMs in clinical practice have been developed and recommendations include36,38 (1) engage clinicians early by defining and communicating the goals of data collection; (2) select measures that are valid, reliable, and easily administered; (3) determine who will complete the measures (eg, patients, proxies), and the setting and timing of completion (eg, inpatient, outpatient, home); (4) determine the mode of data collection (electronic versus paper), recognizing that electronic data permits more rapid calculation and integration of scores into the electronic health record; (5) ensure that the collection of PROMs does not add to clinician workload (eg, reduce administrative burden through collection of previsit information); (6) determine who will receive reports and when they will be provided; (7) provide readily interpretable results to clinicians (eg, clinically meaningful summary scores, graphical information), in time for incorporation into clinical appointments, and with relevant decision supports; and (8) develop mechanisms for responding to issues identified through PROM assessments.Application of PROMs for Performance Measurement in Stroke CareThere is growing interest in the use of PROM-based performance measures to improve healthcare quality.4 A performance measure is a numeric quantification of healthcare quality39 and quality refers to the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge.39 Performance measures can be used for accountability purposes, such as public reporting and value-based purchasing, to incentivize quality improvement. Although many performance measures are in use, with the exception of patient satisfaction measures, few rely on PROMs.4Implementation of PROM-based performance measures involves the routine collection of PROM data across multiple providers (eg, clinics, hospitals), submission of data to a cent" @default.
• W2803351570 created "2018-06-01" @default.
• W2803351570 creator A5028159515 @default.
• W2803351570 creator A5035503064 @default.
• W2803351570 creator A5038719267 @default.
• W2803351570 creator A5056825353 @default.
• W2803351570 creator A5067120188 @default.
• W2803351570 creator A5084006677 @default.
• W2803351570 creator A5085530932 @default.
• W2803351570 date "2018-06-01" @default.
• W2803351570 modified "2023-10-01" @default.
• W2803351570 title "Patient-Reported Outcome Measures (PROMs) for Acute Stroke: Rationale, Methods and Future Directions" @default.
• W2803351570 cites W1889843344 @default.
• W2803351570 cites W1975792159 @default.
• W2803351570 cites W1985265957 @default.
• W2803351570 cites W1990056830 @default.
• W2803351570 cites W1995422036 @default.
• W2803351570 cites W1999019570 @default.
• W2803351570 cites W2006278682 @default.
• W2803351570 cites W2047671141 @default.
• W2803351570 cites W2065333568 @default.
• W2803351570 cites W2070309415 @default.
• W2803351570 cites W2083527976 @default.
• W2803351570 cites W2096281083 @default.
• W2803351570 cites W2106364690 @default.
• W2803351570 cites W2107748660 @default.
• W2803351570 cites W2137447495 @default.
• W2803351570 cites W2138456319 @default.
• W2803351570 cites W2142499410 @default.
• W2803351570 cites W2156762019 @default.
• W2803351570 cites W2157155346 @default.
• W2803351570 cites W2177727633 @default.
• W2803351570 cites W2189726121 @default.
• W2803351570 cites W2230123319 @default.
• W2803351570 cites W2414795604 @default.
• W2803351570 cites W2480680997 @default.
• W2803351570 cites W2499801133 @default.
• W2803351570 cites W2620022897 @default.
• W2803351570 cites W2739450594 @default.
• W2803351570 cites W4211181167 @default.
• W2803351570 cites W4240899207 @default.
• W2803351570 doi "https://doi.org/10.1161/strokeaha.117.018912" @default.
• W2803351570 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/29789396" @default.
• W2803351570 hasPublicationYear "2018" @default.
• W2803351570 type Work @default.
• W2803351570 sameAs 2803351570 @default.
• W2803351570 citedByCount "72" @default.
• W2803351570 countsByYear W28033515702018 @default.
• W2803351570 countsByYear W28033515702019 @default.
• W2803351570 countsByYear W28033515702020 @default.
• W2803351570 countsByYear W28033515702021 @default.
• W2803351570 countsByYear W28033515702022 @default.
• W2803351570 countsByYear W28033515702023 @default.
• W2803351570 crossrefType "journal-article" @default.
• W2803351570 hasAuthorship W2803351570A5028159515 @default.
• W2803351570 hasAuthorship W2803351570A5035503064 @default.
• W2803351570 hasAuthorship W2803351570A5038719267 @default.
• W2803351570 hasAuthorship W2803351570A5056825353 @default.
• W2803351570 hasAuthorship W2803351570A5067120188 @default.
• W2803351570 hasAuthorship W2803351570A5084006677 @default.
• W2803351570 hasAuthorship W2803351570A5085530932 @default.
• W2803351570 hasBestOaLocation W28033515701 @default.
• W2803351570 hasConcept C126322002 @default.
• W2803351570 hasConcept C127413603 @default.
• W2803351570 hasConcept C144237770 @default.
• W2803351570 hasConcept C148220186 @default.
• W2803351570 hasConcept C159110408 @default.
• W2803351570 hasConcept C177713679 @default.
• W2803351570 hasConcept C1862650 @default.
• W2803351570 hasConcept C2776572282 @default.
• W2803351570 hasConcept C2779517570 @default.
• W2803351570 hasConcept C2779951463 @default.
• W2803351570 hasConcept C2780645631 @default.
• W2803351570 hasConcept C3020166492 @default.
• W2803351570 hasConcept C33923547 @default.
• W2803351570 hasConcept C71924100 @default.
• W2803351570 hasConcept C78519656 @default.
• W2803351570 hasConcept C99508421 @default.
• W2803351570 hasConceptScore W2803351570C126322002 @default.
• W2803351570 hasConceptScore W2803351570C127413603 @default.
• W2803351570 hasConceptScore W2803351570C144237770 @default.
• W2803351570 hasConceptScore W2803351570C148220186 @default.
• W2803351570 hasConceptScore W2803351570C159110408 @default.
• W2803351570 hasConceptScore W2803351570C177713679 @default.
• W2803351570 hasConceptScore W2803351570C1862650 @default.
• W2803351570 hasConceptScore W2803351570C2776572282 @default.
• W2803351570 hasConceptScore W2803351570C2779517570 @default.
• W2803351570 hasConceptScore W2803351570C2779951463 @default.
• W2803351570 hasConceptScore W2803351570C2780645631 @default.
• W2803351570 hasConceptScore W2803351570C3020166492 @default.
• W2803351570 hasConceptScore W2803351570C33923547 @default.
• W2803351570 hasConceptScore W2803351570C71924100 @default.
• W2803351570 hasConceptScore W2803351570C78519656 @default.
• W2803351570 hasConceptScore W2803351570C99508421 @default.
• W2803351570 hasIssue "6" @default.
• W2803351570 hasLocation W28033515701 @default.
• W2803351570 hasLocation W28033515702 @default.
• W2803351570 hasOpenAccess W2803351570 @default.

• next