Abstract WP181: Timing of Percutaneous Endoscopic Gastrostomy (PEG) for Acute Ischemic Stroke Inpatients in the United States

Background and Purpose: Stroke guidelines recommend time-limited trials of nasogastric feeding prior to placement of percutaneous endoscopic gastrostomy (PEG) tubes. We sought to describe timing of PEG placement and identify factors associated with early PEG for acute ischemic stroke.

Methods: We designed a retrospective observational study to examine time to PEG for ischemic stroke admissions in the Nationwide Inpatient Sample, 2001-2011. We defined early PEG placement as 1-7 days from admission. Using multivariable regression analysis, we identified the effects of patient and hospital characteristics on time to PEG.

Results: We identified 34,623 admissions receiving a PEG from 2001-2011, 53% of which received the PEG early. Among hospitals placing ≥10 PEG tubes, median time to PEG for individual hospitals ranged from 3 days to over 3 weeks (interquartile range: 6-8.5 days). Older adult age groups were associated with early PEG placement (≥85 years vs. 18-54 years: Adjusted Odds Ratio [AOR] 1.68, 95% CI 1.50-1.87). Those receiving a PEG tube and tracheostomy were less likely to receive the PEG early (vs. no tracheostomy; AOR 0.27, 95% CI 0.24-0.29), and these patients were more often younger compared to PEG only recipients ( Figure ). Those admitted to high volume hospitals were more likely to receive their PEG early (≥350 vs. <150 hospitalizations; AOR 1.26, 95% CI 1.17-1.35).

Conclusions: More than half of PEG recipients received their surgical feeding tube within 7 days of admission. The oldest old, who may be the most likely to benefit from time-limited trials of nasogastric feeding, were most likely to receive a PEG early.

Timing of Percutaneous Endoscopic Gastrostomy for Acute Ischemic Stroke

Background and Purpose—

Stroke guidelines recommend time-limited trials of nasogastric feeding prior to percutaneous endoscopic gastrostomy (PEG) tube placement. We sought to describe timing of PEG placement and identify factors associated with early PEG for acute ischemic stroke.

Methods—

We designed a retrospective observational study to examine time to PEG for ischemic stroke admissions in the Nationwide Inpatient Sample, 2001 to 2011. We defined early PEG placement as 1 to 7 days from admission. Using multivariable regression analysis, we identified the effects of patient and hospital characteristics on PEG timing.

Results—

We identified 34 623 admissions receiving a PEG from 2001 to 2011, 53% of which received the PEG 1 to 7 days from admission. Among hospitals placing ≥10 PEG tubes, median time to PEG for individual hospitals ranged from 3 days to over 3 weeks (interquartile range 6–8.5 days). Older adult age groups were associated with early PEG (≥85 years versus 18–54 years: adjusted odds ratio 1.68, 95% confidence interval 1.50–1.87). Those receiving a PEG and tracheostomy were more likely to receive the PEG beyond 7 days, and these patients were more often younger compared with PEG only recipients. Those admitted to high-volume hospitals were more likely to receive their PEG early (≥350 versus <150 hospitalizations; adjusted odds ratio 1.26, 95% confidence interval 1.17–1.35).

Conclusions—

More than half of the PEG recipients received their surgical feeding tube within 7 days of admission. The oldest old, who may benefit most from time-limited trials of nasogastric feeding for ≥2 to 3 weeks, were most likely to receive a PEG within 7 days.

Are patients with amyotrophic lateral sclerosis at risk of a therapeutic misconception?

OBJECTIVES

To assess whether persons with amyotrophic lateral sclerosis (ALS) are at risk of a therapeutic misconception (TM) in which they misconceive research as treatment or overestimate the likelihood of its benefit.

METHODS

72 patients with ALS recruited via academic and patient organisations were surveyed using a hypothetical first-in-human intervention study scenario. We elicited their understanding of the purpose of the study ('purpose-of-research question') and then asked how they interpreted the question. We then asked for an estimate of the likelihood that their ALS would improve by participating and asked them to explain the meaning of their estimates.

RESULTS

Although 10 of 72 (14%) subjects incorrectly said that the intervention study was 'mostly intending to help [me]' in response to the purpose-of-research question, 7 of those 10 thought that the question was asking them about their own motivations for participating. Overall, only one of 72 respondents (1.4%) both understood the purpose-of-research question as intended and gave the incorrect response. Subjects' mean estimate of likelihood of benefit was 31% (SD 26). This was due to 29 of 72 of respondents providing high estimates (50%-54% likelihood), which they said were expressions of hope and need for a positive attitude; among those who said their estimates meant 'those are the facts' or 'there is a lot of uncertainty', the estimates were much lower (12.6% and 18.5%, respectively).

CONCLUSIONS

In this group of patients with ALS considering a hypothetical first-in-human intervention study, apparent TM responses have alternative explanations and the risk of true TM appears low.

Pramipexole v. Levodopa as Initial Treatment for Parkinson’s Disease: A Randomized Clinical-Economic Trial

Purpose. To determine the 2-year incremental cost effectiveness of initial pramipexole treatment compared with initial levodopa treatment in patientswith early Parkinson’s disease (PD). Methods. 301 subjects with early PD were randomized to either pramipexole or levodopa and followed every 3 months over a 2-year period. Costs were assigned to patient collected health utilization data using a variety of methods. Health state preferences were estimated using the EuroQol. Results. Pramipexole strategy was an estimated $2,138 (SE = $1,182) more expensive than levodopa strategy. The incremental cost-effectiveness of pramipexole compared with levodopa was $106,900/QALY (EQ-5D), compared with pramipexole being dominated by levodopa using the EQVAS. Conclusions.Although considerable uncertainty exists in the 2-year cost-effectiveness of initial pramipexole compared with initial levodopa in the treatment of early PD, our estimates suggest that pramipexole may not be welfare enhancing during the first 2 years of treatment. If initial pramipexole results in long-term improvements in quality of life, its cost-effectiveness will become more favorable.

The lack of documentation of preferences in a cohort of adults who died after ischemic stroke

OBJECTIVE

To measure the extent and timing of physicians' documentation of communication with patients and families regarding limitations on life-sustaining interventions, in a population cohort of adults who died within 30 days after hospitalization for ischemic stroke.

METHODS

We used the California Office of Statewide Health Planning and Development Patient Discharge Database to identify a retrospective cohort of adults with ischemic strokes at all California acute care hospitals from December 2006 to November 2007. Of 326 eligible hospitals, a representative sample of 39 was selected, stratified by stroke volume and mortality. Medical records of 981 admissions were abstracted, oversampled on mortality and tissue plasminogen activator receipt. Among 198 patients who died by 30 days postadmission, overall proportions and timing of documented preferences were calculated; factors associated with documentation were explored.

RESULTS

Of the 198 decedents, mean age was 80 years, 78% were admitted from home, 19% had mild strokes, 11% received tissue plasminogen activator, and 42% died during the index hospitalization. Preferences about at least one life-sustaining intervention were recorded on 39% of patients: cardiopulmonary resuscitation 34%, mechanical ventilation 23%, nasogastric tube feeding 10%, and percutaneous enteral feeding 6%. Most discussions occurred within 5 days of death. Greater stroke severity was associated with increased in-hospital documentation of preferences (p < 0.05).

CONCLUSIONS

Documented discussions about limitations on life-sustaining interventions during hospitalization were low, even though this cohort died within 30 days poststroke. Improving the documentation of preferences may be difficult given the 2015 Centers for Medicare and Medicaid 30-day stroke mortality hospital performance measure that is unadjusted for patient preferences regarding life-sustaining interventions.

Neurologists as primary palliative care providers: Communication and practice approaches

PURPOSE OF REVIEW

To present current knowledge and recommendations regarding communication tasks and practice approaches for neurologists as they practice primary palliative care, including discussing serious news, managing symptoms, aligning treatment with patient preferences, introducing hospice/terminal care, and using the multiprofessional approach.

RECENT FINDINGS

Neurologists receive little formal palliative care training yet often need to discuss prognosis in serious illness, manage intractable symptoms in chronic progressive disease, and alleviate suffering for patients and their families. Because patients with neurologic disorders often have major cognitive impairment, physical impairment, or both, with an uncertain prognosis, their palliative care needs are particularly challenging and they remain largely uncharacterized and often unmanaged.

SUMMARY

We provide an overview of neuropalliative care as a fundamental skill set for all neurologists.

Palliative Care Needs in the Neuro-ICU

OBJECTIVES

Patients admitted to the neurological or neurosurgical ICU are likely to have palliative care needs. The goals of this project are to encourage the ICU team to identify palliative care needs for patients and their families and potential ways to meet those needs.

DESIGN

Quality improvement project using a parallel-group prospective cohort design.

SETTING

Single neuro-ICU at a large, academic medical center.

PATIENTS

All patients admitted to the neuro-ICU from September 1, 2013, to November 30, 2013.

INTERVENTIONS

We developed a palliative care needs screening tool consisting of four questions: 1) Does the patient have distressing physical or psychological symptoms? 2) Are there specific support needs for patient or family? 3) Are treatment options matched with patient-centered goals? 4) Are there disagreements among teams and family? We implemented this daily screening tool on morning rounds for one of two neurocritical care services that alternate admitting days to a single neuro-ICU. We examined prevalence and nature of palliative care needs and actions to address those needs, comparing the services with and without screening.

MEASUREMENTS AND MAIN RESULTS

Over the 3-month period, 130 patients were admitted to the service with screening and 132 patients to the service without screening. The two groups did not differ with regard to age, gender, Glasgow Coma Scale, or diagnosis. Palliative care needs were identified in 62% of screened patients (80/130). Needs were mainly social support (53%) and establishing goals of care (28%). Screening was associated with more documented family conferences (p = 0.019) and a trend toward more palliative care consultations (p = 0.056).

CONCLUSIONS

We developed a brief palliative care needs screening tool that identified palliative care needs for 62% neuro-ICU patients. This tool was associated with actions to meet these needs, potentially improving care for patients and their families.

Patient Preferences and Surrogate Decision Making in Neuroscience Intensive Care Units

In the neuroscience intensive care unit (NICU), most patients lack the capacity to make their own preferences known. This fact leads to situations where surrogate decision makers must fill the role of the patient in terms of making preference-based treatment decisions, oftentimes in challenging situations where prognosis is uncertain. The neurointensivist has a large responsibility and role to play in this shared decision-making process. This review covers how NICU patient preferences are determined through existing advance care documentation or surrogate decision makers and how the optimum roles of the physician and surrogate decision maker are addressed. We outline the process of reaching a shared decision between family and care team and describe a practice for conducting optimum family meetings based on studies of ICU families in crisis. We review challenges in the decision-making process between surrogate decision makers and medical teams in neurocritical care settings, as well as methods to ameliorate conflicts. Ultimately, the goal of shared decision making is to increase knowledge amongst surrogates and care providers, decrease decisional conflict, promote realistic expectations and preference-centered treatment strategies, and lift the emotional burden on families of neurocritical care patients.

"It is not guaranteed that you will benefit": True but misleading?

BACKGROUND

Participants of early-phase intervention trials for serious conditions provide high estimates of likelihood of benefit, even when informed consent forms do not promise such benefits. However, some technically correct, negatively stated benefits statements—such as "it is not guaranteed that you will benefit"—could play a role in raising expectations of benefit because in ordinary English usage such statements denote a likely but not a certain-to-occur event.

METHODS

An experimental online survey of 584 English-speaking adults recruited online. They were randomized to receive one of two benefit statements ("not guaranteed" vs "some but very small chance"), using a hypothetical scenario of an early-phase clinical trial testing an intervention to treat amyotrophic lateral sclerosis. We assessed respondents' willingness to consider participating in the amyotrophic lateral sclerosis trial, their estimates of likelihood of benefit, and their explanations for those estimates.

RESULTS

The two arms did not differ in willingness to consider participation in the amyotrophic lateral sclerosis trial. Those receiving "not guaranteed" benefit statement had higher estimates of benefit than those receiving "some but very small chance" statement (35.7% (standard deviation 20.2) vs 28.3% (standard deviation 22.0), p < 0.0001). A total of 43% of all respondents chose expressions of positive sentiment (hope and need to stay positive) as explanations of their estimates; these respondents' estimates of benefit were higher than others but similar between the two arms. The effect of benefit statements was greatest among those who chose "Those are just the facts" as the explanation for their estimate (31.0% (standard deviation 22.4%) in "not guaranteed" arm vs 18.9% (standard deviation 21.0%) in comparison arm, p = 0.008).

CONCLUSION

The use of "not guaranteed" language in benefit statements, when compared to "small but very small chance" language, appeared to increase the perception of likelihood of benefit of entering an early-phase trial, especially among those who view their estimates of benefits as "facts." Such "no guarantee" benefit statements may be misleading and should not be used in informed consent forms.

Are therapeutic motivation and having one's own doctor as researcher sources of therapeutic misconception?

BACKGROUND

Desire for improvement in one's illness and having one's own doctor functioning as a researcher are thought to promote therapeutic misconception (TM), a phenomenon in which research subjects are said to conflate research with treatment.

PURPOSE

To examine whether subjects' therapeutic motivation and own doctor functioning as researcher are associated with TM.

METHODS

We interviewed 90 persons with advanced Parkinson's disease (PD) enrolled or intending to enrol in sham surgery controlled neurosurgical trials, using qualitative interviews. Subjects were compared by motivation (primarily therapeutic vs primarily altruistic or dually motivated by altruistic and therapeutic motivation), and by doctor status (own doctor as site investigator vs not) on the following: understanding of purpose of study; understanding of research procedures; perception of chance of direct benefit; and recollection and perceptions concerning the risks.

RESULTS

60% had primarily therapeutic motivation and 44% had their own doctor as the site investigator, but neither were generally associated with increased TM responses. Overall level of understanding of purpose and procedures of research were high. Subjects responded with generally high estimates of probability of direct benefit, but their rationales were personal and complex. The therapeutic-motivation group was more sensitive to risks. Five (5.6%) subjects provided incorrect answers to the question about purpose of research, and yet, showed excellent understanding of research procedures.

CONCLUSIONS

In persons with PD involved in sham surgery clinical trials, being primarily motivated by desire for direct benefit to one's illness or having one's own doctor as the site investigator were not associated with greater TM responses.

Racial/Ethnic differences in process of care and outcomes among patients hospitalized with intracerebral hemorrhage

BACKGROUND AND PURPOSE

Although racial/ethnic differences in care are pervasive in many areas of medicine, little is known whether intracerebral hemorrhage (ICH) care processes or outcomes differ by race/ethnicity.

METHODS

We analyzed 123 623 patients with ICH (83 216 white, 22 147 black, 10 519 Hispanic, and 7741 Asian) hospitalized at 1199 Get With The Guidelines-Stroke hospitals between 2003 and 2012. Multivariable logistic regression with generalized estimating equation was used to evaluate the association among race, stroke performance measures, and in-hospital outcomes.

RESULTS

Relative to white patients, black, Hispanic, and Asian patients were significantly younger, but more frequently had more severe stroke (median National Institutes of Health Stroke Scale, 9, 10, 10, and 11, respectively; P<0.001). After adjustment for both patient and hospital-level characteristics, black patients were more likely to receive deep venous thrombosis prophylaxis, rehabilitation assessment, dysphagia screening, and stroke education, but less likely to have door to computed tomographic time ≤25 minutes and smoking cessation counseling than whites. Both Hispanic and Asian patients had higher odds of dysphagia screening but lower odds of smoking cessation counseling. In-hospital all-cause mortality was lower for blacks (23.0%), Hispanics (22.8%), and Asians (25.3%) than for white patients (27.6%). After risk adjustment, all minority groups had lower odds of death, of receiving comfort measures only or of being discharged to hospice. In contrast, they were more likely to exceed the median length of stay when compared with white patients.

CONCLUSIONS

Although individual quality indicators in ICH varied by race/ethnicity, black, Hispanic, and Asian patients with ICH had lower risk-adjusted in-hospital mortality than white patients with ICH.

Current practices in feeding tube placement for US acute ischemic stroke inpatients

OBJECTIVE

We sought to identify current US hospital practices for feeding tube placement in ischemic stroke.

METHODS

In a retrospective observational study, we examined the frequency of feeding tube placement among hospitals in the Nationwide Inpatient Sample with ≥30 adult ischemic stroke admissions annually with length of stay greater than 3 days. We examined trends from 2004 to 2011 and predictors using data from more recent years (2008-2011). We used multilevel multivariable regression models accounting for a hospital random effect, adjusted for patient-level and hospital-level factors to predict feeding tube placement.

RESULTS

Feeding tube insertion rates did not change from 2004 to 2011 (8.1 vs 8.4 per 100 admissions; p trend = 0.11). Among 1,540 hospitals with 164,408 stroke hospitalizations from 2008 to 2011, a feeding tube was placed 8.8% of the time (n = 14,480). Variation in the rate of feeding tube placement was high, from 0% to 26% between hospitals (interquartile range 4.8%-11.2%). In the subset with available race/ethnicity data (n = 88,385), after controlling for patient demographics, socioeconomics, and comorbidities, hospital factors associated with feeding tube placement included stroke volume (odds ratio [OR] 1.28 highest vs lowest quartile; 95% confidence interval [CI] 1.10-1.49), for-profit status (OR 1.13 vs nonprofit; 95% CI 1.01-1.25), and intubation use (OR 1.66 highest vs lowest quartile; 95% CI 1.47-1.87). In addition, hospitals with higher rates of black/Hispanic stroke admissions had increased risk of feeding tube placement (OR 1.28 highest vs lowest quartile; 95% CI 1.14-1.44).

CONCLUSIONS

Variation in feeding tube insertion rates across hospitals is large. Differences across hospitals may be partly explained by external factors beyond the patient-centered decision to insert a feeding tube.

Ethical considerations in elective amputation after traumatic peripheral nerve injuries

Traumatic peripheral nerve injuries often complicate extremity trauma, and may cause substantial functional deficits. We have encountered patients who request amputation of such injured extremities, with the goal of prosthetic replacement as a means to restore function. Data on long-term outcomes of limb salvage vs amputation are limited and somewhat contradictory, leaving how to respond to such requests in the hands of the treating physician. We present example cases, drawn from our experience with wounded soldiers in a peripheral nerve injury clinic, in order to facilitate discussion of the ways in which these patients stress the system of medical decision-making while identifying ethical questions central to responding to these requests.

Variation in do-not-resuscitate orders for patients with ischemic stroke: implications for national hospital comparisons

BACKGROUND AND PURPOSE

Decisions on life-sustaining treatments and the use of do-not-resuscitate (DNR) orders can affect early mortality after stroke. We investigated the variation in early DNR use after stroke among hospitals in California and the effect of this variation on mortality-based hospital classifications.

METHODS

Using the California State Inpatient Database from 2005 to 2011, ischemic stroke admissions for patients aged≥50 years were identified. Cases were categorized by the presence or the absence of DNR orders within the first 24 hours of admission. Multilevel logistic regression models with a random hospital intercept were used to predict inpatient mortality after adjusting for comorbidities, vascular risk factors, and demographics. Hospital mortality rank order was assigned based on this model and compared with the results of a second model that included DNR status.

RESULTS

From 355 hospitals, 252,368 cases were identified, including 33,672 (13.3%) with early DNR. Hospital-level-adjusted use of DNR varied widely (quintile 1, 2.2% versus quintile 5, 23.2%). Hospitals with higher early DNR use had higher inpatient mortality because inpatient mortality more than doubled from quintile 1 (4.2%) to quintile 5 (8.7%). Failure to adjust for DNR orders resulted in substantial hospital reclassification across the rank spectrum, including among high mortality hospitals.

CONCLUSIONS

There is wide variation in the hospital-level proportion of ischemic stroke patients with early DNR orders; this variation affects hospital mortality estimates. Unless the circumstances of early DNR orders are better understood, mortality-based hospital comparisons may not reliably identify hospitals providing a lower quality of care.

Abstract 19: Hospital Level Variation in Do-Not-Resuscitate Orders for Ischemic Stroke Patients: Implications for National Hospital Comparisons

Background: The Center for Medicare and Medicaid Services (CMS) plans to publicly report hospital-level adjusted ischemic stroke mortality as a marker of hospital quality. Do-not-resuscitate (DNR) orders, which are a well-known predictor of mortality after stroke, are not accounted for in the current CMS risk-adjustment model.

Methods: Using the California State Inpatient Database from 2005-2011, ischemic stroke admissions (primary ICD-9 codes 433.x1, 434.x1, 436) over age 50 were identified. Cases were categorized by the presence/absence of a DNR order within the first 24 hours of admission. Multi-level logistic regression with a random hospital intercept was used to predict inpatient mortality after adjusting for comorbidities, vascular risk factors and demographics. Hospital mortality rank order was assigned based on this model and compared to the results of a second model that included DNR status.

Results: From 356 hospitals, 261,921 cases were identified: 34,436 with early DNR (13.1%) and 227,485 without (86.9%). 43.1% of all mortality occurred in patients with DNR orders. Hospital-level utilization of DNR varied widely (quintile 1, 2.5% vs. quintile 5, 30.7%; p < 0.001). Compared to a model with DNR, a model not accounting for DNR overestimated mortality by at least 1% (absolute) in 36 hospitals (maximum change 2.4%) and underestimated mortality by at least 1% (absolute) in 15 hospitals (maximum change 1.8%). Failure to adjust for DNR orders resulted in substantial hospital reclassification across the rank spectrum (Figure).

Conclusions: There is wide variation in the hospital-level proportion of ischemic stroke patients with DNR orders. To the extent variation in hospital-level DNR orders represents variation in patient preferences, CMS risk-adjusted mortality measures may result in hospital misclassification, and specifically punish hospitals treating a high proportion of patients who prefer lower intensity treatment.

Abstract T MP73: Effect of Hospital Factors on Individual-Level Early Do-Not-Resuscitate Orders in Ischemic Stroke

Background: Race-ethnic (R/E) minorities are less likely to have Do-Not-Resuscitate (DNR) orders than non-Hispanic whites. Since race and ethnicity cluster by hospital, institutional factors may contribute to this variability in DNR orders. We investigated the influence of hospital-level proportion of R/E minorities on R/E differences in individual-level DNR decisions.

Methods: Cases of ischemic stroke were identified from the California State Inpatient Database (age >=50, 2005-2011, ICD-9 codes 433.x1, 434.x1, 436 in primary position). Hospitals were classified by the proportion of R/E minority (Black or Hispanic) patients seen at each institution (reported as quartiles). Individual-level early DNR orders (within 24 hours) were reported by individual-level race-ethnicity (White, Asian, Hispanic, or Black), stratified by hospital-level proportion of R/E minority (Black or Hispanic) patients using multilevel logistic regression with a random hospital intercept.

Results: A total of 259,953 cases of ischemic stroke across 370 hospitals were included (Mean age 74, Female 53%, White 59%, Black 8.8%, Hispanic 17%, Asian 8.9%, Others 6.0%). There was variability in the proportion of R/E minority patients seen at the hospital level (Black: range 0-82%, median 2.3%; Hispanic: range 0-96%, median 12.2%). Increasing hospital-level proportion of minority patients seen was associated with less use of DNR orders across all R/E groups. Similar effects were seen when stratified by hospital-level quartile of Black (Figure “A”) or Hispanic (Figure “B”) patients (p<0.001 for both). White patients with ischemic stroke admitted to hospitals that cared for more minority patients used early DNR orders at a frequency similar to R/E minorities.

Conclusions: The probability of an early DNR order after ischemic stroke is due not only to different treatment patterns across R/E groups, but also to differential treatment patterns at hospitals that care for more R/E minorities.

Risk adjustment of ischemic stroke outcomes for comparing hospital performance: a statement for healthcare professionals from the American Heart Association/American Stroke Association

BACKGROUND AND PURPOSE

Stroke is the fourth-leading cause of death and a leading cause of long-term major disability in the United States. Measuring outcomes after stroke has important policy implications. The primary goals of this consensus statement are to (1) review statistical considerations when evaluating models that define hospital performance in providing stroke care; (2) discuss the benefits, limitations, and potential unintended consequences of using various outcome measures when evaluating the quality of ischemic stroke care at the hospital level; (3) summarize the evidence on the role of specific clinical and administrative variables, including patient preferences, in risk-adjusted models of ischemic stroke outcomes; (4) provide recommendations on the minimum list of variables that should be included in risk adjustment of ischemic stroke outcomes for comparisons of quality at the hospital level; and (5) provide recommendations for further research.

METHODS AND RESULTS

This statement gives an overview of statistical considerations for the evaluation of hospital-level outcomes after stroke and provides a systematic review of the literature for the following outcome measures for ischemic stroke at 30 days: functional outcomes, mortality, and readmissions. Data on outcomes after stroke have primarily involved studies conducted at an individual patient level rather than a hospital level. On the basis of the available information, the following factors should be included in all hospital-level risk-adjustment models: age, sex, stroke severity, comorbid conditions, and vascular risk factors. Because stroke severity is the most important prognostic factor for individual patients and appears to be a significant predictor of hospital-level performance for 30-day mortality, inclusion of a stroke severity measure in risk-adjustment models for 30-day outcome measures is recommended. Risk-adjustment models that do not include stroke severity or other recommended variables must provide comparable classification of hospital performance as models that include these variables. Stroke severity and other variables that are included in risk-adjustment models should be standardized across sites, so that their reliability and accuracy are equivalent. There is a pressing need for research in multiple areas to better identify methods and metrics to evaluate outcomes of stroke care.

CONCLUSIONS

There are a number of important methodological challenges in undertaking risk-adjusted outcome comparisons to assess the quality of stroke care in different hospitals. It is important for stakeholders to recognize these challenges and for there to be a concerted approach to improving the methods for quality assessment and improvement.