Care Quality and Outcomes of Ischemic Stroke in Patients With Premorbid Dementia: Get With The Guidelines-Stroke Registry

Abstract

BACKGROUND:

Patients with premorbid dementia have been generally excluded from trials of stroke therapies, and their dementia diagnosis may affect the care received. There are few data on the quality of stroke care and outcomes in these patients.

METHODS:

We compared the quality of care and outcomes for acute ischemic stroke patients with versus without premorbid dementia using national data from the Get With The Guidelines-Stroke registry between July 1, 2020, and December 31, 2021. Process outcomes included receiving intravenous thrombolysis, endovascular thrombectomy, and additional national quality measures. Clinical outcomes included ambulatory status at discharge, discharge destination, and mortality. The analyses were adjusted for patient and hospital characteristics.

RESULTS:

Among 609 350 patients with acute ischemic stroke, 29 751 of 546 407 (5.4%) had documented prestroke dementia (median age, 84 [interquartile range, 78–89]; 62.8% female). Patients with versus without premorbid dementia were more likely to arrive via emergency medical services (70.5% versus 46.8%) and had more severe strokes (median National Institutes of Health Stroke Scale score, 7 [interquartile range, 3–15] versus 3 [interquartile range, 1–8]). They were less likely to be admitted to a comprehensive stroke center (17.9% versus 22.7%; P<0.0001), to receive intravenous thrombolysis (9.6% versus 11.1%; adjusted odds ratio [aOR], 0.91 [95% CI, 0.87–0.95]) or endovascular thrombectomy (4.5% versus 7.4%; aOR, 0.62 [95% CI, 0.56–0.68]), attain each of the Get With The Guidelines-Stroke Achievement Measures and Quality Measures, or attain defect-free stroke care (92.0% versus 95.0%; aOR, 0.75 [95% CI, 0.71–0.78]). Patients with premorbid dementia had longer door-to-needle times (adjusted difference, 3.17 minutes [95% CI, 1.34–5.01]), lower odds of being discharged home (43.8% versus 60.1%; aOR, 0.96 [95% CI, 0.93–1.00]), and higher odds of being nonambulatory at discharge (25.8% versus 9.3%; aOR, 1.62 [95% CI, 1.54–1.69]), and of in-hospital mortality or hospice admission (23.3% versus 8.6%; aOR, 1.38 [95% CI, 1.32–1.43]). Symptomatic intracranial hemorrhage after intravenous thrombolysis/endovascular thrombectomy did not differ (5.3% versus 3.7%; aOR, 1.13 [95% CI, 0.97–1.31]).

CONCLUSIONS:

Patients with premorbid dementia experienced slightly poorer quality of stroke care across multiple measures, were less likely to receive acute stroke interventions, and had worse poststroke outcomes than patients without dementia in a large nationwide registry. Our findings underscore the need for concerted efforts to further improve care quality and outcomes in this population.

GRAPHIC ABSTRACT:

Patients with dementia have been generally excluded from randomized-controlled trials of acute stroke therapies like intravenous thrombolysis (IVT) or endovascular thrombectomy (EVT), despite accounting for around one-tenth of patients with ischemic stroke.¹ Besides the absence of trial data to guide treatment decisions, these patients can face cognitive biases that may result in care disparities.²

Indeed, small or single-center studies have suggested that patients with dementia are less likely to receive acute stroke therapies or rehabilitation.^3–6 A Canadian Stroke Network analysis a decade ago, before EVT became standard of care, indicated that such patients were less likely to receive IVT or stroke unit care.³ The extent to which such disparities persist is unknown; most registries do not routinely collect data on prestroke dementia. A detailed examination of the current quality of stroke care in patients with premorbid dementia remains to be undertaken to identify areas of disparity and inform quality improvement initiatives, as emphasized by a recent Scientific Statement from the American Heart Association (AHA).² Therefore, we compared the quality of care for acute ischemic stroke among patients with versus without prestroke dementia using national data from the Get With The Guidelines-Stroke (GWTG-Stroke) registry.

METHODS

Study data are confidential and cannot be shared according to the terms of the contracts signed between participating hospitals and the AHA.

GWTG-Stroke is a nationwide registry maintained by the AHA and American Stroke Association to support continuous quality improvement of care for patients with stroke and transient ischemic attack. Details of the program design have been previously published.^7,8 GWTG-Stroke uses a Web-based Patient Management Tool to collect clinical data on consecutively admitted patients. IQVIA (Parsippany, NJ) serves as the data collection and coordination center. Hospitals received either approval to enroll patients without individual patient consent under the Common Rule or a waiver of authorization and exemption from subsequent review by their institutional review board. Advarra, the institutional review board for the AHA, determined that this study is exempt from board oversight. This study followed Strengthening the Reporting of Observational Studies in Epidemiology (http://www.equator-network.org/reporting-guidelines/strobe/) reporting guidelines.

We included patients admitted to participating hospitals between July 1, 2020 (when the dementia variable was incorporated into routine data collection) and December 31, 2021, with a final diagnosis of ischemic stroke. The exposure of interest was a medical history of dementia versus no dementia. This was ascertained by trained abstractors who reviewed the medical records for evidence of dementia in the past medical history recorded by treating clinicians. The primary outcomes included performance on process measures after the stroke defined by GWTG-Stroke as achievement or quality measures (summarized in Figure 1).⁷ Clinical outcomes included ambulatory status at discharge, discharge destination, and mortality. The analyses were adjusted for patient and hospital characteristics.

Figure 1.

Summary of the key Get With The Guidelines-Stroke (GWTG-Stroke) achievement measures and quality measures assessed in the study for patients with versus without premorbid dementia.

Details of exclusion criteria, measurements, and statistical analyses are presented in the Supplemental Methods. The AHA Precision Medicine Platform (https://precision.heart.org) was used for analyses.

RESULTS

Among 609 350 patients in the GWTG-Stroke registry between July 1, 2020, and December 31, 2021, 552 556 patients met analysis criteria, of whom 546 407 had available data (98.9%; Figure 2). In all, 29 751 patients (5.4%) had documented prestroke dementia and, compared with patients without dementia, were older; more often female, non-Hispanic White, and on Medicare insurance; more likely to have cardiovascular comorbidities; less likely to ambulate independently prestroke; and had more severe strokes (median National Institutes of Health Stroke Scale score, 7 [interquartile range, 3–15] versus 3 [interquartile range, 1–8] without dementia; P<0.0001; Table S1).

Figure 2.

Flow chart for the patients included in our analysis from the Get With The Guidelines-Stroke registry.

When comparing prehospital care (Table S2), patients with premorbid dementia were more likely to arrive with emergency medical services (20 986 [70.5%] versus 241 682 [46.8%]; P<0.0001), but when doing so, were less likely to have advanced hospital notification (9323 [44.4%] versus 143 101 [59.2%] without dementia; P<0.0001). They were less likely to be admitted to comprehensive stroke centers (5332 [17.9%] versus 117 287 [22.7%]; P<0.0001) or transferred in from another hospital (3120 [10.5%] versus 89 020 [17.2%]; P<0.0001).

When comparing in-hospital care (Table S3), patients with premorbid dementia were less likely to receive IVT (2851 [9.6%] versus 57 301 [11.1%] without dementia; adjusted odds ratio [aOR], 0.91 [95% CI, 0.87–0.95]) and had longer door-to-needle times (adjusted difference, 3.17 minutes [95% CI, 1.34–5.01]). Among those arriving within 2 hours of last-known-well, patients with premorbid dementia were less likely to have IVT within 3 hours of last-known-well (1913 [84.3%] versus 38 471 [89.2%]; aOR, 0.76 [95% CI, 0.68–0.84]). They were less likely to receive EVT overall (1341 [4.5%] versus 38 363 [7.4%]; aOR, 0.62 [95% CI, 0.56–0.68]) and among those arriving within 6 hours of last-known-well with a National Institutes of Health Stroke Scale score ≥6 (494 [8.9%] versus 12 319 [21.0%]; aOR, 0.45 [95% CI, 0.40–0.50]); however, they had similar rates of recanalization (Thrombolysis in Cerebral Infarction grade ≥2b) when receiving EVT (88.7% versus 89.1%; aOR, 1.01 [95% CI, 0.85–1.19]). They were also less likely to attain each of the GWTG-Stroke Achievement Measures and Quality Measures—except stroke rehabilitation assessment (aOR, 1.02 [95% CI, 0.84–1.24])—and less likely to attain defect-free stroke care (aOR, 0.75 [95% CI, 0.71–0.78]).

Patients with premorbid dementia had worse poststroke outcomes (Table S4), including lower odds of being discharged home if not previously in a care facility (10 354 [43.8%] versus 297 111 [60.1%] without dementia; aOR, 0.96 [95% CI, 0.93–1.00]), higher odds of in-hospital mortality or hospice admission (6297 [23.3%] versus 44 320 [8.6%]; aOR, 1.38 [95% CI, 1.32–1.43]) and being unable to ambulate independently at discharge if ambulatory preadmission (4386 [25.8%] versus 30 949 [9.3%]; aOR, 3.40 [95% CI, 3.28–3.53]), and longer length of stay (median, 5 days; interquartile range, 3–7 versus 4 days; interquartile range, 2–6; aOR [LOS >4 days], 1.17 [95% CI, 1.14–1.21]). However, they had similar rates of symptomatic intracranial hemorrhage (250 [5.3%] versus 4119 [3.7%]; aOR, 1.13 [95% CI, 0.97–1.31]) or serious systemic hemorrhage within 36 hours of IVT.

DISCUSSION

In this large national registry of ischemic stroke patients, we found that 5.4% had premorbid dementia. These patients received poorer stroke care and experienced worse outcomes.

Patients with premorbid dementia were less likely to receive advanced emergency medical services notification or be taken to specialized stroke hospitals, despite more severe strokes. This may reflect reduced care aggressiveness—as they were less likely to be transferred in from another hospital—and the challenges of emergency medical personnel in recognizing strokes in patients with preexisting dementia.

Patients with premorbid dementia were less likely to receive IVT/EVT. The absolute difference in IVT utilization among eligible patients was 5%, similar to the 5.2% difference reported by the Canadian Stroke Network 16 years ago.³ Among patients presenting within 6 hours with a National Institutes of Health Stroke Scale score ≥6, the EVT utilization difference was 12%. This suggests that enthusiasm to offer reperfusion therapies to these patients has changed little in the past decade, despite observational evidence that such therapies have comparable safety outcomes in patients with/without dementia.⁹ Indeed, we observed similar rates of symptomatic intracranial and serious hemorrhage in patients treated with IVT/EVT. While AHA guidelines do not clearly exclude patients with prestroke disability or dementia from IVT/EVT,¹⁰ patients with premorbid disability have accounted for substantial proportions of EVT-treated patients in some centers (eg, 5.5% had a prestroke modified Rankin Scale score of >3 in 1 cohort from 2 stroke centers).¹¹ Importantly, given the limitations of our registry data, we do not know why some patients were treated while others were not; it may well be that the values and goals of care of patients or families influenced these decisions. Concerningly, we found that even when patients with premorbid dementia were treated with IVT, they had longer door-to-needle times. Delayed treatment initiation may erode potential IVT benefits, so fast treatment should still be prioritized in this group.¹²

Patients with premorbid dementia were less likely to achieve GWTG-Stroke Achievement Measures, Quality Measures, and defect-free stroke care. There is little rationale for why these patients would benefit less from well-established stroke care strategies like anti-thrombotic initiation, venous thromboembolism prophylaxis, or statins. The findings from our large data set add clarity to conflicting findings of prior smaller studies and should encourage stroke teams to pursue defect-free stroke care in these patients.^4,5

Patients with prestroke dementia also had worse poststroke outcomes even after adjusting for treatment variables. The magnitude of differences in outcomes exceeded absolute differences in care-quality measures; for example, rates of home discharge were 27% lower in patients with prestroke dementia. Prior smaller studies have had conflicting findings about whether premorbid dementia is independently associated with poststroke outcomes.^4,5 The finding that these patients are less likely to achieve conventional favorable outcomes focused on independence is perhaps unsurprising since they by definition already require assistance for daily activities.² Future studies should compare outcomes with and without reperfusion therapies in dementia patients and incorporate mixed-methods approaches to understand how these patients and caregivers perceive treatment-related trade-offs and what they would consider acceptable poststroke outcomes, which may differ from conventional definitions of favorable outcomes.^2,13

This study has important limitations. Registry data can have variable accuracy and completeness of abstraction. That said, GWTG-Stroke includes several efforts to optimize data quality, including detailed training of site chart abstractors, standardized case definitions and coding instructions, predefined logic and range checks on data fields at data entry, audit trails, and regular data quality reports across sites. Determination of premorbid dementia was limited by information available in patients records, and there was no structured assessment to determine prestroke dementia. As such, we likely underestimated the prevalence of dementia since mild dementia is underdiagnosed in the community.¹⁴ We did not have information on dementia severity; greater severity may be expected to result in lower aggressiveness of care and worse outcomes. Our ability to evaluate EVT utilization was limited by missing data on vessel occlusions among untreated patients. Information on socioeconomic status was lacking.

In conclusion, patients with premorbid dementia experienced slightly poorer quality of stroke care across multiple measures, were less likely to receive acute stroke interventions, and had worse poststroke outcomes than patients without dementia in a nationwide registry. Our findings should motivate concerted efforts to improve care quality and outcomes in this population.

Supplementary Material

Supplemental Methods

Tables S1–S4

STROBE Checklist

Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/STROKEAHA.124.049027.

Nonstandard Abbreviations and Acronyms

AHA

American Heart Association

aOR

adjusted odds ratio

EVT

endovascular thrombectomy

GWTG-Stroke

Get With The Guidelines-Stroke

IVT

intravenous thrombolysis