Determining Differences in the Association Between Atrial Fibrillation and Ischemic Stroke Outcomes by Treatment Received

Abstract

Introduction: Whether the association between atrial fibrillation (AF) and ischemic stroke (IS) outcomes differs by IS treatment type is unknown. We hypothesize that patients with IS who have AF will have a worse NIH Stroke Scale (NIHSS) and 90-day modified Rankin Scale (mRS) score than non-AF, with differences by IS treatment type. Methods: Patients with, and without AF admitted to Johns Hopkins (2020–2023) with confirmed IS and complete covariates were eligible for inclusion. Consecutive patients either received acute IS treatment (intravenous tissue plasminogen activator (IVtPA), mechanical thrombectomy (MT), or both) or did not receive treatment (2:1 ratio). Multivariable regression models were used to determine the association between AF and discharge NIHSS, or 90-day mRS, separately, with interaction terms for IS treatment type as appropriate. Results: Among 353 IS patients (mean age 69 years, 52.1% female, 54.7% Black), 62 received IVtPA only, 66 received IVtPA then MT, 108 received MT only, and 117 were not treated. Patients with AF (N = 152) were, on average, 11 years older and had more comorbidities than non-AF. AF was associated with higher odds of an NIHSS > 5, even after adjusting for demographics and comorbidities (OR 2.09, 95% CI 1.29–3.40). AF increased the odds of a worse 90-day mRS (Ordinal OR 1.65, 95% CI 1.03–2.64). The association between AF and NIHSS differed by whether MT was received (p-interaction 0.037), but not by IVtPA (p-interaction 0.105). AF and 90-day mRS differed by whether MT was received (p-interaction 0.020), but not by IVtPA (p-interaction 0.139). Patients with AF who did not receive MT had a worse NIHSS (OR 4.24, 95% CI 1.38–13.00) and 90-day mRS (OR 2.79, 95% CI 1.30–1.97) compared to non-AF. The individual effect estimates were not significant for those treated with MT when comparing AF to non-AF. Conclusions: The association between AF (vs non-AF) and both NIHSS and 90-day mRS differed by whether MT was received, but not by IVtPA. Patients with AF who did not receive MT had more severe strokes and worse outcomes than non-AF. These findings suggest that while AF is typically linked to more severe strokes, not receiving MT when eligible is particularly detrimental. Receipt of IVtPA did not appear to make a difference, possibly due to treatment contraindications and delays among those with AF.

1. Introduction

Stroke is a leading cause of long-term neurological disability and ranks as the second leading cause of death globally [1,2]. Atrial fibrillation (AF), the most common cardiac arrhythmia, is an independent risk factor of acute IS, increasing the risk for stroke by fivefold and accounting for 20% to 25% of all acute ischemic strokes (IS) [2,3,4,5,6,7,8,9,10]. The association between AF and worse stroke outcomes has been well documented. For example, the Framingham study found an 84% increase in the odds of 30-day post-stroke mortality among patients with AF after adjusting for age and other cardiovascular risk factors [11].

Intravenous tissue plasminogen activator (IV-tPA) and mechanical thrombectomy (MT) are both widely recognized as important treatments for acute IS, both proven to improve outcomes among those who are eligible [6,7]. Patients with AF would be anticipated to be frequently eligible to receive these therapies, especially MT since an embolic stroke in the setting of AF usually involves the internal carotid artery or the middle cerebral artery, which trials have shown benefit from acute MT [12,13]. AF-related strokes are also maximal in onset, suggesting that patients with AF should arrive at the hospital faster, perhaps, than their non-AF acute IS counterparts, and so, therefore, should present more often in a window of time in which therapies could be administered [14]. This would then suggest that patients with AF should, in fact, have a higher chance of a favorable outcome, which the literature does not support [15].

Perhaps the aforementioned poor outcomes are due to the fact that patients with AF are more likely to be on anticoagulants, which is a contraindication to IV-tPA, making MT the only treatment available [16]. Perhaps patients with AF have concomitant comorbidities that might instead make it more difficult to administer treatment or simply lead to delays when trying to administer the treatment [16]. Regardless of the hypothesized reason, it remains uncertain whether the association between having AF and both stroke severity and post-stroke outcomes, as well as whether receiving acute IS treatment, changes this association.

This study aims to determine whether there is a difference in stroke severity or post-stroke outcomes among those patients with versus without AF and to determine if there are differences in this association by acute IS treatment. The significance of this research lies in its potential to enhance clinical practice, optimizing stroke management among acute patients with IS who have AF.

2. Materials and Methods

2.1. Study Population

This study was conducted at Johns Hopkins Hospital (JHH), a comprehensive and tertiary referral center for strokes. Inclusion criteria for the study were patients admitted consecutively to JHH between January 2020 and April 2023 with an acute IS defined as a clinical deficit lasting over 24 h with imaging evidence of acute ischemia on brain MRI in a vascular territory that corresponded to the clinical deficit. We included patients who were both eligible and went on to receive treatment (IVtPA, MT, or IVtPA followed by MT) or those who were not eligible and, therefore, did not receive treatment. We excluded individuals who did not present with a confirmed ischemic stroke or with unavailable demographic or clinical characteristics that were felt to represent important confounders. We could also not include patients in our analyses if they did not have either a discharge NIHSS or a 90-day modified Rankin Scale.

2.2. Exposure

AF was defined as either a previously documented history of AF, patient-reported AF at the time of admission, or newly diagnosed AF (using ECG or telemetry) during admission.

2.3. Acute Ischemic Stroke Treatment

IVtPA was defined as patients who received IVtPA treatment upon admission, and MT was defined as patients who received MT treatment upon admission. IVtPA, followed by MT, was defined as patients who received IVtPA and then went on to receive MT as they were eligible for both interventions.

2.4. Stroke Severity and 90-Day Outcomes

Stroke severity was defined using the NIH Stroke Severity Scale (NIHSS) performed by a trained cerebrovascular neurologist at the time of discharge. The 90-day modified Rankin Scale (mRS) [17] served as a measure of 90-day stroke outcome and was ascertained at 90 days by a telephone call with trained nursing staff at Johns Hopkins following discharge.

2.5. Covariates

Patient demographics and other clinical risk factors were collected at the time of admission through reports from the patient and were documented in the JHH electrical health records (EHRs). These included age, sex, race, admission NIHSS, history of hypertension, and hemoglobin A1C. Anticoagulation medication use and statin medication use were also obtained from the medical record. Patients were considered to be taking these medications if they were listed as prior medications or if the patient reported them at the time of admission.

The Charlson Comorbidity Index (CCI) was used to define the comorbidity burden of the individual patients. As has been previously described, each of the following conditions was assigned one point: history of myocardial infarction, congestive heart failure, peripheral vascular disease, dementia, chronic obstructive pulmonary disease, connective tissue disease, peptic ulcer disease, liver disease, diabetes mellitus, chronic kidney disease, solid tumor (defined as an active prior history of solid tumor weighted as whether localized or metastatic), and leukemia or lymphoma (defined as an active prior history of leukemia or lymphoma and weighed equally regardless of acuity or type). Following the modified Charlson Index that has been previously used for IS outcomes, we excluded hemiplegia and history of cerebrovascular disease since they are reflected in the evaluation of stroke [18].

2.6. Statistical Methods

Descriptive statistics summarized the baseline characteristics of acute IS patients with and without AF. Categorical variables are reported as percentages with differences compared using the χ² test, while continuous variables are reported as means (SD), and differences are compared using the t-test.

Our primary dependent variable, discharge NIHSS, was dichotomized by the mean (NIHSS = 5) due to the skewed distribution of discharge NIHSS in our study population. Additionally, an NIHSS of 5 is a clinically meaningful score, as thrombectomy trials often label patients as “severe” if the NIHSS is higher than 6 [19].

Logistic regression models were used to determine the association between AF and stroke severity, which was defined using the discharge NIHSS. Ordinal logistic regression models were used to determine the association between AF and stroke outcomes, which were defined using the 90-day mRS. These regression models were adjusted for the following: Model 1 included age, sex, and race; Model 2 included sex, race, and CCI (age excluded as included in the CCI); Model 3 included Model 2 plus admission NIHSS. We assume there is a statistically significant association between AF and discharge NIHSS or 90-day mRS when p-value ≤ 0.05 for all models unless otherwise specified. We also performed a sensitivity analysis where we considered anticoagulation use as a potential mediator and included an interaction term in our regression models to see if the primary results changed.

In order to determine if there was a difference in the association between AF and either stroke severity (NIHSS) or post-stroke outcome (mRS), we included an interaction term for acute IS treatment type (IVtPA; MT) in separate models. After determining whether or not there was a difference with p-interaction set at ≤0.1 as evidence of possible statistical significance, we then determined the individual effect estimates for each treatment group (i.e., acute IS patients with AF versus without AF among those who received IVtPA; acute IS patients with AF versus without AF among those who did not receive IVtPA; similar analyses were used for MT).

All statistical analyses were completed by using STATA v.18.0 [20].

3. Results

3.1. Demographics and Clinical Characteristics of Patients With and Without AF

Among the 353 patients eligible for the study, patients with AF were, on average, older (75 vs. 64 years, p < 0.001) and less likely to be Black individuals (48.0% vs. 59.7%, p = 0.029). Patients with AF had a higher admission NIHSS (12.5 vs. 10.4, p = 0.019) and Charlson Comorbidity Index score (4.8 vs. 3.8, p < 0.001) compared to patients without AF (

Table 1. Demographic and Clinical Characteristics (N = 353).

	Without AF (N = 201)	With AF (N = 152)	p-Value
Age (years)	63.9 (13.6)	74.8 (12.0)	<0.001
Female	96 (47.8%)	88 (57.9%)	0.06
Black Race	120 (59.7%)	73 (48.0%)	0.03
Treatment Type
No acute treatment received	51/201 (25.4%)	66/152 (43.4%)	<0.001
IVtPA Only	50/201 (24.9%)	12/152 (7.9%)	<0.001
MT Only	60/201 (29.9%)	48/152 (31.6%)	0.73
IVtPA followed by MT	40/201 (19.9%)	26/152 (17.1%)	<0.001
Any IVtPA †	90/201 (44.8%)	38/152 (25.0%)	<0.001
Any MT §	100/201 (49.8%)	74/152 (48.7%)	0.84
Anticoagulant use at time of admission	17 (8.5%)	62 (40.8%)	<0.001
Statin use at time of admission	93 (46.3%)	86 (56.6%)	0.06
Current Smoker	59 (29.4%)	22 (14.5%)	<0.001
History of Hypertension	157 (78.1%)	130 (85.5%)	0.08
First Hemoglobin A1c (%) at time of admission	6.2 (1.5)	6.2 (1.5)	0.81
Admission NIHSS	10.4 (8.4)	12.5 (8.5)	0.02
Charlson Comorbidity Index	3.8 (2.3)	4.8 (2.4)	<0.001
CHA2DS2-VASc	3.4 (1.8)	4.1 (1.7)	<0.001
Discharge NIHSS (N = 328)	5.0 (6.2)	7.0 (7.7)	0.007
History of prior stroke	55 (27.4%)	56 (36.8%)	0.06

Continuous variables were displayed with mean (SD); Categorical variables were displayed with shown N (%), ^† Any IVtPA refers to acute IS patients who received either IVtPA Only or IVtPA followed by MT. ^§ Any MT refers to acute IS patients who received either MT Only or IVtPA/MT. Abbreviation: IVtPA: intravenous tissue plasminogen activator, MT: mechanical thrombectomy, NIHSS: National Institutes of Health Stroke Severity.

). Of those patients on anticoagulation, 42 patients were on Eliquis (11%), 11 patients were on Coumadin (3%), 18 patients were on Xarelto (5%), 5 patients were on a Heparin-based product (2%), and 3 patients were on Pradaxa (0.8%).

3.2. Association Between AF and Stroke Severity (NIHSS) and Stroke Outcome (mRS)

AF was significantly associated with higher odds of having a worse NIHSS (>5) at discharge after adjusting for demographics (Model 1: OR 1.78; 95% CI 1.08–2.94) and after additional adjustment for vascular risk factors (Model 2: OR 2.09; 95% CI 1.29–3.40), but not after adjusting for admission NIHSS (Model 3: OR 1.75; 95% CI 0.99–3.10;

Table 2. The Association Between Atrial Fibrillation and Stroke Severity, defined using the NIH Stroke Scale (NIHSS) at Time of Patient Discharge (N = 328), and 90 Day Stroke Outcomes, measured using the Modified Rankin Scale (mRS) (N = 273).

Association Between Atrial Fibrillation and Stroke Severity, or Discharge NIHSS (N = 328)
NIHSS > 5	Model 1 OR	95% CI	Model 2 OR	95% CI	Model 3 OR	95% CI
Atrial Fibrillation	1.78	(1.08, 2.94)	2.09	(1.29, 3.40)	1.75	(0.99, 3.10)
Association Between Atrial Fibrillation and 90-day Post-Stroke Outcome, or mRS (N = 273)
mRS	Model 1 Ordinal OR	95% CI	Model 2 Ordinal OR	95% CI	Model 3 Ordinal OR	95% CI
Atrial Fibrillation	1.59	(0.99, 2.56)	2.02	(1.27, 3.22)	1.65	(1.03, 2.64)

Model 1: Adjusted for age, sex, race (black vs. other); Model 2: sex, race, CCI (Charlson Comorbidity Index); Model 3: Model 2 plus admission NIHSS (NIH Stroke Scale), Abbreviation: OR: odds ratio, CI: Confidence Interval.

).

AF was not associated with higher odds of having worse 90-day stroke outcomes after adjusting for demographics (Model 1: Ordinal OR 1.59; 95% CI 0.99–2.56). However, a significant association was seen after additional adjustment for vascular risk factors (Model 2: Ordinal OR 2.02; 95% CI 1.27, 3.22) and admission NIHSS (Model 3: Ordinal OR 1.65; 95% CI 1.03–2.64; Table 2).

3.3. Sensitivity Analysis Considering Anticoagulation Use as a Potential Mediator Effect

In consideration of the fact that anticoagulation use may be a potential mediator of the effect or be on the causal pathway between a history of AF and either stroke severity at the time of discharge or post-stroke 90-day outcomes, we performed a sensitivity analysis where we considered anticoagulation use as a potential mediator and stratified by anticoagulation use. We did not see a difference in the association between AF and discharge NIHSS (p-interaction = 0.89; No-Anticoagulation: Model 3 OR 1.65, 95% CI 0.82–3.32; Anticoagulation-use: Model 3 OR 1.66, 95% CI 0.41–6.79), or a difference in the association between AF and 90-day mRS (p-interaction = 0.86; No-Anticoagulation: Model 3 OR 1.56 95% CI 0.91–2.71, Anticoagulation: OR 1.76, 95% CI 0.57–5.46) when stratifying by anticoagulation use with effect estimates similar to our primary results.

3.4. Association Between AF and Stroke Severity (NIHSS) and Stroke Outcome (mRS) by Treatment Received

There was a significant difference in the association between AF and stroke severity based on whether or not the patient received MT, which was significant after adjusting for demographics, vascular risk, and admission NIHSS (Model 2, p-interaction = 0.037;

Table 3. The Effect of Acute Stroke Treatment on the Association Between Atrial Fibrillation and Stroke Severity at Time of Discharge (NIH Stroke Scale (NIHSS > 5)).

Association of Atrial Fibrillation and NIHSS, Stratified by Acute Stroke Treatment (N = 328)
	Model 1		Model 2
	p-interaction	OR (95% CI)	p-interaction	OR (95% CI)
Any MT Received (N = 154)	0.003	1.25 (0.63, 2.48)	0.037	1.20 (0.58, 2.45)
MT Not Received (N = 174)		5.64 (2.45, 13.01)		4.24 (1.38, 13.00)
Any IVtPA Received (N = 120)	0.114	3.25 (1.37, 7.71)	0.105	2.92 (1.12, 7.65)
IVtPA Not Received (N = 208)		1.63 (0.90, 2.98)		1.16 (0.56, 2.44)

Model 1: sex, race (black vs. other), CCI (Charlson Comorbidity Index); Model 2: Model 1 plus admission NIHSS (NIH Stroke Scale); p-interaction term significance level set at p ≤ 0.1, Any IVtPA included patients who received IVtPA only and IVtPA followed by MT with available discharge NIHSS. Any MT included patients who received MT only and IVtPA followed by MT with available discharge NIHSS. Abbreviation: OR: odds ratio, CI: Confidence Interval.

). Among those who did not receive MT treatment, patients with AF had a worse discharge NIHSS compared to patients without AF (Model 2: OR 4.24; 95% CI 1.38–13.00). There was no difference in NIHSS between AF and non-AF among those who received MT (Model 2: OR 1.20; 95% CI: 0.58, 2.45). No difference was found between AF and discharge NIHSS based on whether IVtPA treatment was received (Model 2, p-interaction = 0.105).

There was a significant difference in the association between AF and 90-day stroke outcomes by whether or not the patients received MT (Model 2, p-interaction = 0.020;

Table 4. The Effect of Acute Stroke Treatment on the Association Between Atrial Fibrillation 90-day Stroke Outcomes (Modified Rankin Scale (mRS)).

Association of Atrial Fibrillation and mRS, Stratified by Acute Stroke Treatment (N = 273)
	Model 1		Model 2
	p-interaction	Ordinal OR (95% CI)	p-interaction	Ordinal OR (95% CI)
Any MT Received (N = 144)	0.002	1.09 (0.58, 2.08)	0.020	1.08 (0.57, 2.05)
MT Not Received (N = 129)		4.19 (1.98, 8.88)		2.79 (1.30, 5.97)
Any IVtPA Received (N = 107)	0.431	2.83 (1.26, 6.36)	0.139	2.78 (1.23, 6.28)
IVtPA Not Received (N = 166)		1.61 (0.90, 2.87)		1.08 (0.59, 1.97)

Model 1: sex, race (black vs. other), CCI (Charlson Comorbidity Index); Model 2: Model 1 plus admission NIHSS (NIH Stroke Scale); p-interaction term significance level set at p ≤ 0.1, Any IVtPA included patients who received IVtPA only and IVtPA followed by MT with available mRS. Any MT included patients who received MT only and IVtPA followed by MT with available mRS, Abbreviation: Ordinal OR: Ordinal odds ratio, CI: Confidence Interval.

). Patients with AF who did not receive MT had a worse mRS compared to patients without AF (Model 2: Ordinal OR 2.79; 95% CI 1.30, 5.97). No association was found between AF versus non-AF and mRS among those who received MT (Model 2: Ordinal OR 1.08; 95% CI 0.57, 2.05). There was also no difference in the association between AF versus non-AF and mRS by receipt of IVtPA (Model 1, p-interaction = 0.431; Model 2, p-interaction = 0.139).

4. Discussion

Our study now lends support to prior research that suggests that having AF is associated with a more severe stroke and poorer 90-day outcomes among patients presenting to a tertiary care comprehensive stroke center [4,5,6,11,21]. Our analysis by treatment type also suggests that these outcomes do differ based on whether or not patients with AF receive certain acute stroke therapies. Patients with AF who did not receive MT had a more severe stroke at the time of discharge and worse 90-day functional outcomes when compared to those who did not have AF. Notably, there was no difference between having AF versus not and either stroke severity (NIHSS) or stroke outcomes (mRS) by whether or not they received IVtPA.

AF has been consistently associated with worse outcomes following acute IS [5]. A prospective study involving 10,528 patients with acute IS found that older age and increased stroke severity largely account for the association between AF and poor stroke outcomes, a finding supported by multiple studies [5,6,21]. In our study, patients with AF were generally older, had more comorbidities, and presented with more severe baseline strokes; however, even after controlling for these variables, AF remained associated with more severe strokes and worse outcomes.

There are several potential reasons for this difference. It could be that patients with AF have a greater likelihood of hemorrhagic transformation post-stroke, but we did not find that in this cohort. It could be that patients with AF have larger volumes or larger territory infarcts, but we would anticipate that this would be captured by the admission NIHSS, which we included in our adjustment model. It could be that patients with AF have other comorbidities or shared vascular risks that would influence our outcome that was not captured by the Charlson Index, although this measure is a fairly extensive way to measure overall health [18]. Finally, it may be related to a feature of having AF itself, which has been linked previously to other cerebrovascular manifestations (i.e., poor collaterals, white matter changes, microhemorrhages), which could also lead to higher points on the NIHSS, or a worse mRS [4].

Both IVtPA and MT are widely recognized as beneficial treatments for acute IS. However, the degree of benefit from these therapies may vary depending on the presence of AF and remains an area of ongoing investigation. Several studies report lower odds of favorable outcomes in patients with AF receiving IVtPA compared to patients without AF, though these differences were not statistically significant [5]. For instance, the National Institute of Neurological Disorders (NINDS) tPA trials subgroup analysis found patients with AF had lower odds of excellent recovery three months after stroke (OR 0.57; 95% CI 0.38–0.86) but no significant difference in clinical response between IVtPA and placebo [8,9,22]. Similarly, the European Cooperative Acute Stroke Study (ECASS) III trial subgroup analysis suggested lower odds of a favorable 90-day mRS outcome in patients with AF receiving alteplase, though this was not statistically significant (OR, 0.68; 95% CI, 0.30–1.55) [9,23]. Conversely, the VISTA Collaboration showed that tPA provided similar benefits to patients with AF (OR, 1.44; 95% CI, 1.12–1.73) as to non-AF (OR, 1.53; 95% CI, 1.39–1.69) [8,9]. Our study aligns with previous findings, showing no association between AF and IS outcomes.

Predicting the clinical response to thrombolysis in patients with AF with acute IS is challenging due to diverse interactions between AF and age, comorbidities, and the fact that being on anticoagulation is a contraindication for IVtPA. Our study adjusts for demographic factors, vascular risk, and severity of stroke on admission. AF-associated treatment delays, driven by the challenges of determining anticoagulation status and the risk of hemorrhagic complications, could affect the timeliness and effectiveness of acute stroke therapies such as IVtPA and MT. Contraindications to IVtPA, such as concurrent use of anticoagulants, may influence the decision to treat patients with AF when history regarding medication use is not known [13]. Strong contraindications may dominate the anticoagulation decision in many older and/or debilitated patients with AF, even those at the highest risk for future IS [21].

While long-term oral anticoagulation effectively reduces IS risk in patients with AF for both primary and secondary prevention [13], our study found overall a small subset of patients receiving acute IS treatment on anticoagulation at the time of stroke admission. This suggests a need for increased adherence to anticoagulation for stroke prevention. Future investigation into the difference in the association between AF and stroke outcomes based on time to treatment may help explain our findings on why there is no association between AF and IS outcomes based on IVtPA administration.

The inclusion of MT in stroke treatment has significantly reduced stroke-related disability for patients meeting the criteria, with the window for eligibility continuing to expand [22]. Studies remain inconclusive on whether the association between AF and stroke outcomes differ by whether MT is received, although our data strongly suggests that not receiving this therapy is detrimental. The MR CLEAN study endorses that although the effect of treatment appeared smaller in patients with AF than in patients without AF, formal interaction testing was not statistically significant, meaning that it could not be formally proven that endovascular thrombectomy was less effective in patients with AF versus patients without AF [24]. In contrast, the HERMES meta-analysis suggested that MT was effective in improving functional outcomes, regardless of the presence of AF, with no significant difference in outcomes between patients with and without AF [13]. A recent systematic review and meta-analysis of 10 studies concluded that stroke from large-vessel occlusion and AF experienced worse 90-day outcomes than patients without AF despite having similar rates of successful reperfusion, which was determined to possibly be associated with older age and more comorbidities among patients with AF [25], which are variables our study controlled for. Our study observed that not receiving MT was associated with worse stroke severity and outcomes and that having AF was associated with even worse outcomes if MT was not received compared to those without AF. Overall, our findings support that MT is an effective and safe treatment for patients with AF with acute IS compared to those who do not receive it. This may be explained by the concept that cardiac emboli in patients with AF are more likely to result in large vessel occlusion strokes affecting the anterior cerebral circulation [4], where intervention with IVtPA may be limited. The time window for MT treatment is not as acute as receiving IVtPA, and MT is a procedure that allows more time for informed clinical decision-making in light of possible contraindications [1]. Further exploration as to why there may be a significant association between AF and worse stroke outcomes in patients with acute IS who did not receive MT is paramount to prevent this subgroup from more adverse outcomes.

We recognize that there are some limitations to our analysis. Although we carefully adjusted for potential confounders in our analysis, we recognize that there is the potential for other unknown or unmeasured confounders to potentially bias our results. In collecting data on anticoagulation use, we are unable to provide insight into compliance, and the care of the patient prior to presentation to the hospital is not known. However, since this information is often unknown at the time of hospital presentation in a real-world treatment setting, we believe that our results are generalizable to other similar clinical settings. Due to limitations in power, we were not able to perform regression analyses separately by the type of AF that patients had; however, we did perform t-tests to see if there was a difference in mean outcome measures between those with known AF at the time of presentation to the hospital and AF that was diagnosed during admission. There was no difference in either the mean discharge NIHSS (p-value = 0.14) or mean mRS (p-value = 0.84) between the two groups, making us more confident in our decision to include both groups in one exposure variable. However, we recognize that there may be differences in how these two groups respond to treatment, and due to numbers, we were unable to provide stratified results by known versus newly diagnosed AF in our manuscript. We did not collect data on patient complications related to stroke treatment itself. However, there was no difference in the number of deaths in the hospital between those with AF and those without AF. We believe that the 90-day mRS scores and in-hospital mortality data provide at least some way by which to capture this information, but we recognize that this may be a limitation of our results. Finally, we did not collect information on the subtype of ischemic stroke, and we also did not consider recanalization scores post-thrombectomy.

5. Conclusions

In conclusion, we believe that our results support that AF is associated with worse stroke severity and functional outcomes when compared to those with acute IS who do not have AF. Furthermore, there appears to be a difference in the association between AF and stroke severity or outcomes by treatment in that patients with AF who did not receive MT had even worse outcomes when compared to those who did not have AF, even when adjusting for initial severity and other measures that might confound this association. Future studies focusing on time-to-treatment and stroke outcomes in AF subgroups are ongoing and necessary to further clarify these findings.