Transient ischemic attack events and incident cardiovascular and non-cardiovascular complications: Observations from a large diversified multimorbid cohort

Abstract

Background:

Transient ischemic attack (TIA) is a strong signal prompting the incidence of future cardiovascular and non-cardiovascular complications, in light of recent debate on the so-called “stroke-heart syndrome.” We aimed to investigate the relation of TIAs to incident clinical events.

Methods:

Patients were drawn from three health plans with a wide spectrum of age groups and a wide mix of socio-economic/disability status. Two TIA cohorts in a retrospective design were used to achieve the study specific aims: (i) to investigate the incidence of TIA and associated cardiovascular and non-cardiovascular complications within 30 and 90 days from the onset of incident TIA events; and (ii) to examine the potential risk factors for developing incident TIA events in the general population with/without a history of prior stroke.

Results:

The incident TIA cohort consisted of 53,716 patients with an average age of 64.2 years (SD 15.2) and 46.1% male. Following TIA, the incidence proportions of ischemic stroke within 30 and 90 days were 2.7% and 3.8%, respectively, and for incident acute coronary syndrome being 0.94 and 1.84, respectively. Ventricular arrhythmia had proportions of 1.2 and 2.14, respectively within 30 and 90 days, with acute heart failure having values of 0.49 and 0.923. About 45% or more of the cardiovascular and non-cardiovascular complications occurred in the first 30 days following the incident TIA cases. About one-third of the recurrent TIA cases followed the incident TIA cases within a span of 30 days. Amongst comorbidities with stroke in the comorbid history, prior stroke provided the strongest risk factor in terms of odds ratio (OR = 8.34, 95% CI 7.21–9.66) for incident TIA events. Age was strongly associated with incident TIA events. Without a prior history of stroke (ischemic stroke/transient ischemic attack/thrombo-embolic events), valvular disease was the strongest risk factor from among the comorbidities (OR-1.87, 95% CI 1.51–2.32). Age also provided strong associations with incident TIA events.

Conclusions:

Following a TIA, there was a high risk of stroke, acute coronary syndrome, ventricular arrhythmia, acute heart failure, and non-cardiovascular complications.

Keywords

Transient ischemic attack complications risk

Introduction

Stroke complications are frequent and include cardiovascular and non-cardiovascular events.^1,2 They result in increased hospitalizations and costs of care as well as hinder full recovery from the debilitating stroke events.^1,3,4 Furthermore, most medical complications develop within the first few weeks of stroke.^5,6

With respect to the above, attention has recently focused to cardiac complications following acute stroke events in systematic reviews on the subject and has been labeled as “stroke-heart syndrome.”^7–9 For example, Buckley et al.¹⁰ reported a high risk of incident cardiovascular complications within 4 weeks of the onset of ischemic stroke. This has led to a recent position paper from the European Society of Cardiology Council on Stroke, advocating an integrated care approach to post-stroke management.¹¹

While transient ischemic attack (TIA) is one of the strong signals for subsequent stroke complications, there is currently a lack of information on cardiovascular and non-cardiovascular complications of TIA as well as the potential risk factors associated with these complications, in population health studies from large/diversified populations including a wide age spectrum and varied levels of combined socio-economic status and disability.

The specific aims of this study were firstly, to examine the incidence TIA and associated cardiovascular (e.g. heart failure, atrial fibrillation) and non-cardiovascular (e.g. falls, gastrointestinal bleeding, urinary incontinence) complications within 30 and 90 days from the onset of incident TIA events. Second, we investigated the potential risk factors for developing incident TIA events in the general population.

Methods

Population description and data sources

The patient population are drawn from the Commercial, Medicare and Medicaid health plans. The Commercial plan represents a working cohort with their family members, characterized by average to above average socio-economic status and funded by private corporate insurance. The Medicare plan is a governmental program with a mix of socio-economic status and disability leaned toward the elderly. The Medicaid is funded by both Federal and state governments and populated by a combined lower socio-economic and disability status. All patients fitting the criteria for the design and criteria discussed below were included in this study.

A retrospective cohort design was used to achieve the study aims. For Specific Aim 1 of the study (incident TIA and associated complications), the incident TIA event was selected as the first case for a given patient during the 2020 calendar year starting from January 1, 2020. Cardiovascular and non-cardiovascular complications were defined as occurring by at least 1 day past the TIA index date up to 90 days with no later than March 31, 2021. Recurrent TIA events were also recorded past the incident events. A minimum of 2 years of comorbid history were obtained prior to the index date. Patients had medical benefits with continuous enrollment for the length of the study, that is, incident TIA event + 90 days of complication follow-up + 2 years of prior comorbid history.

For Specific Aim 2 (examination of potential risk factors for incident TIA events), the general patient population were selected as having both medical and pharmacy benefits with a minimum continuous enrollment of 48 months up to a maximum of 72 months for a period spanning from January 1 2016 to December 31 2021. The incident TIA event had to have at least 2 years of comorbid history.

All data were extracted from administrative databases. They were subject to US and corporate privacy laws to maintain record confidentiality. Each patient had a unique identifier. Consequently, personal information such as first and last names as well as social security numbers were not dealt with.

Definition of incident events and variables

The index date for an incident TIA case was qualified as the date corresponding to the first medical claim with a TIA ICD 10 code from January 1, 2020 until December 31, 2020. Incident cardiovascular and non-cardiovascular complications were defined within a period of 30 or 90 days past the index date. Additionally, the list of comorbid conditions was identified during a baseline period of 2 years preceding the TIA index date.

The list of index and baseline comorbid conditions were described by Lip et al.^12,13 and were identified from medical claims using primary and/or secondary diagnoses, as summarized in Supplemental Table S1 (for ICD 10 codes). For example, TIA was described in terms of ICD-10 codes symbolizing vertebra-basilar artery syndrome, carotid artery syndrome (hemispheric), multiple and bilateral artery syndromes, amaurosis fugax, transient global amnesia, other transient cerebral ischemic attacks and related syndromes, and unspecified transient ischemic attack. Each comorbid or index condition was treated as a binary variable, with 1 for condition presence and 0 for its absence. The ICD 10 codes for TIA complications are documented in Supplemental Table S2.

Non-clinical variables included age, health plan factor, gender, and smoking. Age was defined as a continuous variable in years or a nominal categorical variable consisting of multiple levels: 18–44 being the reference group or 0; 45–54 or 1; 55–64 or 2; 65–74 or 3; and 75–90 or 4. The health plan factor was defined with the Commercial plan as the referent group or 0, Medicare as 1 and Medicaid as 2. Sex had males as the referent group or 0 and females as 1.

Quantitative analyses

The descriptive and inferential statistics were performed using the SAS Enterprise package (version 9.4). Logistic regression analysis was used to assess the potential risk factors impacting the incidence of TIA events. The stepwise procedure was used to identify the final list of variables in a main effect model (sl-entry = 0.25 “significance level for entry into the model”; sl-stay = 0.05 “significance level for staying in the model formulation”).

Results

The incident TIA cohort consisted of 53,716 patients with an average age 64.2 years and 46.1% male patients (Table 1). The Commercial, Medicare, and Medicaid health plans included 54.0%, 31.6%, and 14.4%, respectively of the population make-up, with the collective plans providing a composite index reflecting a combination of socio-economic status, disability, and age.

Table 1.

Baseline characteristics for examination of TIA events and associated complications.

Baseline characteristic	Commercial	Medicare	Medicaid	Overall
Age group (years)
18–44	4030 (65.1)	123 (2.0)	2037 (32.9)	6190 (100)
45–54	4584 (70.1)	316 (4.8)	1641 (25.1)	6541 (100)
55–64	9000 (69.5)	1288 (10.0)	2654 (20.5)	12,942 (100)
65–74	5276 (43.5)	6243 (51.5)	607 (5.0)	12,126 (100)
75–90	6119 (38.4)	9030 (56.7)	768 (4.8)	15,917 (100)
Age (years), mean (SD)	60.8 (14.7)	74.8 (9.0)	53.3 (14.1)	64.2 (15.2)
Gender
Males	14,076 (56.9)	7648 (30.9)	3023 (12.2)	24,747 (100)
Females	14,933 (51.5)	9352 (32.3)	4684 (16.2)	28,969 (100)
Total	29,009 (100.0)	17,000 (100.0)	7707 (100.0)	53,716 (100)

With respect to the 18–44 age group, the crude incidence of TIA per 100 person-years was low (0.05 per 100 person-years) then increased almost exponentially until it reached 0.805 cases per 100 person-years (Figure 1(a)). Medicare recorded the highest crude rates, followed by Medicaid then the Commercial health plan (Figure 1(b)). Female patients were slightly higher incidence than males.

Figure 1.

Crude incidence rates (new cases/100 person-years) for TIA as a function of age group (years), health plan type or gender.

Outcomes within 30 and 90 days of TIA

Within 30 and 90 days of TIA incidence, there were high proportions of cardiovascular and non-cardiovascular complications (Supplemental Table S2) based on complications documented from various healthcare records, including hospitalizations and outpatient visits. For cardiovascular complications, the incidence proportions of ischemic stroke were 2.7% and 3.8% within 30 and 90 days, respectively (Figure 2(a)), with 72.4% of the cases occurring within the first 30 days (Figure 2(b)).

Figure 2.

Incidence of cardiovascular/non-cardiovascular complications following incident TIA cases.

Acute coronary syndrome (ACS) had incidence proportions of 0.94 and 1.84 within 30 and 90 days, respectively, with 51.2% occurring within the first 30 days. Unstable angina and acute myocardial infarction contributed the most to ACS. The 30- and 90-day incidence proportions were 0.51/0.46 and 0.98/0.91 for unstable angina and myocardial infarction, respectively.

Ventricular arrhythmia had incidence proportions of 1.2 and 2.14, respectively within 30 and 90 days, with acute heart failure having values of 0.49 and 0.923. Atrial fibrillation scored proportions of 1.65 and 2.71, respectively for the 30- and 90-day incidences.

For non-cardiovascular complications, dysphagia incidence proportions were among the highest, with 2.29 and 3.51 values for the 30- and 90-day incidences, respectively. Urinary tract infection and falls were also high, with proportions of 1.49/3.15 and 1.14/2.23, respectively. Furthermore, gastrointestinal bleeding, depression, and acute pulmonary embolism were contributors to the non-cardiovascular complications.

About 45% or more of the cardiovascular and non-cardiovascular complications occurred in the first 30 days following the incident TIA cases (Figure 2(a)). About one-third of recurrent TIA cases followed the incident TIA cases within a span of 30 days.

Incident versus prior history of stroke

Incident TIA patients without a prior history of stroke had a higher incidence proportion of complications relative to those with a prior history of TIA for either the 30- or 90-day incidence (Figure 3 and Supplemental Table S4). The risk ratios for the 30-day incidence were much pronounced than those for the 90 days incidence in all complications except for recurrent TIA and depression.

Figure 3.

30- or 90-day incidence ratio for without prior history of stroke to a prior history of stroke.

Survival times and risk factors for incident TIA events

The general cohort used to examine the TIA survival times consisted of a total of 1,453,083 patients, with an average age of 45.6 years (SD 16.2), with a minimum enrollment of 4 years and up to a maximum of 6 years of follow-up. We investigated at least 2 years of comorbid history to elicit the impactful risk variables leading to incident TIA cases.

The mean and median incident times for the TIA patients were 36.3 (SD 15.8) and 35.8 (IQR 24.0–48.8) months, respectively. The highest prevalence rates of conditions in the comorbid history linked to incident TIA were hypertension (71.8%), lipid disorders (66.9%), spondylosis/intervertebral disks (53.9), stroke (42.7%), and osteoarthritis (34.8%) (Figure 4).

Figure 4.

Prevalence rate (%) for co-morbid history for incident transient ischemic (TIA) events.

With stroke in the comorbid history, prior stroke provided as expected the highest odds ratio (OR = 8.34) from among those risk factors in the main effect modeling contributing to incident TIA events (Table 2(a)). Patients with peripheral artery disease, on the other hand, were at a lesser risk of incident TIA relative to those without (OR = 0.77). Similarly, female patients were at a less risk than males (OR = 0.76) (Table 2(a)). Age was also strongly associated with incident TIA events, with the highest associations for age 55–64 (OR = 3.04) and age 45–54 (OR = 2.42), respectively. Smoking was similarly a potential risk factor for incident TIA (OR = 1.19). Based on these risk factors, the c index value for the main effect logistic regression model for incident TIA was equal to 0.789 (95% CI 0.777–0.801) and the residual chi-square was non-significant at the 5% level.

Table 2.

Main effect modeling using Logistic regression analysis to assess impactful variables for incident TIA events.

(a) With stroke in the comorbid history
Effect	Level	Odds ratio	95% CI
Stroke	1 vs 0	8.34	7.21	9.66
Valvular disease	1 vs 0	1.57	1.33	1.86
Sleep apnea	1 vs 0	1.47	1.14	1.89
Major bleeding	1 vs 0	1.45	1.23	1.72
Hypertension	1 vs 0	1.44	1.26	1.65
Osteoarthritis	1 vs 0	1.38	1.21	1.57
Cognitive impairment	1 vs 0	1.35	1.02	1.79
Lipid disorders	1 vs 0	1.30	1.14	1.47
Spondylosis	1 vs 0	1.29	1.15	1.45
Atrial fibrillation	1 vs 0	1.25	1.02	1.53
Peripheral artery disease	1 vs 0	0.77	0.62	0.95
Age group	75–90 vs 18–44	1.62	1.25	2.08
	65–74 vs 18–44	1.27	1.01	1.61
	55–64 vs 18–44	3.04	2.57	3.60
	45–54 vs 18–44	2.42	2.04	2.87
Smoking	1 vs 0	1.19	1.07	1.34
Gender	1 vs 0	0.76	0.68	0.85
(b) Without stroke in the comorbid history
Valvular disease	1 vs 0	1.87	1.51	2.32
Major bleeding	1 vs 0	1.58	1.26	1.97
Osteoarthritis	1 vs 0	1.39	1.18	1.62
Chronic kid disease	1 vs 0	1.31	1.02	1.68
Hypertension	1 vs 0	1.30	1.12	1.51
Lipid disorders	1 vs 0	1.27	1.10	1.47
Spondylosis	1 vs 0	1.27	1.11	1.46
Age group	75–90 vs 18–44	2.47	1.83	3.34
	65–74 vs 18–44	1.68	1.27	2.22
	55–64 vs 18–44	4.07	3.37	4.92
	45–54 vs 18–44	2.86	2.36	3.47
Smoking	1 vs 0	1.40	1.23	1.59
Gender	1 vs 0	0.78	0.69	0.89

Without a prior history of stroke (ischemic stroke/transient ischemic attack/thrombo-embolic events), valvular disease was the strongest risk factor from among the comorbidities (OR-1.87, 95% CI 1.51–2.32) (Table 2(b)). Age also provided strong associations with incident TIA events in a way similar to the model with stroke in the comorbid history (Table 2(b)). The c index value was equal to 0.740 (95% CI 0.73–0.75) and the residual chi-square was non-significant at the 5% level.

Discussion

In this study of large cohorts of TIA diagnosed patients, our principal findings are as follows: (i) A lower socio-economic/disability status was significantly associated with an increase in crude incidence rates (Medicare/Medicaid vs Commercial); (ii) The incident TIA cohort were at risk of adverse clinical complications including cardiovascular (acute ischemic stroke, recurrent TIA, acute coronary syndrome, ventricular arrhythmia, acute heart failure, and atrial fibrillation) and non-cardiovascular (dysphagia, urinary tract infection, falls, gastrointestinal bleeding, depression, and acute pulmonary embolism); (iii) with stroke in the comorbid history, prior stroke, various comorbidities and male sex are at a higher risk of incident TIA events, while the 45–64 age groups have the highest risk of incident TIA events relative to the 18–44 age group; and (vi) without a history of stroke, valvular disease offered the strongest risk factor for incident TIA events.

As far as we are aware, this is the largest cohort of confirmed TIA diagnosis patients where the risks of incident adverse clinical events are studied, as well as the factors contributing to incident TIA, including socio-economic/disability status. Based on this study, the overall crude incidence rate was 0.211 per 100 person-years. This value is much greater than the overall rate of 0.11 per 100 person-years reported for the US based on the National Hospital Ambulatory Medical Care Survey gathered for the 1992–2000 time period.¹⁴ Furthermore, from the Greater Cincinnati/Northern Kentucky population between 1993 and 1994, the overall race-, age-, and gender-adjusted incidence rate for TIA was found to be 0.083 per 100 person-years.¹⁵ Lioutas et al.¹⁶ found crude rates of 0.119 cases/100 person for a white population with an average age of 70.1 years. These differences in results could be largely attributed to the large diversified contemporary cohort utilized in this investigation.

The results of this study were obtained from three health plans representing a wide mix of age, socio-economic status and disability level. The Commercial plan, with an average age of 60.8 years and characterized by an average to above average socio-economic level, reported an incidence proportion of 0.148 cases/100 person-years. The Medicaid plan, with average age of 53.3 years and representing a combined mix of poor socio-economic/disability status, had a crude incidence proportion of 0.213 cases/100 person-years which is 79% higher than that reported for the Commercial plan. The Medicare plan, which is dominated by the elderly “average age = 74.8 years” with some level of poor socio-economic/disability status, demonstrated the highest level of incidence proportion or 0.762 cases/100 person-years. Although there is no data for TIA in the published literature for these health plans, data reported for myocardial infarction by Lip et al.¹⁷ are in line with those found for incident TIA events. Indeed, these results were much higher than those reported by Lioutas et al.¹⁶ for gender-race combinations (white males: 0.101/100 person-years; white females: 0.068 cases/100 person-years; black males 0.107 cases/100 person-years; black females: 0.093 cases/100 person-years).

The age 18–44 group recorded the lowest TIA crude incident rates with an increase in the rates with an increase in the age group, consistent what is being reported in the literature for crude rates.^15,18 When adjusted for clinical and non-clinical variables, we found that the 55–64 and 45–54 age groups have the highest risk associations followed by the 75–90 and 65–74 age groups. As commonly established in the literature, older age is among the highest risk factors in relation to TIAs, and there is a possibility that the main effects of conditions such as spondylosis and osteoarthritis found more in older age may act as a proxy for additional correlations with age in the absence of complex models. In statistical modeling, this may appear to reduce the effect of collinearity.

Our observations on the complications associated with incident TIA events are consistent with the emerging recognition of the stroke-heart syndrome. For example, Buckley et al.¹⁰ reported that a cohort of incident ischemic patients developed rates of 11.1% acute coronary syndrome, 8.8% atrial fibrillation/flutter, 6.4% heart failure, 1.2% severe ventricular arrythmia, and 0.1% Takotsubo syndrome within 4 weeks following stroke. The mechanisms inducing post-stroke cardiac and non-cardiac injures have been described^7,9; yet, the data are still lacking to better understand these mechanisms. At the present, the data demonstrated in this study suggest that TIA complications do not differ from those recently reported from ischemic stroke¹⁰ and in some cases may be much more pronounced.

However, one should be alerted that specific complications such as dysphagia should be interpreted with caution. Dysphagia incidence is found to be very significant after TIA in this study; however, dysphagia is also a symptom of TIA and stroke. Furthermore, dysphagia is consequential to ischemic stroke as cerebral, cerebellar, or brain stem strokes can impair swallowing physiology.¹⁹ In light of the above, this could also lead to an overestimation of 90-day dysphagia incidence rate. In addition, there is a possibility for evaluation bias. After a patient suffers a TIA, doctors typically search and treat risk factors for cerebrovascular disease as well as associated conditions. Whereas a condition such as depression, atrial fibrillation, or ventricular arrhythmia is more actively searched and detected after TIA, some of these conditions may lead to TIA. As such, there is a potential for an overestimation of these complications.

Given the high risk of incident major adverse cardiovascular events following presentation with a cerebrovascular event, the ESC recently published a position paper on integrated care for post-stroke management.¹¹ This proposes a framework for holistic post-stroke management, based on the ABC strategy devised by Lip.²⁰

One way to prevent future complications for TIA (i.e. ischemic stroke and related adverse cardiac/non-cardiac outcomes) is to predict high risk events. Gupta et al.²¹ provided, in a systematic review, a summary of prior clinical rules used to predict TIA events.^22–29 These rules lack the significant cardiovascular and non-cardiovascular multi-morbidities fueling incident TIA events. Consequently, improved methods such as machine learning techniques ought to be deployed to predict these high-risk events.^12,13 The above discussion is particularly important in the absence of stroke in the comorbid history as the complications following incident TIA tend to be more severe as found from this study findings and previously hypothesized by Sposato et al.⁹

It should be noted that non-cardiovascular morbidities (other than diabetes mellitus) have not been commonly reported in prior clinical rules. In this study, the prevalence of important conditions such as spondylosis/intervertebral disks was quite high. Though uncommon, people suffering from cervical spondylosis may present with TIA because of the presence of abnormal bone spurs or osteophytes that may block blood vessels supplying blood to the brain.³⁰ Such attacks often happen unexplained and can be the main reason for shock among patients. In a recent UK study, osteoarthritis was reported as a potential risk factor for TIA or ischemic stroke.³¹ While the mechanism linking osteoarthritis to TIA is far from being clear, Wang et al.³² have attributed this relationship to the fact that osteoarthritis patients are less physically active due to severe pain in the joints compared to the general population, particularly those with knee or hip osteoarthritis. Additionally, ischemia of the bone decreases cartilage nutrition and induces multiple bone infarcts, which is one potential explanation of the interrelationship with cardiovascular disease.

Limitations

The present investigation has some limitations. First, this was a retrospective observational investigation with potential for biases well described in the scientific literature. Although from epidemiological standpoint, the randomized clinical trial is the gold standard for clinical research, its external validity to real life is quite limited thereby hampering the real values of best practices derived from these trials. Therefore, the extraction of findings from observational studies, which have been subjected to a rigorous validation process, will extend the wings of clinical trials to improve the realm of external validity.

Second, the data were extracted from administrative databases which may be subject to errors and missing data. Additionally, the complications were not directly linked to any respective medications. All databases are, however, constantly subject to robust validation methods to enhance the data accuracy and integrity. With such an approach, the data quality can be substantially improved.³³

Finally, the accuracy of TIA diagnosis can be questioned,³⁴ but we have used validated claims codes for this, suggesting that this would be a physician-based diagnosis, rather than self-reported. In particular, we have used physician-based claims characterizing different types of healthcare services such as ER, inpatient hospitalization, pharmacy, outpatient visits, and outpatient hospital visits. Such an approach appears to improve the detection of incident TIA cases as evidenced in part by the above discussion of overall crude incidence rates of TIA obtained from this study relative to those obtained in the scientific literature.

Conclusions

Following a TIA, there was a high risk of stroke, acute coronary syndrome, ventricular dysfunction, acute heart failure, and non-cardiovascular complications. Given the associations of TIA with comorbidities this supports the need for a more holistic or integrated care approach to post-stroke care.

Supplemental Material

sj-docx-1-eso-10.1177_23969873221146044 – Supplemental material for Transient ischemic attack events and incident cardiovascular and non-cardiovascular complications: Observations from a large diversified multimorbid cohort

Supplemental material, sj-docx-1-eso-10.1177_23969873221146044 for Transient ischemic attack events and incident cardiovascular and non-cardiovascular complications: Observations from a large diversified multimorbid cohort by Gregory YH Lip, Ash Genaidy, Cara Estes, Deborah McKay and Tina Falks in European Stroke Journal

Footnotes

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: the authors report no conflicts of interest in relation to this work. GYHL has been a consultant and speaker for BMS/Pfizer, Boehringer Ingelheim, Daiichi-Sankyo, Anthem. No fees are received personally. GYHL is co-principal investigator of the AFFIRMO project on multimorbidity in AF, which has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 899871.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Informed consent/ethical approval

The data gathered was not subject to requirement of informed consent and ethical approval. However, the study data usage is subject to US privacy laws and Anthem corporate agreements. Our own approvals to use the Anthem data sources for the current study do not allow us to distribute or make patient data directly available to other parties.

Guarantor statement

Ashraf Genaidy

Author contributions

GYHL was involved in conception and design of the study and critical revision of the manuscript. AG was involved in the conception and design of the study, data acquisition, statistical analysis and interpretation of data, drafting and critical revision of the manuscript. CE was involved in interpretation of data and critical revision of the manuscript. DM was involved in the interpretation of data and critical revision of the manuscript. TF was involved in the interpretation of data and critical revision of the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

ORCID iD

Ash Genaidy

Data availability statement

Data are available as presented in the paper. According to US laws and corporate agreements, our own approvals to use the Anthem data sources for the current study do not allow us to distribute or make patient data directly available to other parties.

Supplemental material

Supplemental material for this article is available online.

References

Kumar

Selim

Caplan

LR.

Medical complications after stroke. Lancet Neurol 2010; 9: 105–118.

Asgedom

Gidey

, et al Medical complications and mortality of hospitalized stroke patients. J Stroke Cerebrovasc Dis 2020; 29: 104990.

Langhorne

Stott

Robertson

, et al Medical complications after stroke: a multicenter study. Stroke 2000; 31: 1223–1229.

Roth

Lovell

Harvey

, et al Incidence of and risk factors for medical complications during stroke rehabilitation. Stroke 2001; 32: 523–529.

Indredavik

Rohweder

Naalsund

, et al Medical complications in a comprehensive stroke unit and an early supported discharge service. Stroke 2008; 39: 414–420.

Bovim

Askim

Lydersen

, et al Complications in the first week after stroke: a 10-year comparison. BMC Neurol 2016; 16: 133.

Scheitz

Nolte

Doehner

, et al Stroke-heart syndrome: clinical presentation and underlying mechanisms. Lancet Neurol 2018; 17: 1109–1120.

Nolte

Endres

Scheitz

JF.

Myocardial injury in acute ischemic stroke. Clin & Trans Neurosci 2021; 5: 2514183X2110185.

Sposato

Hilz

Aspberg

, et al Post-stroke cardiovascular complications and neurogenic cardiac injury: JACC state-of-the-art review. J Am Coll Cardiol 2020; 76: 2768–2785.

10.

Buckley

BJR

Harrison

Hill

, et al Stroke-heart syndrome: incidence and clinical outcomes of cardiac complications following stroke. Stroke 2022; 53: 1759–1763.

11.

Lip

Lane

Lenarczyk

, et al Integrated care for optimizing the management of stroke and associated heart disease: a position paper of the European Society of Cardiology Council on Stroke. Eur Heart J 2022; 43: 2442–2460.

12.

Lip

GYH

Genaidy

Tran

, et al Improving stroke risk prediction in the general population: a comparative assessment of common clinical rules, a new multimorbid index, and machine-learning-based algorithms. Thromb Haemost 2022; 122: 142–150.

13.

Lip

GYH

Tran

Genaidy

, et al Improving dynamic stroke risk prediction in non-anticoagulated patients with and without atrial fibrillation: comparing common clinical risk scores and machine learning algorithms. Eur Heart J Qual Care Clin Outcomes 2022; 8: 548–556.

14.

Easton

Saver

Albers

, et al American Heart Association. Definition and evaluation of transient ischemic attack: a scientific statement for healthcare professionals from the American Heart Association/American Stroke Association Stroke Council; Council on Cardiovascular Surgery and Anesthesia; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular Nursing; and the Interdisciplinary Council on Peripheral Vascular Disease. The American Academy of Neurology affirms the value of this statement as an educational tool for neurologists. Stroke 2009; 40: 2276–2293.

15.

Kleindorfer

Panagos

Pancioli

, et al Incidence and short-term prognosis of transient ischemic attack in a population-based study. Stroke 2005; 36: 720–723.

16.

Lioutas

V-A

Ivan

Himali

, et al Incidence of transient ischemic attack and association with long-term risk of stroke. JAMA 2021; 325: 373–381.

17.

Lip

GYH

Genaidy

Tran

, et al Incident and recurrent myocardial infarction (MI) in relation to comorbidities: prediction of outcomes using machine-learning algorithms. Eur J Clin Invest 2022; 52: e13777.

18.

Degan

Ornello

Tiseo

, et al Epidemiology of transient .ischemic attacks using time- or tissue-based definitions: a population-based study. Stroke 2017; 48: 530–536.

19.

Martino

Foley

Bhogal

, et al Dysphagia after stroke incidence, diagnosis, and pulmonary complications. Stroke 2005; 36: 2756–2763.

20.

Lip

GYH

. The ABC pathway: an integrated approach to improve AF management. Nat Rev Cardiol 2017; 14: 627–628.

21.

Gupta

Farrell

Mittal

MK.

Transient ischemic attacks: predictability of future ischemic stroke or transient ischemic attack events. Ther Clin Risk Manag 2014; 10: 27–35.

22.

Johnston

Gress

Browner

, et al Short-term prognosis after emergency department diagnosis of TIA. JAMA 2000; 284: 2901–2906.

23.

Rothwell

Giles

Flossmann

, et al A simple score (ABCD) to identify individuals at high early risk of stroke after transient ischaemic attack. Lancet 2005; 366: 29–36.

24.

Johnston

Rothwell

Nguyen-Huynh

, et al Validation and refinement of scores to predict very early stroke risk after transient ischaemic attack. Lancet 2007; 369: 283–292.

25.

Sciolla

Melis

Rapid identification of high-risk transient ischemic attacks: prospective validation of the ABCD score. Stroke 2008; 39: 297–302.

26.

Giles

Albers

Amarenco

, et al Addition of brain infarction to the ABCD² score (ABCD² I). Stroke 2010; 41: 1907–1913.

27.

Merwick

Albers

Amarenco

, et al Addition of brain and carotid imaging to the ABCD(2) score to identify patients at early risk of stroke after transient ischaemic attack: a multicentre observational study. Lancet Neurol 2010; 9: 1060–1069.

28.

Ghandehari

Ahmadi

Ebrahimzadeh

, et al Assessment of ABCD² scale in patients with transient ischaemic attack or stroke. Neurol Neurochir Pol 2012; 46: 421–427.

29.

Purroy F

Jiménez

Caballero

Gorospe

, et al Prediction of early stroke recurrence in transient ischemic attack patients from the PROMAPA study: a comparison of prognostic risk scores. Cerebrovasc Dis 2012; 33: 182–189.

30.

Denis

Shedid

Shehadeh

, et al Cervical spondylosis: a rare and curable cause of vertebrobasilar insufficiency. Eur Spine J 2014; 23 Suppl 2: 206–213.

31.

Jacob

Tanislav

Kostev

Osteoarthritis and incidence of stroke and transient ischemic attack in 320,136 adults followed in general practices in the United Kingdom. Joint Bone Spine 2021; 88: 105104.

32.

Wang

Bai

, et al Osteoarthritis and the risk of cardiovascular disease: a meta-analysis of observational studies. Sci Rep 2016; 6: 39672.

33.

Widdifield

Bernatsky

Paterson

, et al Accuracy of Canadian health administrative databases in identifying patients with rheumatoid arthritis: a validation study using the medical records of rheumatologists. Arthritis Care Res 2013; 65: 1582–1591.

34.

Edwards

Koehoorn

Boyd

, et al Diagnostic accuracy of transient ischemic attack from physician claims. Can J Neurol Sci 2017; 44: 397–403.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB