Abstract
Background:
Evaluation and hospitalization rates after a transient ischemic attack (TIA)-like presentation vary widely in clinical practice. This study aimed to examine variations in care settings at initial TIA diagnosis in the United States.
Methods:
We retrospectively analyzed an adult cohort with a first TIA principal diagnosis between January 1, 2015, and December 31, 2019, from TriNetX Diamond Network. Care settings at TIA diagnosis were defined as hospital care (including inpatient services and observation unit care without admission) and outpatient care (including any outpatient or emergency department visits). We estimated the distribution of care settings at TIA diagnosis and examined the associations of the hospital care setting with baseline age, sex, race, ethnicity, region, and stroke history.
Results:
Among the 554,315 included patients, 38.8% received hospital care at their initial TIA diagnosis. A higher percentage of hospital care was observed in the age group of 50-64 years (40.3%), Black (46%), Hispanic (41.2%), South (40.9%) and Midwest (43.0%) Regions, and with a history of stroke (39.6%). Multivariable logistic regression consistently showed patients who were aged 50-64 years (Odds Ratio=1.09, 95% CI: [1.07, 1.11]), Black (1.28, [1.24, 1.32]), Hispanic (1.13, [1.09, 1.18]), from South (1.20, [1.18, 1.22]) and Midwest Region (1.33, [1.30, 1.35]), and had a history of stroke (1.02, [1.00, 1.04]) to more likely receive hospital care.
Conclusions:
Although there are TIA care disparities based on demographics, most patients with initial TIA received acute care in outpatient settings. It is imperative to ensure primary providers can risk-stratify TIA patients and provide rapid and proper management.
Keywords: transient ischemic attack, care setting, disparities, electronic health records
INTRODUCTION
Transient ischemic attack (TIA) precedes about 20% of stroke events.1–3 Appropriate and timely preventive strategies should be obtained to prevent subsequent stroke among patients with TIA. Two decades ago, the EXisting PREventive Strategies for Stroke (EXPRESS) study4 proposed referral for acute management of patients with transient ischemic attack.
Currently, there are several inpatient, outpatient, and hybrid care pathways for the evaluation of patients with TIA worldwide.5–8 In the United States (US), The American Heart / American Stroke Association (AHA/ASA) guideline9 recommends an urgent evaluation and hospitalization of TIA patients if they present within 72 hours and indicate a high risk of early recurrence. However, evaluation and hospitalization rates after a TIA-like presentation vary widely among practitioners, hospitals, and regions.10–12 In addition, the care is influenced by regions, settings, sociodemographic characteristics, and insurance status.13–15
Understanding the current TIA care pathways and re-examining variations in the care is the first step to addressing TIA care disparities in the US. This study aims to systematically examine TIA care settings and the associated disparities by analyzing real-world multi-center health records data from the TriNetX Network, a large health research network. The results of this study will provide insights for the stroke community, facilitating a better understanding of disparities in TIA care across the US and helping identify opportunities to improve care.
METHODS
Data Source
Deidentified patient-level electronic health records (EHR) data were accessed through the TriNetX Network, a global health research network that combines longitudinal clinical data from over 120 healthcare organizations (HCO).16,17 TriNetX is certified to the ISO 27001:2013 standard and maintains an Information Security Management System (ISMS) to ensure the protection of the healthcare data to which it has access and to meet the requirements of the HIPAA Security Rule. The clinical data included patient demographics (sex, year of birth, race, ethnicity, geographic location by 3-digit zip code), longitudinal medical records of diagnoses (with International Classification of Diseases [ICD]-9 or ICD-10 codes), and procedures (with Current Procedural Terminology [CPT]/Healthcare Common Procedure Coding System [HCPCS] codes) at each encounter. TriNetX data have been widely used in clinical and health service research literature,17–21 including applications in the clinical areas of stroke, cerebrovascular diseases, and neurological conditions.19,22,23 Specifically, in this study, we used the TriNetX Diamond Network with a coverage of over 200 million population in the US. The primary sources of the data are third-party longitudinal data from ambulatory and primary care EHR and medical and pharmacy claims from claims clearinghouses. This study was exempted from Institutional Review Board approval because it used only de-identified patient records and did not involve the collection, use, or transmittal of individually identifiable data.16
Study Cohort
We defined the index event of each patient as the first encounter with TIA principal diagnosis (ICD-9: 435.0-435.9 or ICD-10: G45.0-G45.9). The study cohort was selected based on the following inclusion and exclusion criteria. Patients were included if they (1) had at least one TIA diagnosis, with the index event (first TIA diagnosis) between January 1, 2015, and December 31, 2019; (2) TIA was designated as the principal diagnosis at the index event; (3) aged between 18-89 years old at the index event; (4) had medical records (diagnoses or procedures) and well-defined care settings (see further details for the care setting definitions in the next section) within ±2 days of the index event, and (5) had known sex variable and were located in the US. We used year range of 2015-2019 with relatively stable sample sizes in each year from the TriNetX data, which was also before the COVID-19 pandemic to avoid potential biases in the analysis results due to the pandemic. The upper limit of age was chosen since any patient older than 89 years was suppressed in the TriNetX dataset during the de-identification procedure. If a patient had procedure codes for hospital discharges, the procedures, and the diagnoses after the discharge records were considered as an encounter separate from the one with the index event.
Among the patients meeting the above inclusion criteria, we excluded those who had stroke diagnosis (ICD-9: 433.01-433.91, ICD-10: I63.0-I63.9) within −7 to +2 days of the index event, or between +2 and +7 days after the index event if the diagnosis was in the same encounter of the index event. This ensured that patients with a stroke or an initial working diagnosis of TIA with different follow-up image findings (e.g., a patient who was hospitalized with TIA, had a stroke on MRI, and was discharged with the stroke diagnosis) were excluded. The time windows, −7 to +2 days and +2 to +7 days, specified in the exclusion criteria were chosen since we observed from preliminary data analysis that the majority of patients received a co-diagnosis of stroke within this time frame (Figure S1), which was also in alignment with the clinician’s practical experiences. A comparison of included and excluded patients is provided in Table S1.
Care Settings as the Outcome Measure
We considered four main care settings involved in the patient’s encounter related to TIA: emergency department (ED) visits, inpatient (IP) services (i.e., hospital admission), outpatient (OP) visits, and observation unit (OU) care. We used CPT codes for different types of services under the section of “Evaluation and Management Services”24 (Table S2), to determine the care settings for the TIA patients. It is possible that one patient had multiple care settings within the ±2 days window of the index event. For instance, a patient may have had an OP visit followed by an ED visit and a hospital admission (IP).
The primary outcome was the distribution of five mutually exclusive groups of care settings at the index event of TIA diagnosis: (1) IP, which may or may not involve other care settings within the same encounter; (2) OU without hospital admission, (3) OP only visits, (4) ED only visits, and (5) both ED and OP visits on the same day. We refer to IP and OU as hospital care, and to OP and ED as outpatient care for brevity. We did not distinguish the sequence of multiple care settings within the same day given that the available data only had minimum time resolution as the level of days.
Statistical Analysis
The distribution of the five distinct groups of care settings at the index event was summarized for the overall cohort and by each subgroup of sex (male and female), race (White, Black or African American, Asian, and Unknown), ethnicity (Hispanic, Non-Hispanic, Unknown), and census region (Northeast, Midwest, South, West, Unknown or Others; derived based on 3-digit zipcode), respectively. We used a two-sample chi-square test to compare the distribution of care settings between patient subgroups. For the multivariable analysis, we applied logistic regression to examine the association between the patient’s characteristics and hospital care (i.e., IP or OU). Independent variables included patients’ sex, age at the index event, race, ethnicity, and geographic region. Odds Ratios (ORs) and 95% Confidence Intervals (CIs) were reported. To account for the possible interaction effects between multiple baseline characteristics, we performed the analyses for each subgroup by sex, race, ethnicity, and census region and stratified by all other patient characteristics.
Sensitivity Analysis
Considering differences in the timing and variability of reporting health records in clinical practice, we relaxed the time window of ±2 days for the index event in the base case to a wider window of −2 to 7 days in our sensitivity analysis. We compared results to the base case results to assess if a wider time window led to different distributions of care settings at the TIA index event.
RESULTS
Baseline Characteristics
A total of 554,315 patients with TIA principal diagnosis between January 1, 2015, and December 31, 2019, were included in this study (Figure S2). The baseline characteristics are summarized in Table 1. The average age of the study cohort was 66.6 (standard deviation: 13.9) years at the TIA diagnosis. Among the patients with known race (25.2%) and ethnicity (27.2%), 85.7% were White, 13.9% Black, and 9.2% Hispanic. Most patients were from the South (35.5%); 9.1% of patients had a history of stroke prior to the index event of TIA diagnosis.
Table 1.
Patient characteristics for the study cohort from the TriNetX Diamond Network.
| N (%) | |
|---|---|
| Overall cohort | 554,315 |
| Age at the index event (years) | |
| <50 | 67,300 (12.1%) |
| 50-64 | 146,341 (26.4%) |
| ≥65 | 340,674 (61.5%) |
| Mean (SD) | 66.6 (13.9) |
| Median [Min, Max] | 69.0 [18.0, 89.0] |
| Sex | |
| Female | 312,147 (56.3%) |
| Male | 242,168 (43.7%) |
| Race | |
| White | 119,722 (21.6%) |
| Black or African American | 19,143 (3.5%) |
| Asian | 1,065 (0.2%) |
| Unknown | 414,385 (74.8%) |
| Ethnicity | |
| Hispanic | 13,839 (2.5%) |
| Non-Hispanic | 136,859 (24.7%) |
| Unknown | 403,617 (72.8%) |
| Geographical region | |
| West | 66,698 (12%) |
| Midwest | 90,387 (16.3%) |
| South | 196,828 (35.5%) |
| Northeast | 98,196 (17.7%) |
| Unknown | 102,206 (18.4%) |
| History of Stroke | |
| Yes | 50,554 (9.1%) |
Abbreviations: SD, standard deviation; Min, minimum; Max, maximum.
Distribution of the Care Settings at TIA Diagnosis
The distribution of care settings at the TIA diagnosis are summarized in Table 2. The majority (61.2%) of the patients received outpatient care, with most patients being seen in the ED only setting (30.6%), followed by the OP only setting (27.1%). There were 38.8% of patients who received hospital care,, including 24.4% admitted to inpatient services (IP) and 14.4% provided care in OU without hospital admission. The stratified analyses showed disparities in the TIA care settings by different subgroups. Older patients tended to have a lower proportion of OP only visits (25.6% for ≥65 years vs. 30.7% for <50 years) and a higher percentage of IP (25.3% vs. 21.4%). Men showed a marginally higher proportion of ED only visits (31.2% vs. 30.2%) and IP (24.9% vs. 24.1%).
Table 2.
Distribution of the care settings in the overall study cohort and subgroup analyses.
| Hospital Care | Outpatient Care | ||||||
|---|---|---|---|---|---|---|---|
| IP | OU | Subtotal* | OP only | ED only | ED and OP | Subtotal* | |
| Overall study cohort | 135,456 (24.4%) | 79,873 (14.4%) | 215,329 (38.8%) | 150,005 (27.1%) | 169,871 (30.6%) | 19,110 (3.4%) | 338,986 (61.2%) |
| Age group | |||||||
| <50 | 14,426 (21.4%) | 11,298 (16.8%) | 25,724 (38.2%) | 20,666 (30.7%) | 17,836 (26.5%) | 3,074 (4.6%) | 41,576 (61.8%) |
| 50-64 | 34,704 (23.7%) | 24,250 (16.6%) | 58,954 (40.3%) | 42,032 (28.7%) | 39,810 (27.2%) | 5,545 (3.8%) | 87,387 (59.7%) |
| ≥65 | 86,326 (25.3%) | 44,325 (13%) | 130,651 (38.4%) | 87,307 (25.6%) | 112,225 (32.9%) | 10,491 (3.1%) | 210,023 (61.6%) |
| Sex | |||||||
| Female | 75,263 (24.1%) | 46,103 (14.8%) | 121,366 (38.9%) | 85,443 (27.4%) | 94,341 (30.2%) | 10,997 (3.5%) | 190,781 (61.1%) |
| Male | 60,193 (24.9%) | 33,770 (13.9%) | 93,963 (38.8%) | 64,562 (26.7%) | 75,530 (31.2%) | 8,113 (3.4%) | 148,205 (61.2%) |
| Race | |||||||
| White | 29,468 (24.6%) | 17,725 (14.8%) | 47,193 (39.4%) | 31,006 (25.9%) | 37,048 (30.9%) | 4,475 (3.7%) | 72,529 (60.6%) |
| Black or African American | 6,066 (31.7%) | 2,745 (14.3%) | 8,811 (46%) | 4,525 (23.6%) | 5,159 (26.9%) | 648 (3.4%) | 10,332 (54%) |
| Asian | 209 (19.6%) | 120 (11.3%) | 329 (30.9%) | 356 (33.4%) | 338 (31.7%) | 42 (3.9%) | 736 (69.1%) |
| Unknown | 99,713 (24.1%) | 59,283 (14.3%) | 158,996 (38.4%) | 114,118 (27.5%) | 127,326 (30.7%) | 13,945 (3.4%) | 255,389 (61.6%) |
| Ethnicity | |||||||
| Hispanic | 3,777 (27.3%) | 1,921 (13.9%) | 5,698 (41.2%) | 4,395 (31.8%) | 3,231 (23.3%) | 515 (3.7%) | 8,141 (58.8%) |
| Non-Hispanic | 34,782 (25.4%) | 19,833 (14.5%) | 54,615 (39.9%) | 35,522 (26%) | 41,727 (30.5%) | 4,995 (3.6%) | 82,244 (60.1%) |
| Unknown | 96,897 (24%) | 58,119 (14.4%) | 155,016 (38.4%) | 110,088 (27.3%) | 124,913 (30.9%) | 13,600 (3.4%) | 248,601 (61.6%) |
| Geographical region | |||||||
| Northeast | 25,150 (25.6%) | 10,624 (10.8%) | 35,774 (36.4%) | 39,814 (40.5%) | 19,274 (19.6%) | 3,334 (3.4%) | 62,422 (63.6%) |
| Midwest | 22,186 (24.5%) | 16,668 (18.4%) | 38,854 (43%) | 22,119 (24.5%) | 25,682 (28.4%) | 3,732 (4.1%) | 51,533 (57%) |
| South | 48,694 (24.7%) | 31,758 (16.1%) | 80,452 (40.9%) | 46,158 (23.5%) | 63,749 (32.4%) | 6,469 (3.3%) | 116,376 (59.1%) |
| West | 12,576 (18.9%) | 9,608 (14.4%) | 22,184 (33.3%) | 15,981 (24%) | 25,979 (39%) | 2,554 (3.8%) | 44,514 (66.7%) |
| Unknown | 26,850 (26.3%) | 11,215 (11%) | 38,065 (37.2%) | 25,933 (25.4%) | 35,187 (34.4%) | 3,021 (3%) | 64,141 (62.8%) |
| History of stroke | |||||||
| No | 121,591 (24.1%) | 73,712 (14.6%) | 195,303 (38.8%) | 136,383 (27.1%) | 154,433 (30.7%) | 17,642 (3.5%) | 308,458 (61.2%) |
| Yes | 13,865 (27.4%) | 6,161 (12.2%) | 20,026 (39.6%) | 13,622 (26.9%) | 15,438 (30.5%) | 1,468 (2.9%) | 30,528 (60.4%) |
Proportions of hospital care (or outpatient care) across different subpopulations were compared using the two-sample chi-square test. First subgroup within each category was used as the reference group for the comparison. All comparisons showed statistically significant differences with a small p-value<0.01 between the subgroups.
Abbreviations: OP, outpatient visits; ED, emergency department visits; IP, inpatient services (hospital admission); OU, observation unit.
We further estimated the distribution of care settings for each patient subgroup (Figure 1). Across subgroups by race, Black patients had the highest percentage of hospital care (IP and OU, 46.0%). Whites and Asians had a higher percentage of outpatient care (OP and ED) above 60%. Compared with non-Hispanic patients, Hispanic patients were more commonly seen in IP (27.3% vs. 25.4%) or OP only (31.8% vs. 26%) settings. For patients with known geographic information, those from the Midwest and South regions had the highest proportion of hospital care (IP and OU), 43.0% and 40.9%, respectively; patients from the West Region had the highest proportion of outpatient care (OP and ED, 66.7%). Moreover, patients with a history of stroke were more likely to be in the IP setting at TIA diagnosis compared with those without a stroke history. Our subgroup analysis by sex (Table S3–S4), race (Table S5–S7), and ethnicity (Table S8–S9) also showed consistent findings in the distribution of care settings.
Figure 1.
Distribution of the care settings stratified by patient characteristics.
Multivariable Logistics Regression for Care Setting
The multivariable logistic regression analysis examined the association of each patient’s characteristics with hospital care (i.e., IP or OU) at the patient’s TIA diagnosis (Table 3). We found that patients aged 50-64 years (OR=1.09, 95% Confidence Interval: [1.07, 1.11]), Black (OR= 1.28, [1.24, 1.32]), and Hispanic (OR=1.13, [1.09, 1.18]) were more likely to receive hospital care at TIA diagnosis, whereas Asian patients (OR=0.74, [0.65, 0.84]) were less likely to receive hospital care. No significant association with sex was observed. Compared with the Northeast region, patients living in the Midwest (OR=1.33, [1.30, 1.35]) or South (OR=1.20, [1.18, 1.22]) were more likely to present to the hospital, while those in the West (OR=0.88, [0.86, 0.90]) were more likely to present in outpatient settings. When limited to the outcome of IP setting only (Table S10), older age, Black, Hispanic, and having a history of stroke remained strong risk factors associated with the outcome. Analysis using the subpopulation that excluded patients with unknown values of baseline characteristics variables showed similar findings (Table S11). We further stratified the analysis by geographical region (Table S12) and found that being Black was significantly positively associated with hospital care across most regions, and Hispanic patients were significantly more likely to be admitted to the hospital in all regions except for the Northeast. In additional stratified analysis by sex, race, and ethnicity (Tables S13–15), we found a stronger positive association of hospital care with Black patients aged of 50-64 and with White or non-Hispanic patients older than 65 years. Black and Hispanic patients were consistently more likely to receive hospital care for either sex.
Table 3.
Associations of patient characteristics and hospital care at initial TIA diagnosis.
| Variable (reference group) | Odds Ratio (95% CI) | p-value |
|---|---|---|
| Sex (female) | ||
| Male | 1.00 (0.99, 1.01) | 0.99 |
| Age group (<50 years) | ||
| 50-64 years | 1.09 (1.07, 1.11) | <0.01 |
| ≥65 years | 1.02 (1.00, 1.03) | 0.06 |
| Race (White) | ||
| Black or African American | 1.28 (1.24, 1.32) | <0.01 |
| Asian and others | 0.74 (0.65, 0.84) | <0.01 |
| Unknown | 0.97 (0.95, 0.99) | <0.01 |
| Ethnicity (non-Hispanic) | ||
| Hispanic | 1.13 (1.09, 1.18) | <0.01 |
| Unknown | 1.01 (0.99, 1.03) | 0.28 |
| Geographical region (Northeast) | ||
| West | 0.88 (0.86, 0.90) | <0.01 |
| Midwest | 1.33 (1.30, 1.35) | <0.01 |
| South | 1.20 (1.18, 1.22) | <0.01 |
| Unknown | 1.07 (1.05, 1.09) | <0.01 |
| History of stroke (no) | ||
| Yes | 1.02 (1.00, 1.04) | 0.02 |
Abbreviations: CI, confidence interval.
Sensitivity Analysis
Results from the sensitivity analysis (Tables S16–17) showed consistent patterns similar to the base case. The proportion of patients utilizing different care settings and the direction and magnitude of the association between patient characteristics and the likelihood of hospitalization were similar across the wider time window, indicating the robustness of our findings. However, we observed moderate decreases in the percentage of patients receiving ED only, OP only, and OU without IP and moderate increases in the percentage of patients receiving IP and ED followed by OP visits. This observation suggests that when we expanded the time window for the index event, more patients were observed to receive a combination of healthcare services rather than services in just one setting. The decrease in IP only, ED only, and OU without IP could be attributed to the fact that patients could access a broader range of care settings within the wider time window.
DISCUSSIONS
Using the real-world data, our analysis showed wide variations in the care settings of TIA patients in clinical practice. We observed that 40% of patients were admitted to the inpatient services or observation units, and the rest received their initial care in outpatient settings (ED or OP). We found that age (≥50 years), Black race, Hispanic ethnicity, and South or Midwest regions were the significant factors positively associated with hospital care after a TIA.
There are only a few studies that have investigated the care setting and delivery for patients with TIA. The results of a study13 on all ED-treated TIA cases in eleven states using data from the 2002 Healthcare Cost and Utilization Project indicated that there were significant variations in diagnostic procedures and dispositions (i.e., admitted to hospital or released from ED). The investigators suggested that the care was influenced by regions, settings, sociodemographic characteristics, and insurance status. A survey of patients residing southeastern US (i.e., the Stroke Belt region) with a high incidence of stroke found that nearly 15% of people would first contact their primary care physician rather than accessing hospital care (e.g., emergency services or hospital visits) at an occurrence of a stroke.25 A higher proportion of patients managed without hospital care for TIA patients was expected, due to the transient nature of the symptoms and the lack of definitive criteria for hospitalization. In our more recent data, we found that on average 27.1% of patients went to an outpatient clinic only after a TIA event and that the proportion was the highest in the Northeast (40.5%) and the lowest in the South (23.5%).
Overall, our study indicates that the majority of TIA patients receive acute care in outpatient settings. This observation highlights the role of primary providers in initial TIA care. Although a considerable proportion of primary providers demonstrate a good understanding of TIA definition and management,26,27 a qualitative study underscored uncertainty regarding triage and management differences between high-risk and low-risk TIA patients.28 It is essential to ensure primary providers can risk-stratify TIA patients and have enough resources to provide rapid and proper management and referrals when needed.
While our results indicate that a majority of patients with initial TIA received acute care in outpatient settings, there were variations in TIA care by age, race, and region. Blacks had higher odds of receiving inpatient TIA care. There are several reasons for this observation. Black people in the US have higher rates of chronic conditions, such as diabetes29 and hypertension.30 Similar to the age factor, having chronic conditions can place the patient in a high-risk category and justify an admission. At the same time, Blacks have less access to outpatient settings and primary care physicians. Several studies have documented Black-White inequalities in the use of ambulatory care.31,32 In addition, how TIA patients are managed in practical settings could also be influenced by the severity of patients’ symptoms, as well as other provider-side factors, such as the availability of facilities, the availability of rapid evaluation, and the potential for following up with the patients, which were not directly available in our data for analysis and warrant further investigation in future research.
A strength of this study is that our analysis was based on a large, diverse, and geographically distributed patient population from the TriNetX Network, which significantly bolsters the reliability of our findings. The substantial sample size increases the statistical power of our analysis and enables us to examine a wide range of patient characteristics and care settings associated with TIA diagnosis. Our selection process in defining the study cohort and index events, combined with the sensitivity analyses to assess our findings’ robustness, further strengthens our study’s validity and its potential implications for healthcare systems and policymakers.
Our study has some limitations that must be considered when interpreting the results. First, our reliance on the quality and accuracy of the data provided by the TriNetX Network comes with challenges. As with any study using electronic health records, our findings may be affected by potential coding errors, misclassification, or incomplete data. This is particularly relevant when examining a condition like TIA, which can be challenging to diagnose in clinical settings due to its transient nature and the potential overlap of symptoms with other neurological disorders.33 The difficulty in accurately diagnosing TIA may lead to potential misclassification in our study cohort, which could affect the associations identified between patient characteristics and hospital admission patterns. Second, our data had a high missing rate in the race and ethnicity variables, posing challenges in assessing disparities between race and ethnicity subgroups. However, our study still included the largest sample size, to our best knowledge, for studying care settings of TIA. Besides, our extensive stratified analysis by demographics and regions showed consistent results, validating the main findings in our observed variations in TIA care settings by race and ethnicity. Nevertheless, results from the large and diverse patient population in this analysis provide valuable insights into the variations in TIA care from real-world data. Third, despite its extensive coverage of patient population and health plans in the US, the TriNetX data are not nationally representative in its geographical distribution. We could not assess the geographic distribution at a more granular level due to the lack of provider and healthcare facility information in the database. As a result, the observed variations by geographic regions did not necessarily represent the systemic disparities as a more general conclusion. Lastly, we conservatively selected the cohort by excluding patients with subsequent stroke shortly after the TIA, due to the concerns of not being able to separate the hospitalization because of stroke given the data limitation. This could potentially lead to a less severe patient cohort with TIA and thus a lower hospitalization rate. Future research is needed to further examine how stroke risk stratifications and the clinical risk factors of stroke are associated with different care settings of TIA.
CONCLUSIONS
Although significant variations in TIA care were observed from the real-world data by the factors of age, race, ethnicity, and region, the majority of patients with initial TIA receive acute care in outpatient settings. It is imperative to ensure primary providers can risk-stratify TIA patients and provide rapid and proper management.
Supplementary Material
ACKNOWLEDGMENTS
The authors acknowledge Dr. Durgesh Chaudhary for his valuable inputs and suggestions in analytical details and conceptualizing the analysis at the early stage of this study. Research reported in this publication was supported in part by the National Institute Of Neurological Disorders and Stroke of the National Institutes of Health under Award Number R01NS128986 and the National Center for Advancing Translational Sciences of the National Institutes of Health through Grant UL1 TR002014. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
DECLARATION OF INTEREST
None.
Contributor Information
Zixuan Feng, The Harold and Inge Marcus Department of Industrial and Manufacturing Engineering, The Pennsylvania State University.
Qiushi Chen, The Harold and Inge Marcus Department of Industrial and Manufacturing Engineering, The Pennsylvania State University.
Paul Griffin, The Harold and Inge Marcus Department of Industrial and Manufacturing Engineering, The Pennsylvania State University.
Jiang Li, Department of Molecular and Functional Genomics, Weis Center for Research, Geisinger Health System.
Vida Abedi, Department of Public Health Sciences, College of Medicine, The Pennsylvania State University.
Ramin Zand, Department of Neurology, College of Medicine, The Pennsylvania State University.
REFERENCES
- 1.Lavallee PC, Meseguer E, Abboud H, et al. A transient ischaemic attack clinic with round-the-clock access (SOS-TIA): feasibility and effects. Lancet Neurol. 2007;6(11):953–960. [DOI] [PubMed] [Google Scholar]
- 2.Prabhakaran S, Silver AJ, Warrior L, McClenathan B, Lee VH. Misdiagnosis of transient ischemic attacks in the emergency room. Cerebrovasc Dis. 2008;26(6):630–635. [DOI] [PubMed] [Google Scholar]
- 3.Hackam DG, Spence JD. Combining multiple approaches for the secondary prevention of vascular events after stroke: a quantitative modeling study. Stroke. 2007;38(6):1881–1885. [DOI] [PubMed] [Google Scholar]
- 4.Rothwell PM, Giles MF, Chandratheva A, et al. Effect of urgent treatment of transient ischaemic attack and minor stroke on early recurrent stroke (EXPRESS study): a prospective population-based sequential comparison. Lancet. 2007;370(9596):1432–1442. [DOI] [PubMed] [Google Scholar]
- 5.Evans BA, Ali K, Bulger J, et al. Referral pathways for patients with TIA avoiding hospital admission: a scoping review. BMJ Open. 2017;7(2):e013443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sanders LM, Srikanth VK, Jolley DJ, et al. Monash transient ischemic attack triaging treatment: safety of a transient ischemic attack mechanism-based outpatient model of care. Stroke. 2012;43(11):2936–2941. [DOI] [PubMed] [Google Scholar]
- 7.Jarhult SJ, Howell ML, Barnaure-Nachbar I, et al. Implementation of a Rapid, Protocol-based TIA Management Pathway. West J Emerg Med. 2018;19(2):216–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lim A, Singhal S, Lavallee P, et al. An International Report on the Adaptations of Rapid Transient Ischaemic Attack Pathways During the COVID-19 Pandemic. J Stroke Cerebrovasc Dis. 2020;29(11):105228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Easton JD, Saver JL, Albers GW, et al. 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(6):2276–2293. [DOI] [PubMed] [Google Scholar]
- 10.Edlow JA, Kim S, Pelletier AJ, Camargo CA Jr., National study on emergency department visits for transient ischemic attack, 1992-2001. Acad Emerg Med. 2006;13(6):666–672. [DOI] [PubMed] [Google Scholar]
- 11.Ramirez L, Kim-Tenser MA, Sanossian N, et al. Trends in Transient Ischemic Attack Hospitalizations in the United States. J Am Heart Assoc. 2016;5(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Khare S. Risk factors of transient ischemic attack: An overview. J Midlife Health. 2016;7(1):2–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Coben JH, Owens PL, Steiner CA, Crocco TJ. Hospital and demographic influences on the disposition of transient ischemic attack. Acad Emerg Med. 2008;15(2):171–176. [DOI] [PubMed] [Google Scholar]
- 14.Goldstein LB, Bian J, Samsa GP, Bonito AJ, Lux LJ, Matchar DB. New transient ischemic attack and stroke: outpatient management by primary care physicians. Arch Intern Med. 2000;160(19):2941–2946. [DOI] [PubMed] [Google Scholar]
- 15.Brown DL, Lisabeth LD, Garcia NM, Smith MA, Morgenstern LB. Emergency department evaluation of ischemic stroke and TIA: the BASIC Project. Neurology. 2004;63(12):2250–2254. [DOI] [PubMed] [Google Scholar]
- 16.TriNetX. https://trinetx.com/. Accessed October 15, 2022.
- 17.Palchuk MB, London JW, Perez-Rey D, et al. A global federated real-world data and analytics platform for research. JAMIA Open. 2023;6(2):ooad035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Harrison SL, Fazio-Eynullayeva E, Lane DA, Underhill P, Lip GYH. Comorbidities associated with mortality in 31,461 adults with COVID-19 in the United States: A federated electronic medical record analysis. PLoS Med. 2020;17(9):e1003321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Taquet M, Geddes JR, Husain M, Luciano S, Harrison PJ. 6-month neurological and psychiatric outcomes in 236 379 survivors of COVID-19: a retrospective cohort study using electronic health records. Lancet Psychiatry. 2021;8(5):416–427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Taquet M, Luciano S, Geddes JR, Harrison PJ. Bidirectional associations between COVID-19 and psychiatric disorder: retrospective cohort studies of 62 354 COVID-19 cases in the USA. Lancet Psychiatry. 2021;8(2):130–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tang K, Tiu BC, Wan G, et al. Pre-Existing Autoimmune Disease and Mortality in Patients Treated with Anti-PD-1 and Anti-PD-L1 Therapy. J Natl Cancer Inst. 2022;114(8):1200–1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Proietti R, Rivera-Caravaca JM, Lopez-Galvez R, et al. Clinical implications of different types of dementia in patients with atrial fibrillation: Insights from a global federated health network analysis. Clin Cardiol. 2023;46(6):656–662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Wang W, Wang CY, Wang SI, Wei JC. Long-term cardiovascular outcomes in COVID-19 survivors among non-vaccinated population: A retrospective cohort study from the TriNetX US collaborative networks. EClinicalMedicine. 2022;53:101619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Codify AAoPC. https://www.aapc.com/codes/cpt-codes-range/99091-99499/. Accessed Octobor 5, 2022.
- 25.Goldstein LB, Gradison M. Stroke-related knowledge among patients with access to medical care in the stroke belt. J Stroke Cerebrovasc Dis. 1999;8(5):349–352. [DOI] [PubMed] [Google Scholar]
- 26.Purroy F, Cruz Esteve I, Galindo Ortego MG, Marsal Mora JR, Oro M, Plana A. Survey of the knowledge and management of transient ischemic attacks among primary care physicians and nurses. Neurologia. 2011;26(4):200–207. [DOI] [PubMed] [Google Scholar]
- 27.Al Salman S, Al Sultan AA, Aldawood MA, Alradhi HK, AlMuhaish MA, Alsumaeel SA. Bridge to Better Care: Investigating Transient Ischemic Attack (TIA) Management Expertise Among Primary Healthcare Providers in Al-Ahsa, Saudi Arabia. Cureus. 2023;15(12):e50420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Edwards D, Turner GM, Virdee SK, Mant J. The role of the GP in managing suspected transient ischaemic attack: a qualitative study. BMC Fam Pract. 2019;20(1):67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Centers for Disease Control and Prevention. Summary Health Statistics: National Health Interview Survey, 2018. https://www.cdc.gov/nchs/nhis/shs/tables.htm. Published 2019. Accessed February, 2024.
- 30.Muntner P, Hardy ST, Fine LJ, et al. Trends in Blood Pressure Control Among US Adults With Hypertension, 1999-2000 to 2017-2018. JAMA. 2020;324(12):1190–1200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Dieleman JL, Chen C, Crosby SW, et al. US Health Care Spending by Race and Ethnicity, 2002-2016. JAMA. 2021;326(7):649–659. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Dickman SL, Gaffney A, McGregor A, et al. Trends in Health Care Use Among Black and White Persons in the US, 1963-2019. JAMA Netw Open. 2022;5(6):e2217383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Sadighi A, Stanciu A, Banciu M, et al. Rate and associated factors of transient ischemic attack misdiagnosis. eNeurologicalSci. 2019;15:100193. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.