Risk of Subsequent Stroke Among Patients Receiving Outpatient vs Inpatient Care for Transient Ischemic Attack

Key Points Question Does the risk of subsequent stroke differ by care setting among patients with transient ischemic attack (TIA) or minor stroke? Findings In this systematic review and meta-analysis of 226 683 unique patients in 71 unique studies, patients cared for in a TIA clinic vs as inpatients had similar risks of subsequent stroke. Patients who were treated in emergency departments without further follow-up had a higher risk of subsequent stroke than those treated as inpatients or in TIA clinics. Meaning In this study, the risk of subsequent stroke among patients who received treatment in a TIA clinic was not higher than those who were hospitalized.


Trip database
P: "Transient ischemic attack" OR TIA OR "transient neurologic attack" OR TNA OR "mini stroke" OR "minor stroke" I: outpatient OR "out patient" OR "clinic" C: inpatient OR admit OR hospitalization O: "cerebral infarction" OR "brain ischemia" OR stroke OR "cerebrovascular accident" OR death)

CINAHL Complete
("Transient ischemic attack" OR TIA OR "transient neurologic attack" OR TNA OR "mini stroke" OR "minor stroke") AND (outpatient OR "out patient" OR "clinic" ) AND ("cerebral infarction" OR "brain ischemia" OR stroke OR "cerebrovascular accident" OR death) © 2022 Shahjouei S et  * Some cohorts in this study had low risk of bias while there was no information regarding other cohort(s).
ⴕ This value is represented for all included cohorts in the study (110 cohorts).Among the 64 cohorts with specified setting (i.e.TIA clinic, inpatient, and ED) 58 (90.6%) cohorts had low and 6 (9.4%) cohorts had moderate overall risk of bias.

eFigure 4 .
Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 2 Days eFigure 5. Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 7 Days eFigure 6. Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 30 Days eFigure 7. Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 90 Days eReferences.

eFigure 5 .
Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 7 Days Sensitivity analysis included the prospective cohort of patients recruited after 2000.The risk estimate for inpatients was considered as the reference line.* indicates the Stroke Unit, ** indicates the Observation Unit.

eFigure 6 .
Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 30 Days Sensitivity analysis included the prospective cohort of patients recruited after 2000.The risk estimate for inpatients was considered as the reference line.* indicates the Stroke Unit, ** indicates the Observation Unit.

eFigure 7 .
Sensitivity Analysis: Risk of Subsequent Ischemic Stroke Within 90 Days Sensitivity analysis included the prospective cohort of patients recruited after 2000.The risk estimate for inpatients was considered as the reference line.* indicates the Stroke Unit, ** indicates the Observation Unit.
eAppendix.Search Protocols eTable 1. Mixed-Effect Models Considering Different Possible Moderators eTable 2. Mixed-Effect Model Considering ABCD 2 Scores eTable 3. Excluded Studies eTable 4. Heterogeneity and Risk of Stroke Assessment Based on Different Estimators eTable 5. Comparison of Risk Estimates eTable 6. Publication Bias Assessment eTable 7. Risk-of-Bias Assessment Based on ROBINS-E Tool eFigure 1. Funnel Plots eFigure 2. Risk of Subsequent Ischemic Stroke Within 2 Days eFigure 3. Risk of Subsequent Ischemic Stroke Within 30 Days Statement 2: Care Setting -Format search as Part 1 or Part 2 or Part 3 Part 1 (exp ambulatory care facilities/) or (ambulatory care facilit$.mp.) or (clinic$.mp.) or (exp Ambulatory Care/) or (exp Outpatient Clinics, Hospital/) or (exp Outpatients/) or (exp Mobile Health Units/) or (exp Outpatients/) or (exp Outpatient Clinics, Hospital/) or (exp Emergency Service, Hospital/) or (exp Monitoring, Ambulatory/) or (exp Self-Care Units/) or (exp Clinical Observation Units/) or (exp Secondary Prevention/) or ((initial or urgent or rapid or semi?urgent or emergency) adj3 (evaluation$ or assessment$ or care)) 4l. JAMA Network Open.All cohorts evaluated in different intervals were entered to the model.†Datawere sparce for outcome of TIA clinic cohort within 2 days There were inadequate data for running a model for outcome of patients treated in TIA clinics within two days of index TIA.DL, DerSimonian and Laird.QM, indicates test statistic for the omnibus test of coefficients; df, degree of freedom; K, number of outcomes included in the model fitting; τ2, estimated amount of residual heterogeneity; I2 statistic, residual heterogeneity / unaccounted variability; H2, unaccounted variability / sampling variability; R2, amount of heterogeneity accounted.4.Heterogeneity and Risk of Stroke Assessment Based on Different Estimators DerSimonian and Laird estimator; HS, Hunter and Schmidt estimator; Hedges (HE) ML, Maximum Likelihood estimator; REML, Restricted Maximum Likelihood estimator; SJ, Sidik and Jonkman estimator; EB, empirical Bayes, estimator; Q, Cochran's Q test (χ2 test for heterogeneity); p, p-value for Q; I2, I2 statistic; Risk, risk of subsequent stroke following index event.5. Comparison of Risk Estimates Comparisons were conducted under restricted maximum likelihood (REML) estimator.Sensitivity Analysis included the prospective cohorts recruited after 2000.QE indicates test statistics for the tests of heterogeneity; QEP, p-values for the tests of heterogeneity; I2, I2 statistic.
eTable 1. Mixed-Effect Models Considering Different Possible Moderators Evaluation interval was considered within 2, 7, 30, and 90 days; Study design was defined as prospective vs. retrospective; Patient recruitment was defined a and 2007, and after 2007; ABCD2 score was defined as percentage of patients in each cohort who had score of <4 vs. ≥4; Setting of Management was define and unspecified.Restricted Maximum Likelihood (REML) was considered as estimator of all models.©2022Shahjouei S et al.JAMA Network Open.* © 2022 Shahjouei S et al.JAMA Network Open.eTable © 2022 Shahjouei S et al.JAMA Network Open.© 2022 Shahjouei S et al.JAMA Network Open.eTable