Characteristics and Outcomes Among US Patients Hospitalized for Ischemic Stroke Before vs During the COVID-19 Pandemic

This cohort study uses data from the Vizient Clinical Data Base to examine hospital discharge rates and outcomes of hospitalizations as well as demographic factors among US patients with ischemic stroke before and during the COVID-19 pandemic.


Introduction
7][8] Although the burden of COVID-19 increased in the US, 9 additional data on rates of IS through the end of 2020 have not been published, to our knowledge.Previous research has suggested that patients with IS and comorbid COVID-19 have poor outcomes [10][11][12][13][14][15][16] ; however, to our knowledge, that research has not explored the impact of IS being identified at hospital admission vs during hospitalization with COVID-19. 2,17This study used a data set from 478 hospitals throughout the US to examine the clinical characteristics and outcomes among 5517 patients with IS and COVID-19 who were discharged in 2020 compared with those of patients with IS who were discharged in 2019.

Data Set
This retrospective cohort study included data from the Vizient Clinical Data Base (CDB), a health care analytics platform used by participating US hospitals for measuring clinical performance, costs, and outcomes. 18We included patients who received a diagnosis of IS (based on International Statistical Classification of Diseases, Tenth Revision, Clinical Modification [ICD-10-CM] codes I63.x and H34.1 19 ) who were discharged from 478 hospitals that reported complete patient-level data from January 1, 2019, to December 31, 2020.Patients were eligible for inclusion if they were admitted to the hospital on a nonelective basis and were not receiving hospice care at the time of admission.This study was approved by the University of Utah with a waiver of informed consent because of the study's retrospective design and use of deidentified data.The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

Study Cohorts
All patients with IS who were discharged from the hospital were included in monthly counts.For additional analyses, we excluded patients discharged between January 1 and March 31, 2020, because we could not be certain of their COVID-19 status.We created 3 cohorts of patients with IS: (1) patients with IS who were discharged in 2019 (control group), (2) patients with IS without COVID-19 who were discharged between April 1 and December 31, 2020 (non-COVID-19 group), and (3) patients with IS and laboratory-confirmed COVID-19 who were discharged between April and December 2020 (COVID-19 group).We identified patients with comorbid COVID-19 using ICD- 10

Outcomes and Variables
The primary outcome was monthly discharges with ischemic stroke.Use of intravenous (IV) alteplase and endovascular thrombectomy (EVT) for the treatment of acute IS was identified using codes from the International Classification of Diseases, Tenth Revision, Procedure Coding System (ICD-10-PCS) (codes given in eTable 1 in the Supplement).We also abstracted patient scores from the National Institutes of Health Stroke Scale (NIHSS) (score ranges from 0 to 42, with higher scores indicating greater functional impairment), which was coded as R29.7 in the ICD-10-CM in the first available NIHSS data set. 21Because NIHSS scores were not available for all patients, analyses including NIHSS data had smaller samples.

Statistical Analysis
We developed graphical representations of the monthly hospital discharges of patients with IS and and pulmonary embolus (1.9% vs 2.4%); and more likely to be intubated (11.3% vs 12.3%) and experience both favorable discharge (65.9% vs 69.0%) and in-hospital death (6.4% vs 6.8%).
However, the magnitude of the differences between these cohorts was relatively small.
The differences in demographic characteristics between the control group and the COVID- The ORs in the mixed-effects model for baseline demographic characteristics and comorbidities among the COVID-19 group are shown in Figure 2. Older patients and those who were male, Hispanic, or Asian were more likely to die in the hospital.The presence of atrial fibrillation was also associated with increases in the likelihood of in-hospital death, and smoking and the presence of dyslipidemia and hypertension were associated with a reduced likelihood of in-hospital death.Similar    findings in the opposite direction were observed for favorable discharge, with older, male, and Hispanic patients less likely to have favorable discharge; however, the differences were not significant with the exception of older age and male sex.

Discussion
In this retrospective cohort study, after an initial decrease in IS hospitalizations during the beginning of the COVID-19 pandemic in February 2020, hospitalizations returned to prepandemic levels by July 2020 and stayed at those levels for the remainder of 2020.The reduction in hospitalizations for IS at the beginning of the pandemic may have been associated with patients not seeking medical care owing to fear of acquiring COVID-19.The increase in hospital discharges in the later months of 2020 may have been associated with awareness efforts and/or an easing of public fears regarding hospital care. 4,24 adjusted models, patients in the COVID-19 group compared with those in the control group had a 5-fold higher risk of in-hospital death and were 3 times less likely to have a favorable hospital discharge.These associations remained significant in a sensitivity analysis adjusted for stroke Although the incidence of IS and comorbid COVID-19 from April to December 2020 (when it could be accurately measured) was 4.7% among all discharged patients with IS, the association of COVID-19 with in-hospital mortality was observed throughout 2020.In 2019, the in-hospital mortality among discharged patients with IS was 6.4%, and in all of 2020, the in-hospital mortality was 7.6% (P < .001).Of note, compared with the control group, patients with IS who did not have laboratory-confirmed COVID-19 and who were discharged between April and December 2020 also had a higher rate of in-hospital death (6.8%) and more complications, such as acute coronary syndrome, pulmonary embolus, and intubation, suggesting that although COVID-19 was not confirmed in these patients, a small percentage of them may have had infection that was not diagnosed or coded as such.
We also found that discharged patients with IS and comorbid COVID-19 were younger and more likely to be male, Black, and Hispanic or have obesity or diabetes but were less likely to smoke or have hypertension or dyslipidemia compared with patients in the control group.These findings are consistent with previous data on COVID-19 demographic factors and disparities. 12,25The higher incidence of COVID-19 in Black and Hispanic populations may have been associated with socioeconomic and health disparities that, in turn, may be associated with increases in the risk of developing COVID-19 and IS. 26 A definitive explanation for the increased risk of COVID-19 among male patients and those with obesity or diabetes has not been found.Whether the higher risk of IS was associated with vascular risk factors that increase susceptibility to more severe COVID-19 or whether the factors were associated with independent or synergistic prothrombotic consequences in patients with COVID-19 is unknown.
Among discharged patients with IS and comorbid COVID-19, we found that the factors associated with in-hospital death were older age; male sex; Asian, Hispanic, and other race/ethnicity; and the presence of atrial fibrillation.Discharged patients with IS and comorbid COVID-19 were also more likely to have medical complications, such as acute coronary syndrome, pulmonary embolus, or intubation, which is consistent with previous findings 12,13

Limitations
This study has limitations.The Vizient CDB is not representative of all inpatient discharges in the US, and identification of IS cases through billing codes has the potential for misclassification bias.
However, the sample was large, and the data sampling and extraction methods, including those used for hospitals, were consistent across time.We could not capture the severity of COVID-19, which limited our ability to make definitive associations.Apart from the dichotomy of presence or absence of IS at the time of hospital admission, we did not know when the IS event occurred during the course of hospitalization, which prevented us from fully exploring the association between COVID-19 and IS.In addition, the differences between patients with IS in 2019 and patients with IS without COVID-19 in 2020 suggest that some of the patients in the non-COVID-19 group may have had COVID-19 but were not documented as such.Evaluation of longitudinal data with additional variables over subsequent time frames is needed to confirm these findings.

Conclusions
This cohort study found that after an initial decrease in IS discharges during the beginning of the COVID-19 pandemic in February 2020, discharge counts returned to prepandemic levels by July 2020 and stayed at those levels for the remainder of 2020.In adjusted models, discharged patients with IS and comorbid COVID-19 in 2020 compared with discharged patients with IS in 2019 had a 5-fold higher risk of in-hospital death and were 3 times less likely to have a favorable hospital discharge.We also found that discharged patients with IS and comorbid COVID-19 in 2020 compared with discharged patients with IS in 2019 had more severe stroke and were younger; more likely to be male, Black, and Hispanic; and more likely to have obesity and diabetes but less likely to smoke or have hypertension or dyslipidemia.Among discharged patients with IS and comorbid COVID-

Figure 2 .Smoking
Figure 2. In-Hospital Death and Favorable Hospital Discharge Among Patients With Ischemic Stroke and Comorbid COVID-19

-CM JAMA Network Open | Neurology Hospitalizations
for Ischemic Stroke Before vs During the COVID-19 Pandemic CoV-2.20TheICD-10-CM U07.1 code was released in late March 2020 and was widely used by April.Although there were other approaches that could have been used to identify COVID-19 cases, these approaches risked misclassification bias because of the absence of laboratory confirmation of SARS-CoV-2 infection.
JAMA Network Open.2021;4(5):e2110314.doi:10.1001/jamanetworkopen.2021.10314(Reprinted) May 17, 2021 2/12 Downloaded From: https://jamanetwork.com/ on 09/26/2023 code U07.1, which is reserved for cases that have laboratory confirmation of infection with SARS- 16e secondary outcomes included in-hospital death and favorable discharge, defined as discharge to the home or an acute rehabilitation facility.Medical comorbidities were coded as absent or present based on ICD-10-CM codes.Because of data restrictions in the Vizient CDB, patient age could not be reported; thus, we used age categories (18-50 years, 51-64 years, 65-74 years, 75-79 years, and Ն80 years).Race categories (White, Black, Asian, and other or unreported) were self-reported to hospital staff and were independent of Hispanic ethnicity, which was handled as a distinct category.We included race/ ethnicity as a variable of interest in the study based on a previous study indicating racial/ethnic disparities among patients with COVID-19.16

Table 1 .
in the Supplement).The monthly count of discharged patients with COVID-19 is shown in the eFigure in the Supplement, with the highest monthly total occurring in December 2020, the last month of the data collection period.The rate of in-hospital death was 6.4% in 2019 and 7.6% in 2020 COVID-19 group, which comprised 111 418 of 116 935 patients with IS (95.3%; 51.9% male; 62.8% White; 24.6% aged Ն80 years) discharged between April and December 2020 who did not have COVID-19; and (3) the COVID-19 group, which comprised 5517 of 116 935 patients with IS (4.7%; 58.0% male; 42.5% White; 21.3% aged Ն80 years) discharged between April and December 2020 who had laboratory-confirmed COVID-19.The demographic characteristics and outcomes in the 3 cohorts are shown in Compared with the control group, patients in the non-COVID-19 group were younger (age Ն75 years: 38.2% vs 36.3%);more likely to be male (50.7% vs 51.9%) and smoke (16.0%vs 17.2%); more likely to have obesity (16.2% vs 18.4%), dyslipidemia (61.2% vs 63.2%), intracerebral hemorrhage (6.1% vs 6.9%), the number of IV alteplase and EVT interventions received.For the control, COVID-19, and non-COVID-19 cohorts, we reported descriptive statistics and results of tests for differences between the control group vs the non-COVID-19 group and the non-COVID-19 group vs the COVID-19 group, with a 2-sided t test used for interval variables and a χ 2 test used for binary variables.We used mixed-effects logistic regression models to measure (P < .001).Among 324 013 total patients, 41 166 discharged between January and March 2020 were excluded from the analysis because they had unreliable data on COVID-19 status.The remaining 282 847 patients were allocated to 3 cohorts: (1) the control group, which comprised 165 912 patients Figure 1.Monthly Hospital Discharges Among Patients With Ischemic Stroke (IS) A Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec B Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec EVT IV alteplase EVT indicates endovascular thrombectomy; IV, intravenous.JAMA Network Open | Neurology Hospitalizations for Ischemic Stroke Before vs During the COVID-19 Pandemic JAMA Network Open.2021;4(5):e2110314.doi:10.1001/jamanetworkopen.2021.10314(Reprinted) May 17, 2021 4/12 Downloaded From: https://jamanetwork.com/ on 09/26/2023 with IS (50.7% male; 63.4% White; 26.3% aged Ն80 years) discharged in 2019; (2) the non-

Table 1 .
Baseline Demographic Characteristics and Outcomes Among Patients With Ischemic Stroke (continued) h A total of 186 243 patients had an NIHSS score available (score ranges from 0 to 42, with higher scores indicating greater functional impairment).iLength of ICU stay restricted to patients who spent more than 24 hours in the ICU.

Table 2 .
Mixed-Effects Logistic Regression Analysis of Outcomes a Adjusted for patient age, sex, and race/ethnicity; presence of diabetes, congestive heart failure, obesity, and smoking; Elixhauser Comorbidity Index score; National Institutes of Health Stroke Scale score; and receipt of IV alteplase and EVT.This model included 2314 patients with IS and COVID-19 discharged between April and December 2020 compared with the control group of 92 751 patients with IS discharged in 2019.
Abbreviations: EVT, endovascular thrombectomy; IS, ischemic stroke; IV, intravenous; OR, odds ratio.aAnalysis of 5517 patients with IS and COVID-19 discharged between April and December 2020 compared with control group of 165 912 patients with IS discharged in 2019.bAdjusted for patient age, sex, and race/ethnicity; presence of diabetes, congestive heart failure, obesity, and smoking; and Elixhauser Comorbidity Index score.c

Table 3 .
Patients With Diagnosis of IS at Hospital Discharge Stratified by Presence of IS Diagnosis at Admission a The control group comprised patients with IS who were discharged in 2019.b The non-COVID-19 group comprised patients with IS who did not have COVID-19 and were discharged between April and December 2020.c P values for non-COVID-19 group vs control group.d The COVID-19 group comprised patients with IS who had COVID-19 and were discharged between April and December 2020.e P values for COVID-19 group vs control group.
and interventions.The rate of in-hospital mortality was particularly high (46.0%)among patients in the COVID-19 group who did not have IS at admission.This phenomenon, which could reflect the development of IS during hospitalization or the delayed recognition of simultaneous IS after initial presentation with respiratory symptoms, was more than 3 times as common among patients in the COVID-19 group compared with those in the control group.The higher mortality among those without IS at admission may be associated with the severity of COVID-19 (which may have masked neurologic symptoms and led to deferral of neuroimaging evaluations), a higher prevalence of in-hospital thrombotic complications among those with COVID-19, or ineligibility for acute interventions among inpatients with IS.
Hospitalizations for Ischemic Stroke Before vs During the COVID-19 Pandemic In the control group, the mean (SD) NIHSS score was 7.6 (7.9), and in discharged patients with IS and comorbid COVID-19, the mean (SD) score was 11.6(9.6);thisfinding suggests that patients with IS and comorbid COVID-19 may experience more severe IS, that COVID-19 may be identified more frequently in patients with severe IS, or that patients with COVID-19 and minor stroke do not seek medical care.In this cohort study, 65.9% of patients with COVID-19-associated IS presented with IS at admission, meaning that 34.1% of patients did not have IS at admission, which may reflect the inability to recognize IS symptoms in these patients or the occurrence of IS events during hospitalization as a thrombotic complication of COVID-19.Among patients in the control and non-COVID-19 groups, 90.5% and 90.1%, respectively, had IS at admission.Although the rate of in-hospital death was higher among patients in the control group without IS at admission than among those with IS at admission, the comparable rates among patients with COVID-19 were high.For example, the in-hospital death rate among discharged patients without IS at admission in 2019 was 21.5%, whereas it was 46.0% among discharged patients with COVID-19 without IS at admission in 2020.Patients with COVID-19 who did not have IS at admission were also less likely to receive EVT or have favorable discharge, which may represent an opportunity to improve the care of patients with COVID-19 through increased monitoring for IS while in the hospital and/or through strategies to prevent IS.
19, the factors associated with in-hospital death were older age, male sex, Black or Hispanic race/ethnicity, and atrial fibrillation.The rate of in-hospital death was highest among patients with COVID-19 who did not have IS at admission, which may represent an opportunity to improve the care of patients with COVID-19.Hospitalizations for Ischemic Stroke Before vs During the COVID-19 Pandemic JAMA Network Open.2021;4(5):e2110314.doi:10.1001/jamanetworkopen.2021.10314(Reprinted) May 17, 2021 10/12 Downloaded From: https://jamanetwork.com/ on 09/26/2023