Between January 1st, 2019 and February 19th, 2020, a total of 268 individuals who visited the emergency department presented with stroke symptoms within six hours after a final diagnosis of ischemic stroke. Among these patients, about two-thirds were males. The mean age was 69 years old. Their initial National Institutes of Health Stroke Scale (NIHSS) score upon admission was 5 (IQR: 3–12). A significant proportion of patients had a history of hypertension, accounting for more than half of cases (61%), while over 25% of patients were diagnosed with diabetes mellitus. Notably, 49 (18.3%) patients had a prior medical history of stroke or transient ischemic attack. Among stroke subtypes, large artery atherosclerosis accounted for the highest at approximately one-third, followed by cardioembolism (26.9%) and small vessel occlusion (19.0%). In terms of treatment, about one-third and 20% received intravenous thrombolysis and endovascular thrombectomy, respectively [Table 1].
Table 1
Baseline characteristics of the study population (N = 268)
Variables | Value |
Age, mean ± SD | 68.6 ± 12.3 |
Sex, male (%) | 179 (66.8) |
Premorbid mRS, median (IQR) | 0 (0–0) |
Initial NIHSS score, median (IQR) | 5 (3–12) |
Comorbidities, N (%) |
Hypertension | 162 (60.5) |
Diabetes | 77 (28.7) |
Hyperlipidemia | 33 (12.3) |
Atrial fibrillation | 34 (12.7) |
Cancer | 28 (10.5) |
Smoking | 39 (14.6) |
Coronary heart disease | 19 (7.1) |
Stroke or TIA | 49 (18.3) |
Stroke subtype, N (%) |
Large artery atherosclerosis | 84 (31.3) |
Small vessel occlusion | 51 (19.0) |
Cardioembolism | 72 (26.9) |
Other-determined | 15 (5.6) |
Undetermined | 46 (17.2) |
Hyperacute reperfusion treatment, N (%) |
Intravenous thrombolysis | 93 (34.7) |
Endovascular thrombectomy | 54 (20.2) |
SD, standard deviation; mRS, modified Rankin’s scale; NIHSS, National Institute of Health Stroke Scale; IQR, interquartile range; TIA, transient ischemia attack.
When looking into the trend between each quarter of the year and quality indicators, we observed a gradual decrease in the number of patients with a concomitant increase in new COVID-19 cases in the community over time [Figure 1A, Supplementary Table 1]. Additionally, increasing trends were noted for door-to-neurologist referral time, DIT, and DPT. However, other quality indicators such as DNT and discharge or 3-month mRS exhibited no differences [Figure 1. B-F, Supplementary Table 1].
In the multivariable analysis to determine effects of calendar date on quality indicators [Table 2], calendar date seemed to increase DIT and door-to-neurology-referral time in Model 2 after adjusting for age, sex, premorbid mRS, initial NIHSS, and onset-to-arrival time. Furthermore, DPT seemed to be increased after incorporating other covariates (Model 3).
Table 2
Multivariable analysis for effects of calendar date (per 30 days) on quality indicators
Quality indicator | Standardized Beta [95% CI] | P-value |
Model 1 |
Door to first image time | 0.193 [0.074–0.311] | 0.002 |
Door to neurologist referral time | 0.195 [0.077–0.313] | 0.001 |
Door to needle time | -0.047 [-0.256–0.162] | 0.656 |
Door to puncture time | 0.289 [0.022–0.555] | 0.034 |
Discharge mRS 0–2 | 0.0002 [-0.0010–0.0014] | 0.750 |
3-month mRS 0–2 | 0.0004 [-0.0008–0.0015] | 0.543 |
Model 2 |
Door to first image time | 0.178 [0.064– 0.292] | 0.002 |
Door to neurologist referral time | 0.195 [0.078–0.313] | 0.001 |
Door to needle time | -0.061 [-0.275–0.153] | 0.575 |
Door to puncture time | 0.279 [-0.017–0.575] | 0.064 |
Discharge mRS 0–2 | -0.0004 [-0.0019–0.0010] | 0.565 |
3-month mRS 0–2 | 0.0003 [-0.0011–0.0018] | 0.665 |
Model 3 |
Door to first image time | 0.201 [0.085–0.316] | < 0.001 |
Door to neurologist referral time | 0.203 [0.084–0.323] | < 0.001 |
Door to needle time | -0.082 [-0.300–0.136] | 0.456 |
Door to puncture time | 0.538 [0.187–0.890] | 0.004 |
Discharge mRS 0–2 | -0.0002 [-0.0018–0.0014] | 0.814 |
3-month mRS 0–2 | 0.0004 [-0.0012–0.0020] | 0.604 |
Model 1: unadjusted for covariates; |
Model 2: adjusted for covariates: age, sex, premorbid mRS, visit NIHSS, and onset to arrival time; |
Model 3: adjusted for covariates such as age, sex, pre mRS, visit NIHSS, onset to arrival time, history of hypertension, diabetes, dyslipidemia, atrial fibrillation, cancer, smoking, ischemic heart disease, previous stroke or transient ischemia attack, and stroke subtype.
After that, we divided patients into those who arrived in our emergency department before (n = 173) and after (n = 95) the change in in-hospital quarantine policy with mandatory COVID-19 screening. There were no differences in baseline characteristics between these groups except that the premorbid mRS was slightly higher after implementation of the mandatory COVID-19 screening [Supplement Table 2]. However, the proportion of patients treated with endovascular thrombectomy was much higher after the change in in-hospital quarantine policy (15.6% before mandatory COVID-19 screening vs. 28.4% after the mandatory COVID-19 screening). A delay in median DIT was observed comparing before and after the change in quarantine policy (11 minutes vs. 14 minutes). Although it did not reach the statistical significance threshold, median door-to-referral time (20 minutes vs. 23 minutes) and median DPT (137 minutes vs. 151.5 minutes) were also prolonged after the change in the mandatory COVID-19 screening policy. Despite these shifts in quality indicators, no substantial differences were noted in functional outcomes such as discharge mRS scores or 3 months' mRS scores [Table 3].
Table 3
Comparison of quality indicators before and after quarantine in-hospital quarantine policy change
Quality indicator | Before mandatory COVID-19 screening (N = 173) | After mandatory COVID-19 screening (N = 95) | P-value |
Door to first image time, minutes (median (IQR)) | 11 (7–18) | 14 (10–24) | < 0.01 |
Door to neurologist referral time, minutes (median (IQR)) | 20 (13–29) | 23 (16–35) | 0.08 |
Door to needle time, minutes (median (IQR)) | 51 (39–59) | 50 (45–58) | 0.70 |
Door to puncture time, minutes (median (IQR)) | 137 (120–170) | 158 (133–192) | 0.09 |
Discharge mRS, 0 to 2 (%) | 65 (37.6) | 39 (41.1) | 0.67 |
3-months mRS, 0 to 2 (%) | 96 (57.5) | 52 (56.5) | 0.99 |
mRS, modified Rankin’s scale; IQR, interquartile range.
After introducing information of whether the patient was admitted before or after implementing the mandatory COVID-19 screening as a variable in addition to previous multivariable models, the change in the quarantine policy seemed to increase the DIT even after adjusting for other covariates [Table 4].
Table 4
Multivariable analysis for effects of changes of in-hospital quarantine policy on quality indicators
Quality indicator | Estimate [95% CI] | P-value |
Model 1 |
Door to first image time | 0.410 [0.019–0.800] | 0.040 |
Door to neurologist referral time | 0.008 [-0.386–0.402] | 0.968 |
Door to needle time | 0.417 [-0.238–1.072] | 0.209 |
Door to puncture time | 0.225 [-0.523–0.973] | 0.549 |
Discharge mRS 0–2 | -0.547 [-1.398–0.279] | 0.199 |
3-month mRS 0–2 | -0.217 [-1.042–0.605] | 0.604 |
Model 2 |
Door to first image time | 0.404 [0.028–0.779] | 0.035 |
Door to neurologist referral time | 0.024 [-0.367–0.415] | 0.903 |
Door to needle time | 0.435 [-0.241–1.111] | 0.204 |
Door to puncture time | 0.254 [-0.5382–1.047] | 0.521 |
Discharge mRS | -0.605 [-1.664–0.427] | 0.254 |
3 months mRS | 0.024 [ -0.978–1.029] | 0.963 |
Model 3 |
Door to first image time | 0.399 [0.018–0.780] | 0.040 |
Door to neurologist referral time | -0.052 [-0.450–0.346] | 0.797 |
Door to needle time | 0.249 [-0.479–0.976] | 0.498 |
Door to puncture time | 0.062 [-0.825–0.950] | 0.888 |
Discharge mRS 0–2 | -0.701 [-1.840–0.405] | 0.219 |
3-month mRS 0–2 | -0.150 [-1.216–0.913] | 0.781 |
Model 1: adjusted for calendar date; |
Model 2: adjusted for covariates (age, sex, premorbid mRS, visit NIHSS, onset to arrival time, and calendar date); |
Model 3: adjusted for covariates such as age, sex, pre mRS, visit NIHSS, onset to arrival time, history of hypertension, diabetes, dyslipidemia, atrial fibrillation, cancer, smoking, ischemic heart disease, previous stroke or transient ischemia attack, stroke subtype, and calendar date.