Impact of the COVID-19 Pandemic on Emergency Admission for Stroke Patients: Time Series Study in Japan


 BackgroundHospitalization rates have reportedly reduced on account of measures such as the declaration of a COVID-19-related state of emergency. Appropriate treatment of stroke immediately after its onset contributes to improved survival, and delay in hospitalisation after onset affect stroke severity and mortality. This study aimed to determine the impact of the COVID-19 pandemic on emergency hospitalisation of stroke patients in Japan and the associated changes in stroke severity and mortality. MethodsThis was an observational study that used nationwide administrative data. The participants were selected from the data of hospitalised patients diagnosed with stroke. We cross-sectionally observed the background factors of patients during April and May 2020, when COVID-19-related state of emergency was declared, and the same period in 2019. We also modelled monthly trends in the numbers of emergency stroke admissions, of stroke admissions at each level of Japan Coma Scale (JCS), and of deaths within 24 h using interrupted time series regression (ITS). ResultsThere was no difference between the pre- and the COVID-19 pandemic periods in the pre-hospital baseline characteristics of patients. No significant change was seen through ITS in the number of emergency stroke admissions between the COVID-19 and the pre-COVID-19 pandemic periods (risk ratio [RR]=0.96, 95% confidence interval [95%CI]: 0.90–1.02, P=0.159). There was a significant difference in the JCS3 of impaired consciousness in emergency stroke patients, which was more severe during the COVID-19 pandemic than pre-COVID-19 pandemic (RR=1.16, 95%CI: 1.04–1.28, P=0.005). There was no change in the mortality rate of patients with COVID-19 compared to those without COVID-19, but there were significantly more deaths within 24 h of admission. (No deaths within 24 h: RR=0.97, 95%CI: 0.88–1.06, P=0.446; deaths within 24 h: RR=2.80, 95%CI: 2.40–3.27, P<0.001). ConclusionsOVID-19 infection prevalence increased the number of deaths and severity of illness within 24 h of hospitalisation in Japan. However, there was no association between the number of hospital admissions or deaths with baseline characteristics. These results suggest no problems in accepting hospitalised patients during the COVID-19 pandemic, but the delay in visiting the hospital after disease onset may have increased stroke severity.


Results
There was no difference between the pre-and the COVID-19 pandemic periods in the pre-hospital baseline characteristics of patients. No signi cant change was seen through ITS in the number of emergency stroke admissions between the COVID-19 and the pre-COVID-19 pandemic periods (risk ratio [RR]=0.96, 95% con dence interval [95%CI]: 0.90-1.02, P=0.159). There was a signi cant difference in the JCS3 of impaired consciousness in emergency stroke patients, which was more severe during the COVID-19 pandemic than pre-COVID-19 pandemic (RR=1.16, 95%CI: 1.04-1.28, P=0.005). There was no change in the mortality rate of patients with COVID-19 compared to those without COVID-19, but there were signi cantly more deaths within 24 h of admission. ( Conclusions OVID-19 infection prevalence increased the number of deaths and severity of illness within 24 h of hospitalisation in Japan. However, there was no association between the number of hospital admissions or deaths with baseline characteristics. These results suggest no problems in accepting hospitalised patients during the COVID-19 pandemic, but the delay in visiting the hospital after disease onset may have increased stroke severity.

Background
In December 2019, the rst case of the novel coronavirus disease , caused by a severe acute respiratory syndrome, was reported in Wuhan, China [1]. COVID-19 is characterised by rapid spread, and various preventive measures have been implemented worldwide to contain this disease. In Japan, as of 30 August 2020, 67,077 con rmed cases of COVID-19 have been reported [2].
With various countermeasures in place against the infection, hospitalization rates have decreased in overseas emergency hospitals, and a similar decrease in emergency hospital admissions across urban areas of Japan has been reported [3][4] [5].
Stroke is the second leading cause of death worldwide, the fourth leading cause of death in Japan, and is a common cause of emergency hospitalisation [6]. Appropriate treatment immediately after the onset of stroke contributes to the survival rate, and especially for treatments with limited indications, the time spent in the hospital from the onset of stroke is important [7][8] [9] [10]. Therefore, since the delay in hospitalisation from the onset of stroke may affect severity and mortality, it is important to understand the emergency hospitalisation status of stroke patients due to COVID-19 infection.
The Japanese government implemented various strategies to prevent the spread of COVID-19, and from 7 April, 2020 to 25 May, 2020 the government declared a state of emergency. The public was asked to stay home or work remotely, and large-scale events were cancelled or postponed. Additionally, to prevent the spread of infection at medical institutions or when travelling, the Japanese government instructed people to contact medical institutions before seeing a doctor if they had fever or other symptoms [11] [12]. At the time of the declaration of the state of emergency was just declared, the rate of infection slowed down temporarily, but after the declaration of the state of emergency, the rate of infection increased again [2].
Previous studies focusing on COVID-19 have shown a highly signi cant decrease in the number of visits to emergency cardiology departments during lockdown [3] [4]. The fear of a pandemic may have led to changes in patient behaviour, such as avoiding hospitalisation, even when symptoms are present. Studies regarding other past infectious disease pandemics have documented the largest outbreak of severe acute respiratory syndrome in 2003, resulting in limited access to care, lower overall hospitalisation rates, and a signi cant reduction in the use of healthcare services. There was also a decrease in the rate of hospitalisations and emergency room visits due to acute illness among patients with potentially serious conditions [13].
According to Japanese regional ambulance activation data, the proportion of emergency calls in the rst half of 2020 was lower than that during the same period in 2019 [14] [15]. Studies at the regional level have also shown a decrease in the number of emergency stroke admissions [5]. It is possible that the COVID-19 infection has reduced the rate of access to medical care for stroke patients in Japan, which may have an impact on mortality. The purpose of this study was to determine the impact of the COVID-19 pandemic on the admission of stroke patients to emergency hospitals and the resulting changes in stroke severity and mortality throughout Japan.

Study design
This observational study used nationwide administrative data from patients with a stroke who had been admitted to the emergency department between April 2018 and August 2020.

Patient selection
We extracted data from inpatients who were diagnosed with cerebral infarction (CI), cerebral haemorrhage (CH), and subarachnoid haemorrhage (SH) at the time of admission from April 2018 to August 2020. The International Classi cation of Diseases 10 diagnostic codes were used for the identi cation of CI (I63), CH (I61), and SH (I60). Patients aged 20 and older, who visited the emergency room, were included in our study. We excluded patients who stayed in the hospital for over 30 days to exclude non-acute patients. In this study, we used data up to August 2020, which was available until September 2020, and limited the target to patients hospitalised for 30 days or less.

Data source and variables
All data were obtained from the Diagnosis Procedure Combination, which is a national database of acute care inpatients in Japan. Diagnosis Procedure Combination data were used to evaluate the xed medical payment system. Details of the database are presented elsewhere [16] [17][18] [19]. The database contains data on hospital ID, patient ID, age, sex, weight and height, diseases, comorbidities according to ICD-10 codes, length of stay (LOS), Barthel index (BI), modi ed Ranking Scale (mRS) at discharge, admission and discharge destination, Japan Coma Scale (JCS) at admission as a consciousness scale, onset days (four categories: within 3 days, 4-7 days, 8 or more days, asymptomatic), and ambulance use.
We divided patients into the following six groups: 20-44, 45-54, 55-64, 65-74, 75-84, and ≥ 85 years, as the probability of most lifestyle-related diseases increases from approximately 40 years of age [20]. The BMI at admission was categorised based on the modi ed World Health Organization classi cation: < 18.4 kg/m 2 , 18.5-24.9 kg/m 2 , 25.0-29.9 kg/m 2 , 30.0-34.9 kg/m 2 , and > 35.0 kg/m 2 , and missing values [21]. The Charlson comorbidity index (CCI) was calculated using Quan's protocol [22] [23]. In the calculation of the CCI, patients' comorbidity conditions were classi ed into 17 categories. Each condition had a weighted score of 1, 2, 3, or 6, depending on the mortality risk associated with the case. Individual scores were summed up to obtain the total CCI scores. We divided the total CCI scores into the following four groups: 0, 1, 2, and 3 or more. The JCS is a unique Japanese method of assessing the level of consciousness, similar to the Glasgow scale. The JCS code of 0 denotes a lack of consciousness, while a code of 1-3 denotes the state of being awake without stimuli. A code of 10-30 signi es the ability to be roused by some stimuli, and 100-300 signi es the state of not being roused [24]. We divided the JCS into four groups according to the number of digits: 0, 1, 2, and 3. To determine the impact of the COVID-19 infection, the COVID-19 pandemic period was set to start in April, when the state of emergency was declared. Before and after the April-May period, the number of infected people in the country has been increasing [2].
The requirement for informed consent was waived because of the anonymous nature of the data. Study approval was obtained from the institutional review board of the Tokyo Medical and Dental University.

Statistical methods
Continuous variables were expressed as mean and standard error, and categorical variables were expressed as the number of subjects and percentages. Chi-square and t-test were used to analyse the differences in the characteristics of stroke patients between April and May 2019, and April and May 2020 for discrete and continuous variables. The standard mean difference (SMD) was also calculated to measure the effect sizes of both groups. The baseline characteristics were compared with an absolute standardised difference of ≤ 10% considered to denote negligible imbalances between the COVID-19 pandemic and the pre-COVID-19 pandemic periods [25].
We used an interrupted time series regression (ITS) to model monthly trends in the number of emergency stroke admissions, the number of stroke admissions at each level of the JCS (1, 2, and 3), and the number of deaths within 24 hours or otherwise [26] [27]. The Poisson distribution was used as the link function for all models. The month in which the emergency was declared was considered the point in time when the intervention took place, and all subsequent months were considered as post-intervention.
We hypothesised a level change model, such that the slope change is not considered in this model. All outcome variables used in the ITS were examined for a unit roof using the Dickey-Fuller test. We included xed effect terms for the year and season to control for seasonal trends. After checking and adjusting the model for seasonality and trends, autocorrelation and partial autocorrelation were plotted to check for autocorrelation of residuals. We adjusted each model using AIC and adopted one that best t the model. All tests were 2-tailed, and the threshold for signi cance was set at P < .05. Statistical analyses were performed using R statistical software (version 3.3.2; R Foundation for Statistical Computing, Vienna, Austria). Figure 1 shows the total number of eligible patients included in the ITS, the number of patients in April and May 2020 when the state of emergency was declared, and the number of patients during the same period in 2019. A total of 175,166 eligible patients were identi ed between April 2018 to August and August 2020. When the state of emergency was declared, there were 6,040 hospitalised patients in April and 6,444 in May. For the same period in 2019, 6,197 and 6,828 patients were admitted in April and May, respectively. Table 1 shows the pre-hospital baseline characteristics of patients admitted in April and May 2019 and 2020. There was no difference between the COVID-19 and the pre-COVID-19 pandemic periods in any of the pre-hospital baseline variables.    Although no unit root was found for all the variables used in the ITS, a harmonic adjustment term was inserted in the ITS model because of seasonal variation. No autocorrelation or partial autocorrelation was observed in the adjusted model.

Discussion
In this study, emergency stroke admissions during the COVID-19 pandemic were more severe, and the number of deaths within 24 h of admission increased. However, there was no change in the number of deaths, hospital admissions, or characteristics such as age, sex, and dependency. This suggests that the COVID-19 pandemic increased the severity of the disease at the time of emergency admission, although it did not change the patient population admitted to the emergency hospital after stroke. This study was the rst to focus on the changes in the severity of emergency stroke patients during the COVID-19 pandemic in Japan, with a particular focus on the urgency of stroke.
Further, we found that the number of patients with a high level of consciousness on admission, and deaths within 24 h of admission increased during the COVID-19 pandemic compared to before the COVID-19 pandemic. The severity of stroke patients at the time of emergency admission was also higher during the COVID-19 pandemic. A previous study on emergency admissions of heart failure patients reported a decrease in the severity of hospitalisation during the COVID-19 pandemic. This study showed that the number of hospitalisations decreased because heart failure patients avoided hospitalisation due to fear of COVID-19 infection [4]. The results of this study did not agree with the results of a previous study conducted on the severity of illness at the time of hospitalisation during the COVID-19 pandemic. This may be due to differences between heart failure and stroke in the conditions that lead to hospitalisation, and patients may have been in a condition that could have inhibited hospitalisation.
There was no difference in the onset date at the baseline. In a previous study of emergency hospital admissions during the COVID-19 pandemic in Japan, there was no difference in the time between stroke onset and transport to the emergency hospital [5]. However, another previous study that assessed the impact of the COVID-19 pandemic on the emergency system in Osaka city, Japan, found that, since April 2020, the emergency system in the city has been facing di culty regarding hospital acceptance of patients transported to hospitals for acute diseases [28]. It is possible that the COVID-19 pandemic may have caused a change in access to health care, but this cannot be examined in this study because there were no data within 3 days of stroke onset, and future studies are needed.
Furthermore, there was no difference in the number of emergency hospital admissions for stroke patients during the pandemic compared to before. This nding is inconsistent with the results of previous studies conducted in Japan. One previous study aimed to clarify the effect of the COVID-19 pandemic on a comprehensive stroke centre in Japan; the results showed that true emergencies appear to be associated with fewer emergency hospital visits and hospitalisations for stroke [29]. There are two reasons for this. First, the previous study was conducted at a single institution and in an area with a relatively high number of COVID-19 cases, which could have been strongly in uenced by the pandemic. In this study, we used data covering about half of the total population of Japan, and examined the impact of COVID-19 after the declaration emergency, when COVID-19 infections were increasing in Japan as a whole; thus, the impact of COVID-19 may have changed in different regions of Japan [30]. Second, the observation period in the previous study was for a few months and did not consider annual and seasonal variations. In the present study, the period used was approximately 2 years, and the analysis also considered seasonal variations.
Since the number of stroke patients is known to uctuate seasonally, it was necessary to consider the long-term period, which led to different results between this study and previous studies [31].
A previous study comparing the COVID-19 pandemic and non-pandemic periods in Canada found no changes in the number of emergency stroke admissions, which was found to be similar to the present study. However, the referral rate to stroke prevention clinics decreased, and the number of patients with minor symptoms decreased. Patients with transient ischemic attacks or very mild strokes may have decided to stay at home instead of seeking medical care, according to a previous study [32]. The results of this study thus suggest that emergency hospitalisation for stroke is normal in Japan.
In this study, the level of consciousness information was used to determine the severity of stroke at the time of hospitalisation, but it did not consider information on treatment such as thrombolytic therapy, or tests such as CT and MRI, which may have resulted in biased information because we could not correctly determine the severity of stroke and other patient conditions. However, it is clear from data such as the number of 24-hour deaths in which the severity of stroke was higher at the time of admission, that there has been a change in the course of events leading to hospitalization.

Conclusions
We found that the prevalence of the COVID-19 pandemic increased the number of deaths and severity of illness within 24 h of hospitalisation in Japan. However, there was no association between the number of hospital admissions and deaths with baseline characteristics. These results suggest that there was no problem in accepting hospitalised patients during the COVID-19 pandemic, but the delay in visiting the hospital after the onset of the disease may have increased the severity of stroke. The results of this study suggest the need to monitor the risk of stroke and changes in patients' conditions before the onset of stroke as the infection spreads. As the COVID-19 infection pandemic becomes more prolonged, the overall severity of the stroke-prone population may also change. The authorities need to encourage patients with high need to come to the hospital and provide appropriate information about high-risk patients in the community.
Abbreviations CI: Cerebral infarction; CH: Cerebral haemorrhage; SH: Subarachnoid haemorrhage CCI: Carlson comorbidity index; BI: Barthes index; BMI: body mass index; RR: risk ratio; 95%CI: 95% con dence interval; JCS: Japan Coma Scale; LOS: length of hospital stay; mRS: modi ed Ranking Scale; SMD: standard mean difference Declarations TT analysed all the data and was a major contributor to the manuscript preparation. TT, SI, and KF interpreted the data. All authors approved the nal manuscript.

Ethics approval and consent to participate
This study was approved by the ethics committee of the Tokyo Medical and Dental University. The need for informed consent was waived owing to the use of anonymized data.

Consent for publication
Not applicable.

Availability of data and materials
The datasets supporting the conclusions of this article are available from the corresponding author upon reasonable request.