key: cord-0285801-0f4bvfsi authors: Risser, C.; Tran Ba Loc, P.; Binder-Foucard, F.; Fabacher, T.; Lefevre, H.; Sauvage, C.; Sauleau, E.; Wolff, V. title: COVID-19 impact on stroke admissions during France's first epidemic peak: an exhaustive, nationwide, observational study date: 2021-08-05 journal: nan DOI: 10.1101/2021.08.03.21261438 sha: 23f749cd48300d26f311398f3269a449c5fae535 doc_id: 285801 cord_uid: 0f4bvfsi Background and Purpose: The COVID-19 pandemic continues to have great impacts on the care of non-COVID-19 patients. This was especially true during the first epidemic peak in France, which coincided with the national lockdown (17 March 2020 to 10 May 2020). Patients with serious and urgent disease like stroke may have experienced a degradation of care, or may have been hesitant to seek healthcare during this period. The aim of this study was to identify, on a national level, whether a decrease in stroke admissions occurred in spring 2020, by analyzing the evolution of all stroke admissions in France from January 2019 to June 2020. Methods: We conducted a nationwide cohort study using the French national database of hospital admissions (PMSI) to extract exhaustive data on all hospitalizations in France with at least one stroke diagnosis between 1 January 2019 and 30 June 2020. The primary endpoint was the difference in the slope gradients of stroke hospitalizations between pre-epidemic, epidemic peak and post-epidemic periods. Modeling was carried out using Bayesian techniques. Results: Stroke hospitalizations dropped from 10 March 2020 (slope gradient: -11.70), and began to rise again from 22 March (slope gradient: 2.090) to 7 May. In total, there were 23 873 stroke admissions during the period March-April 2020, compared to 29 263 at the same period in 2019, representing a decrease of 18.42%. The percentage change was -15.63%, -25.19%, -18.62% for ischemic strokes, transient ischemic attacks, and hemorrhagic strokes, respectively. In spatial models of French departments, the incidence of COVID-19 explained the ratio of stroke hospitalizations. Conclusions: Stroke hospitalizations in France experienced a decline during the first lockdown period, which cannot be explained by a sudden change in stroke incidence. This decline is therefore likely to be a direct, or indirect, result of the COVID-19 pandemic. Since the emergence of the COVID-19 pandemic, healthcare providers around the globe have been forced to mobilize their resources in order to cope in these unprecedented conditions. The care of non-COVID patients has certainly been impacted as a result. [1] [2] [3] [4] The capacity of healthcare systems worldwide was rapidly overwhelmed following the beginning of the pandemic. France faced a shortage of both medical supplies and human resources during the first epidemic peak, in spring 2020, despite undertaking 55 days of national lockdown (from 17 March to 10 May 2020 included) . 5-7 On the final day of this lockdown, the number of COVID-19 related hospital deaths recorded was 16 820. 8 A reduction in the quality and quantity of care as a result of the extreme pressure experienced by the healthcare system may have resulted in collateral damage with regard to other patients, particularly for those with serious or urgent diseases requiring immediate treatment. 1, 9, 10 In addition, many patients have been hesitant to seek heath care during this period, a phenomenon briefly described in the literature regarding the shorter SARS epidemics. 11, 12 To date, several works have alerted to an unexpected decrease in hospital admissions for non COVID-19 diseases, but very few have estimated this on a national level. [13] [14] [15] The aim of this study was to determine whether the number of hospital admissions for stroke in France decreased during the COVID-19 epidemic peak and lockdown period, by analyzing the evolution in number of all stroke hospitalizations from January 2019 to June 2020. The secondary objectives were to observe the evolution of different types of stroke; to compare selected hospitalization characteristics between the epidemic peak period March-April 2020 and March-April 2019; to study the correlation between the incidence of hospital admission for COVID-19 and the evolution in stroke admissions between March-April 2020 and March-April 2019 by French department. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint The data that support the findings of this study are available from the corresponding author upon reasonable request. This study adheres to French legislation regarding the reuse of anonymized data (MR-005 of Anonymized data were extracted from the French national "Information Systems Medicalisation Programme" (PMSI) database, which includes all hospitalization data transmitted by all public and private hospitals in France. Diagnoses are coded using the International Classification of Diseases, 10th Revision (ICD-10). In France, at the end of a hospital stay, all patient diagnostics and medical procedures are recorded from the patient's medical records according to national coding rules. Medical and administrative information is transmitted monthly to the national database within one month of the patient's discharge. A perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint this study: "hemorrhagic" (I60, I61, I62), "ischemic" (I63), and "transient" (G45). The database request, made retrospectively on 13 October 2020, included all hospital admissions between 1 January 2019 and 30 June 2020, with at least one stroke diagnosis, as defined by ATIH coding, during the stay. In order to prevent any selection bias which may occur by counting the same acute episode of stroke twice, only one admission per patient was counted if two admissions for the same patient were separated by one day or less. The study size was therefore defined by the relevant entries extracted from the database. The age of the patient was recorded as the age upon admission to hospital. The location of hospital stay was determined by the French department in which the patient was initially admitted. The primary endpoint is the difference in slopes of stroke hospitalizations between the different periods (pre-epidemic, epidemic peak, post-epidemic) estimated by the model. The Quantitative variables are presented as the mean with standard deviation, or median with the first and third quartiles of the distributions. Qualitative variables are presented as numbers and percentages. We considered the change in the number of stroke admissions as centered 7-days rolling means. To compare slopes between time periods, we modelled the evolution of stroke All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint 6 hospitalizations in four segments of lines (stability, fall, rise, return to baseline) joined by three estimated knots. Disregarding possible seasonality effects, we assumed relative stability of hospitalizations, including during the summer of 2019, before a sudden change in 2020. Modeling was carried out using Bayesian techniques with a 95% credibility interval. Probability of superiority is rounded at both extremities to <0.0001 and >0.9999. The prior distributions are assumed to be very uninformative except for the position of each of the 3 knots, which is assumed to be uniformly distributed over the time interval between 15 February 2020 and 30 June 2020. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint To identify a cause behind the change in the number of stroke admissions, a spatial correlation study was carried out by graphically representing, by French department, the incidence rates of hospital admissions for COVID-19 in March-April 2020, and the ratio between the number of stroke admissions in March-April 2020 compared to those in March-April 2019. The incidence rates of hospital admissions for COVID-19 were standardized by age and sex, using the official data estimates of French department populations by quinquennial ages and gender in 2020, published by the National Institute of Statistics and Economic Studies (INSEE) 16 ; if two stays for COVID-19 for the same patient were separated by 1 day or less, only one stay was counted. An ecological normal regression was fitted on data collected from metropolitan France between the ratio of stroke hospitalizations to standardized incidence for COVID-19, and a quadratic spatial trend established (on the centroids of each department), in order to adjust for spatial effects. We used SAS Enterprise Guide 8.3 for data requests on the PMSI national database. All statistical analyses and graphs were carried out using R software, version 4.0.2, R Core Team (2020). perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint The posterior medians for three knots were estimated on: 10 March 2020 (K1), 22 March 2020 (K2) and 7 May 2020 (K3), with small credibility intervals (K1, K2, K3 and their credibility intervals are detailed in Table 1 , and credibility intervals shown as horizontal lines at the relevant position in Figure 1 ). Because the 7-day rolling mean number of hospital admissions is modeled, the changes in hospitalizations may have occurred in the three days either side of the given date. After 10 March 2020 (K1), the slope of the second "fall" segment was negative, at -11.70 [-16.25; -8.552 ]. At K3, i.e., after the "rise" period, the mean number of hospital admissions is still -37. 26 per day compared to the number calculated if the slopes of all segments had remained the same as that of the first segment. After 7 May 2020 (K3), the slope of the fourth segment remains higher than that of the first segment (posterior median 0.4663), resulting in an upward slope compared to the "stability" period, albeit lower than that of the third "rise" period (2.090). In Figure 2 , showing the evolution according to the type of stroke, the decrease in hospital admissions is visible for all three types of stroke during the period March-April 2020. The number of admissions in March-April 2019 was 16 776, 6978 and 5509, respectively, for ischemic, TIA and hemorrhagic stroke, compared to 14 154, 5220, and 4483 for the same period in 2020. As shown in Table 2 , decrease in hospitalizations is common to all subtypes of hemorrhagic stroke, in spite of their differing physiopathologies and care. Hospital stays were shorter in March-April 2020 than in March-April 2019, whilst the proportion of men, as well as the ages of admitted stroke patients, were comparable between the two years ( Table 2 ). The death rate was slightly higher in 2020. In March-April 2020, 2.26% of hospitalizations for stroke were associated with a COVID-19 diagnosis. As shown in Figure 3A , the ratio of the number of stroke admissions in March-April 2020 to the number of stroke admissions in March-April 2019 was lower than 1 for 89 out of 96 departments (metropolitan France only). The departments with the greatest decrease in All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint admissions are the region of Paris, the north east and the south east of France. The map of age and sex standardized incidence rates of COVID-19 hospital admissions by department ( Figure 3B ) shows a higher incidence in the north east of France and the Parisian region. Spatial models showed that the incidence of COVID-19 hospitalizations explains the ratio of stroke admissions (as the zero line is outside of the credible interval of the smoothed effect), whether the spatial trend is included in the model or not (Figure 4 ). All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint We showed that the number of patients admitted to hospital for stroke decreased significantly in France during the peak of the COVID-19 epidemic, at the time when a pandemic was declared by the WHO and lockdown measures were taken in France. This decrease is seen across all stroke subtypes. Proportions between the sexes were similar between the two periods. Differences found in the duration of hospital stay and the mortality rate are difficult to interpret because some long-term hospital stays in 2020 may have ended only after data extraction, and/or relevant data may have been uploaded late to the national database. In contrast, long-term hospital stays in 2019 are much less likely to be affected by such factors, as a longer follow-up period is present. This sudden drop in the number of hospital admissions cannot be explained by a change in incidence due to seasonality, nor by a change of hospital facility, because our study is national. It is most likely a direct, as well as indirect, consequence of the COVID-19 pandemic. The apparently identical spatial distribution between COVID-19 and decreased stroke admission does not seem to be due to an identical spatial trend, but rather to the influence of COVID-19 on stroke admissions, because the link between the two persists even when a trend is also modelled. This phenomenon was described in the literature for emergency department visits during the 2003 SARS outbreak in Taipei (up to 51.6% reduction in the number of daily emergency department visits compared to the number of visits prior to the outbreak). 12 It was also described for emergency department visits in Toronto during the 2003 SARS outbreak (21% reduction over the 4-week study period). 17 More recently, authors described the impact of lockdown measures on emergency department visits in Lebanon, determining a 97.11% decrease per day following the "declaration of general mobilization". 18 Equally, in a tertiary pediatric emergency department in Cincinnati, United States, an area which experienced All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in be the result of healthcare providers redirecting a higher proportion of patients out of hospital due to a lack of capacity, or, due to an overall under-diagnosis of stroke in these unprecedented and demanding pandemic conditions. In support of these hypotheses, our study found a greater decline in stroke admissions in several departments which experienced a higher incidence of COVID-19 hospitalization. It seems unlikely that the epidemic context associated to national lockdown could have influenced stroke risk factors within such a short time. In the literature, the incidence of neurological symptoms in patients with COVID-19 is between 1.6% and 2.5%, depending on the study. [24] [25] [26] [27] [28] A bicenter study of two New York City hospitals found that 31 patients out of a total 1916 admitted to the emergency department or hospitalized with COVID-19 had suffered an acute ischemic stroke, including 8 in which stroke was the reason for referral. In comparison, of the 1486 patients with a diagnosis of influenza, 3 had suffered an ischemic All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint stroke. The authors found that even after adjustment for age, sex, and race, there is a higher probability of stroke with SARS-CoV-2 infection than with influenza (OR 7.6; 95% CI 2.3-25.2). 28 The relationship between COVID-19 and stroke therefore exists, and a history of stroke results in a worse prognosis for a patient with COVID-19. The incidence of stroke in COVID-19 patients is difficult to assess, and causality difficult to establish, but it appears to be in favor of increasing rather than decreasing the incidence of stroke. This may be due to the pathophysiology of COVID-19, which involves inflammation and hypercoagulability. 26, [28] [29] [30] [31] [32] [33] Although COVID-19 may be responsible for neurological symptoms, it is also possible that it may have been a concurrent cause of death. 32 As a result, a number of patients who may have otherwise suffered a stroke may have died from COVID-19 before the stroke occurred, or indeed, may have developed a stroke but which remained undiagnosed for a variety of factors linked to the health crisis context. To the best of our knowledge, this is the first nationwide study in France, a country with a relatively high incidence rate of stroke (in 2014, the acute care hospitalization rate was 167.9 per 100,000 habitants, age-standardized according to the 2010 European population 34 ), describing the evolution of stroke admissions during the first epidemic peak of COVID-19. Thanks to a time period of few months, we have an almost exhaustive count of stroke hospitalizations in France from January 2019 to June 2020. Using our statistical methods, we have been able to highlight the decline in national stroke patient care during the peak of the epidemic, while taking into account the evolution of stroke admissions over the entire year. These results alert us to the need for adapting stroke management in such circumstances, and, more broadly, how to ensure appropriate care for non-COVID-19 patients during the epidemic. The impact of the epidemic on cerebrovascular diseases may be seen in the months and years to come. In contrast to other studies, we did not restrict the analyses to stroke emergency department visits, but to hospital admissions for stroke, which is even more All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint alarming. In addition, to measure the magnitude of the impact on stroke, we decided to count admissions rather than patients, taking into account the possibility that patients may suffer multiple strokes during the study period. We also pooled transfers between hospitals during patient care to avoid a stay being counted several times. A limitation of using the PMSI national database is that a possible delay can occur between patient discharge and coding. However, the extent of this bias is limited, as the hospital is reimbursed by the state health insurance only if diagnoses made during hospitalization are sent to the national database no more than one month after discharge. To ensure that data is as complete as possible, the analysis period was ended on 30 June 2020, i.e., leaving more than three months for patient codes to be registered before the date of extraction from the database in October 2020. Homogeneity of the diagnostic coding, which must be as close as possible to those noted during the stay, is provided by strict, national rules with regular checks carried out by the payer, thus limiting misclassification bias. A final limitation of our study is restricted knowledge of other external local events which could influence the activity of the hospital facility. Such biases are nevertheless partially controlled by the study design, as analysis is carried out at a national level, and the annual time periods compared time periods compared are proximal. Within this study, we have not extended the analysis across a sufficient number of previous years to be able to determine whether there is a seasonal effect on stroke admissions. In previous unpublished analyses, carried out for internal purposes using data from Strasbourg University Hospital, we were able to identify a decrease in the trend of hospital admissions for stroke during the summer months of July-August, from 2015 to 2019. It is unclear whether this potential seasonal effect also concerns the period of March-April, in which case it would reduce the effect size of the difference measured during the epidemic peak. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint This study has determined a nationwide decrease in stroke hospital admissions at the time of the first COVID-19 wave in France. The drastic lockdown measures and unprecedented epidemic context have most likely impacted the probability of patients seeking hospital assistance in France, particularly in those regions most affected by COVID-19. In light of these findings, the care provided for stroke should be reconsidered in order to prevent stroke under-diagnosis, to improve outpatient medical care, and to facilitate health provider decisionmaking during the crisis. State news conferences announcing restrictions should also emphasize the importance of continuing to seek expert medical care when needed, for example, in the case of stroke. It should be highlighted to the general population that patient care for stroke remains of high quality during the crisis, and that the emergency services should always be contacted if signs of stroke appear. Equally, the public health authorities should take into account the direct, and indirect, consequences of restrictive measures on stroke admissions, and use this to inform public health decisions. It would also be informative to look at the long-term effects of the epidemic on stroke admissions. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in None Conflict(s)-of-Interest/Disclosure(s) None All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in (A) shows that the reduction of the ratio correlates with an increase in standardized COVID incidence. (B) Map of estimated spatial quadratic trend in metropolitan France, shaded from light yellow (lower reduction of ratio) to dark red (stronger reduction). Because of the different spatial scales between aggregated data for French departments and the continuous quadratic trend, the unit in the map is an estimation of ratio reduction. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted August 5, 2021. ; https://doi.org/10.1101/2021.08.03.21261438 doi: medRxiv preprint Non-COVID-19 Patients with Life-threatening Diseases Who Visited a Fever Clinic: A Single-center What if the worst consequences of COVID-19 concerned non-COVID patients? Challenges of non-COVID-19 patients with chronic illness during the pandemic Non-COVID Diseases during the Pandemic: Where Have All Other Emergencies Gone? Med Kaunas Lith; 56. Epub ahead of print 1 Estimating the burden of SARS-CoV-2 in France Impact of lockdown on COVID-19 epidemic in Îlede-France and possible exit strategies Données relatives à l'épidémie de COVID-19 en France : vue d'ensemble -data Personal protective equipment for surgeons during COVID-19 pandemic: systematic review of availability, usage and rationing Stroke care during the COVID-19 pandemic: experience from three large European countries COVID-19 and stroke-A global World Stroke Organization perspective Declining Emergency Department Visits and Costs During the Severe Acute Respiratory Syndrome (SARS) Outbreak Where have the STEMIs gone during COVID-19 lockdown? Eur Heart J Qual Care Clin Outcomes. Epub ahead of print 29 Estimation de la population au 1 er janvier 2021 | Insee Effect of the SARS outbreak on visits to a community hospital emergency department The impact of COVID-19 lockdown measures on ED visits in Lebanon The impact of public health interventions on critical illness in the pediatric emergency department during the SARS-CoV-2 pandemic Impact of a Stay-at-Home Order on Stroke Admission, Subtype, and Metrics during the COVID-19 Pandemic Decline in stroke alerts and hospitalisations during the COVID-19 pandemic. Stroke Vasc Neurol. Epub ahead of print 27 Analysis of Nationwide Stroke Patient Care in Times of COVID-19 Pandemic in Germany Global Impact of COVID-19 on Stroke Care and Intravenous Thrombolysis Neurologic Features in Severe SARS-CoV-2 Infection Risk of stroke in hospitalized SARS-CoV-2 infected patients: A multinational study. EBioMedicine; 59. Epub ahead of print 1 SARS-CoV-2 and Stroke in a New York Healthcare System Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China Risk of Ischemic Stroke in Patients With Coronavirus Disease 2019 (COVID-19) vs Patients With Influenza Risk Factors of Fatal Outcome in Hospitalized Subjects With Coronavirus Disease 2019 From a Nationwide Analysis in China Clinical and Imaging Characteristics in Patients with SARS-CoV-2 Infection and Acute Intracranial Hemorrhage Intracranial hemorrhage in coronavirus disease 2019 (COVID-19) patients Neurological manifestations and implications of COVID-19 pandemic High Prevalence of Deep Venous Thrombosis in Non-Severe COVID-19 Patients Hospitalized for a Neurovascular Disease Accident vasculaire cérébral All rights reserved. No reuse allowed without permission. perpetuity preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted