Effects of Meteorological Conditions on the Risk of Ischemic Stroke Events in Patients Treated with Alteplase--HEWS-tPA

Medical meteorological forecast for ischemic stroke: random forest regression vs long short-term memory model

Ischemic stroke (IS) is one of the top risk factors for death and disability. Meteorological conditions have an effect on IS attack. In this study, we try to develop models of medical meteorological forecast for IS attack based on machine learning and deep learning algorithms. The medical meteorological forecast would be beneficial to public health in IS events prevention and treatment. We collected data on IS attacks and climatology in each day from 18th September 2016 to 31th December 2020 in Haikou. Data on IS attacks were from the number of hospital admissions due to IS attack among general population. The random forest (RF) regression and long short-term memory (LSTM) algorithms were respectively used to develop the predictive model based on meteorological data. Performance of the model was assessed by mean squared error (MSE) and root mean squared error (RMSE). A total of 42849 IS attacks was included in this study. IS attacks were significantly decreased in winter. The pattern of climatological data was observed the regularity in seasons. For the performance of RF regression model, the MSE is 243, and the RMSE is 15.6. For LSTM model, the MSE is 36, and the RMSE is 6. In conclusion, LSTM model is more accurate than RF regression model to predict IS attacks in general population based on meteorological data. LSTM model showed acceptable accuracy for the prediction and could be used as medical meteorological forecast to predict IS attack among population according to local climate.

High average daily temperature in summer and the incidence of thrombolytic treatment for acute ischemic stroke

INTRODUCTION

Meteorological factors can increase stroke risk; however, their impact is not precisely understood. Heat waves during summer increase total mortality. Therefore, we hypothesized that the average daily temperature in summer may correlate with the incidence of thrombolytic treatment for acute ischemic stroke in Budapest and Pest County, Hungary.

METHODS

We analyzed the relationship between the average daily temperature in summer months and the daily number of thrombolytic treatments (TT) performed with the indication of acute ischemic stroke between 1st June and 31st August each year from 2007 to 2016. The analysis was also performed after the omission of the data of the last day of the months due to possible psychosocial impact reported in our previous study. Spearman's correlation was used for statistical analysis.

RESULTS

No significant correlation was found between the average summer daily temperature and the number of TT in the entire sample of the 10-year period. When omitting the data of the last day of each month, positive correlations were suspected in 2014 (r=0.225, P=0.034) and 2015 (r=0.276, P=0.009).

CONCLUSION

Our findings did not confirm an association between the average daily temperature in summer and the daily number of TT throughout the examined 10-year period. However, importantly, in 2014 and 2015, the years with the highest average daily temperatures in this period, a positive correlation was found. The level of correlation is modest, indicating that risk factors, both meteorological and non-meteorological, other than the average temperature, play equally important roles in determining the incidence of thrombolytic treatment for acute ischemic stroke on the population level.

Association between El Niño-Southern Oscillation events and stroke: a case-crossover study in Kaunas city, Lithuania, 2000-2015

The aim of this study was to determine the association between the daily number of cases of ischemic stroke (IS) and hemorrhagic stroke (HS) in patients aged 25-64 years and the El Niño-Southern Oscillation (ENSO) events during 2000-2015. As an indicator of the effect of the ENSO, the monthly NIÑO 3.4 index (Equatorial Pacific Sea Surface Temperature) was used. During the 5844-day study period, 5600 cases of stroke (3170 (56.61%) in men and 2430 (43.39%) in women) were analyzed. Of these, 4354 (77.8%) cases were IS, and 1041 (18.6%) cases were HS. In 3496 (62.2%) cases, stroke occurred in the age group of 55-64 years. In the analysis, we used the following categories of the ENSO events: strong La Niña, moderate La Niña, moderate El Niño, and strong El Niño. The effect of the ENSO was examined by using the multivariate Poisson regression adjusting for weather variables. The highest risk of both strokes (BS) was observed on days of strong and moderate La Niña (rate ratio (RR) 1.27, 95% CI 1.13-1.42) and RR = 1.15 (1.07-1.23), respectively), while the risk for IS was the highest on days of moderate El Niño (RR = 1.11(1.02-1.20)). A lower risk for BS was found on days of strong El Niño (RR = 0.77(0.62-0.97)). We found that ENSO events affected the occurrence of BS and IS in all age groups, and the strongest effect was observed among females. The results of this study provide new evidence that ENSO events may affect the risk of stroke, especially the risk of IS.

Seasonal patterns and associations in the incidence of acute ischemic stroke requiring mechanical thrombectomy

Background

In order to identify risk periods with an increased demand in technical and human resources, we tried to determine patterns and associations in the incidence of acute ischemic stroke due to embolic large vessel occlusions (eLVO) requiring mechanical thrombectomy (MT).

Methods

We conducted a time series analysis over a 9‐year period (2010–2018) based on observational data in order to detect seasonal patterns in the incidence of MT due to eLVO (n = 2628 patients). In a series of sequential negative binominal regression models, we aimed to detect further associations (e.g., temperature, atmospheric pressure, air pollution).

Results

There was a 6‐month seasonal pattern in the incidence of MT due to eLVO (p = 0.024) peaking in March and September. Colder overall temperature was associated with an increase in MT due to eLVO (average marginal effect [AME], [95% CI]: −0.15 [−0.30–0.0001]; p = 0.05; per °C). A current increase in the average monthly temperature was associated with a higher incidence of MT due to eLVO (0.34 [0.11–0.56]; p = 0.003). Atmospheric pressure was positively correlated with MT due to eLVO (0.38 [0.13–0.64]; p = 0.003; per hectopascal [hPa]). We could detect no causal correlation between air pollutants and MT due to eLVO.

Conclusions

Our data suggest a 6‐month seasonal pattern in the incidence of MT due to eLVO peaking in spring and early autumn. This might be attributed to two different factors: (1) a current temperature rise (comparing the average monthly temperature in consecutive months) and (2) colder overall temperature. These results could help to identify risk periods requiring an adaptation in local infrastructure.

Prediction Model of Deep Learning for Ambulance Transports in Kesennuma City by Meteorological Data

PURPOSE

With the aging population in Japan, the prediction of ambulance transports is needed to save the limited medical resources. Some meteorological factors were risks of ambulance transports, but it is difficult to predict in a classically statistical way because Japan has 4 seasons. We tried to make prediction models for ambulance transports using the deep learning (DL) framework, Prediction One (Sony Network Communications Inc., Tokyo, Japan), with the meteorological and calendarial variables.

MATERIALS AND METHODS

We retrospectively investigated the daily ambulance transports and meteorological data between 2017 and 2019. First, to confirm their association, we performed classically statistical analysis. Second, to test the DL framework's utility for ambulance transports prediction, we made 3 prediction models for daily ambulance transports (total daily ambulance transports more than 5 or not, cardiopulmonary arrest (CPA), and trauma) using meteorological and calendarial factors and evaluated their accuracies by internal cross-validation.

RESULTS

During the 1095 days of 3 years, the total ambulance transports were 5948, including 240 CPAs and 337 traumas. Cardiogenic CPA accounted for 72.3%, according to the Utstein classification. The relation between ambulance transports and meteorological parameters by polynomial curves were statistically obtained, but their r2s were small. On the other hand, all DL-based prediction models obtained satisfactory accuracies in the internal cross-validation. The areas under the curves obtained from each model were all over 0.947.

CONCLUSION

We could statistically make polynomial curves between the meteorological variables and the number of ambulance transport. We also preliminarily made DL-based prediction models. The DL-based prediction for daily ambulance transports would be used in the future, leading to solving the lack of medical resources in Japan.

Preliminary development of a prediction model for daily stroke occurrences based on meteorological and calendar information using deep learning framework (Prediction One; Sony Network Communications Inc., Japan)

BACKGROUND

Chronologically meteorological and calendar factors were risks of stroke occurrence. However, the prediction of stroke occurrences is difficult depending on only meteorological and calendar factors. We tried to make prediction models for stroke occurrences using deep learning (DL) software, Prediction One (Sony Network Communications Inc., Tokyo, Japan), with those variables.

METHODS

We retrospectively investigated the daily stroke occurrences between 2017 and 2019. We used Prediction One software to make the prediction models for daily stroke occurrences (present or absent) using 221 chronologically meteorological and calendar factors. We made a prediction models from the 3-year dataset and evaluated their accuracies using the internal cross-validation. Areas under the curves (AUCs) of receiver operating characteristic curves were used as accuracies.

RESULTS

The 371 cerebral infarction (CI), 184 intracerebral hemorrhage (ICH), and 53 subarachnoid hemorrhage patients were included in the study. The AUCs of the several DL-based prediction models for all stroke occurrences were 0.532-0.757. Those for CI were 0.600-0.782. Those for ICH were 0.714-0.988.

CONCLUSION

Our preliminary results suggested a probability of the DL-based prediction models for stroke occurrence only by meteorological and calendar factors. In the future, by synchronizing a variety of medical information among the electronic medical records and personal smartphones as well as integrating the physical activities or meteorological conditions in real time, the prediction of stroke occurrence could be performed with high accuracy, to save medical resources, to have patients care for themselves, and to perform efficient medicine.

Warm Front Passage on the Previous Day Increased Ischemic Stroke Events

BACKGROUND AND PURPOSE

The influence of a weather front passage is rarely evaluated on stroke events. We hypothesized that a weather front passage on the stroke onset day or during the previous days may play an important role in the incidence of stroke.

METHODS

A multicenter retrospective study was conducted to evaluate the frequency of stroke events and their interaction with weather front passages. Consecutive acute stroke patients (n = 3935, 73.5 ± 12.4 years, 1610 females) who were admitted to 7 stroke hospitals in 3 cities from January 2012 to December 2013 were enrolled in this study. Multivariate Poisson regression models involving time lag variables were used to compare the daily rates of stroke events with the day of a weather front passage and the previous 6 days, adjusting for considerable influences of ambient temperature and atmospheric pressure.

RESULTS

There were a total of 33 cold fronts and 13 warm fronts that passed over the 3 cities during the study period. The frequency of ischemic stroke significantly increased when a warm front passed on the previous day (risk ratio 1.34, 95% confidence interval 1.07-1.69, P= .016).

CONCLUSIONS

This study indicated that a weather front passage on the previous days may be associated with the occurrence of stroke.

Various meteorological conditions exhibit both immediate and delayed influences on the risk of stroke events: The HEWS-stroke study

We hypothesized that meteorological conditions on the onset day and conditions on the former days may play important roles in the modulation of physical conditions. Associations of meteorological factors and their changes in former days with stroke onset are of interest. We conducted a multicenter retrospective study to evaluate the frequency of stroke events and their interaction with meteorological conditions and their daily changes. Acute stroke patients (n = 3935, 73.5±12.4 years, 1610 females) who were admitted to 7 stroke hospitals in three restricted areas were enrolled in this study. Poisson regression models involving time-lag variables was used to compare daily rates of stroke events with mean thermo-hydrological index (THI), atmospheric pressure, and their daily changes. We divided onset days into quintiles based on the THI, atmospheric pressure, and their daily changes for the last 7 days. The frequencies of ischemic stroke significantly increased when THI varied either cooler or warmer from a previous day (extremely cooler, risk ratio (RR) 1.19, 95% confidence interval (CI) 1.05 to 1.34; extremely warmer, RR 1.16, 95% CI 1.03 to 1.31; r2 = 0.001 for the best regression, p = 0.001). Intracerebral hemorrhage frequencies significantly decreased on high-THI days (extremely high, RR 0.72, 95% CI 0.54 to 0.95; r2 = 0.013 for the best regression, p<0.001) and increased in high atmospheric pressure days (high, RR 1.31, 95% CI 1.04 to 1.65; r2 = 0.009 for the best regression, p<0.001). Additionally, even after adjusting for the THI on the onset day and its changes for the other days, intracerebral hemorrhage increased when THI got extremely cooler in 4 days prior (RR 1.33, 95% CI 1.03 to 1.71, r2 = 0.006 for the best regression, p<0.001). Various meteorological conditions may exhibit influences on stroke onset. And, when temperature cooled, there may be a possibility to show delayed influence on the frequency of intracerebral hemorrhage 4 days later.

Anomalous equivalent potential temperature: an atmospheric feature predicting days with higher risk for fatal outcome in acute ischemic stroke-a preliminary study

Acute stroke is a life-threatening condition. Fatal outcome is related to risk factors, some of these affected by climatic changes. Forecasting potentially harmful atmospheric processes may therefore be of practical importance in the acute care of stroke patients. We analyzed the history of all patients with acute ischemic stroke (N = 184) confirmed by neuroimaging including those who died (N = 35, 15 males) at our hospital department in the winter months of 2009. Patient data were anonymized, and the human meteorologists were only aware of patients' age, gender, and exact time of death. Of the meteorological parameters, equivalent potential temperature (EPT) has been chosen for analysis. EPT is generally used for forecasting thunderstorms, but in the case of synoptic scale airflow (10(6) m), it is suitable for characterizing the air mass inflowing from different regions. The behavior of measured EPT values was compared to the climatic (30 years) averages. We developed meteorological criteria for anomalous periods of EPT and tested if such periods are associated with higher rate of fatal outcome. The duration of anomalous and non-anomalous periods was nearly equal during the studied 3 months. Stroke onset distributed similarly between anomalous and non-anomalous days; however, of the 35 deaths, 27 occurred during anomalous periods: on average, 0.56 deaths occurred on anomalous days and 0.19 on non-anomalous days. Winter periods meeting the criteria of anomalous EPT may have a significant adverse human-meteorological impact on the outcome in acute ischemic stroke.

Short-Term Effect of Ambient Temperature and the Risk of Stroke: A Systematic Review and Meta-Analysis

BACKGROUND AND PURPOSE

The relationship between stroke and short-term temperature changes remains controversial. Therefore, we conducted a systematic review and meta-analysis to investigate the association between stroke and both high and low temperatures, and health assessment.

METHODS

We searched PubMed, Embase, Cochrane, China National Knowledge Infrastructure (CNKI) and Wanfang Data up to 14 September 2014. Study selection, quality assessment, and author-contractions were steps before data extraction. We converted all estimates effects into relative risk (RR) per 1 °C increase/decrease in temperature from 75th to 99th or 25th to 1st percentiles, then conducted meta-analyses to combine the ultimate RRs, and assessed health impact among the population.

RESULTS

20 articles were included in the final analysis. The overall analysis showed a positive relationship between 1 °C change and the occurrence of major adverse cerebrovascular events (MACBE), 1.1% (95% confidence intervals (CI), 0.6 to 1.7) and 1.2% (95% CI, 0.8 to 1.6) increase for hot and cold effects separately. The same trends can be found in both effects of mortality and the cold effect for morbidity. Hot temperature acted as a protective factor of hemorrhage stroke (HS), -1.9% (95% CI, -2.8 to -0.9), however, it acted as a risk factor for ischemic stroke (IS), 1.2% (95% CI, 0.7 to 1.8).

CONCLUSION

Short-term changes of both low and high temperature had statistically significant impacts on MACBE.