Predicting the hotspots of age-adjusted mortality rates of lower respiratory infection across the continental United States: Integration of GIS, spatial statistics and machine learning algorithms

Mapping the pandemic: a review of Geographical Information Systems-based spatial modeling of Covid-19

According to the World Health Organization (WHO), COVID-19 has caused more than 6.5 million deaths, while over 600 million people are infected. With regard to the tools and techniques of disease analysis, spatial analysis is increasingly being used to analyze the impact of COVID-19. The present review offers an assessment of research that used regional data systems to study the COVID-19 epidemic published between 2020 and 2022. The research focuses on: categories of the area, authors, methods, and procedures used by the authors and the results of their findings. This input will enable the contrast of different spatial models used for regional data systems with COVID-19. Our outcomes showed increased use of geographically weighted regression and Moran I spatial statistical tools applied to better spatial and time-based gauges. We have also found an increase in the use of local models compared to other spatial statistics models/methods.

Mapping the risk of respiratory infections using suburban district areas in a large city in Colombia

BACKGROUND

Acute respiratory infections (ARI) in Cúcuta -Colombia, have a comparatively high burden of disease associated with high public health costs. However, little is known about the epidemiology of these diseases in the city and its distribution within suburban areas. This study addresses this gap by estimating and mapping the risk of ARI in Cúcuta and identifying the most relevant risk factors.

METHODS

A spatial epidemiological analysis was designed to investigate the association of sociodemographic and environmental risk factors with the rate of ambulatory consultations of ARI in urban sections of Cúcuta, 2018. The ARI rate was calculated using a method for spatial estimation of disease rates. A Bayesian spatial model was implemented using the Integrated Nested Laplace Approximation approach and the Besag-York-Mollié specification. The risk of ARI per urban section and the hotspots of higher risk were also estimated and mapped.

RESULTS

A higher risk of IRA was found in central, south, north and west areas of Cúcuta after adjusting for sociodemographic and environmental factors, and taking into consideration the spatial distribution of the city's urban sections. An increase of one unit in the percentage of population younger than 15 years; the Index of Multidimensional Poverty and the rate of ARI in the migrant population was associated with a 1.08 (1.06-1.1); 1.04 (1.01-1.08) and 1.25 (1.22-1.27) increase of the ARI rate, respectively. Twenty-four urban sections were identified as hotspots of risk in central, south, north and west areas in Cucuta.

CONCLUSION

Sociodemographic factors and their spatial patterns are determinants of acute respiratory infections in Cúcuta. Bayesian spatial hierarchical models can be used to estimate and map the risk of these infections in suburban areas of large cities in Colombia. The methods of this study can be used globally to identify suburban areas and or specific communities at risk to support the implementation of prevention strategies and decision-making in the public and private health sectors.

Machine learning for predicting the risk stratification of 1-5 cm gastric gastrointestinal stromal tumors based on CT

BACKGROUD

To predict the malignancy of 1-5 cm gastric gastrointestinal stromal tumors (GISTs) by machine learning (ML) on CT images using three models - Logistic Regression (LR), Decision Tree (DT) and Gradient Boosting Decision Tree (GBDT).

METHODS

231 patients from Center 1 were randomly assigned into the training cohort (n = 161) and the internal validation cohort (n = 70) in a 7:3 ratio. The other 78 patients from Center 2 served as the external test cohort. Scikit-learn software was used to build three classifiers. The performance of the three models were evaluated by sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV) and area under the curve (AUC). Diagnostic differences between ML models and radiologists were compared in the external test cohort. Important features of LR and GBDT were analyzed and compared.

RESULTS

GBDT outperformed LR and DT with the largest AUC values (0.981 and 0.815) in the training and internal validation cohorts and the greatest accuracy (0.923, 0.833 and 0.844) across all three cohorts. However, LR was found to have the largest AUC value (0.910) in the external test cohort. DT yielded the worst accuracy (0.790 and 0.727) and AUC values (0.803 and 0.700) in both the internal validation cohort and the external test cohort. GBDT and LR performed better than radiologists. Long diameter was demonstrated to be the same and most important CT feature for GBDT and LR.

CONCLUSIONS

ML classifiers, especially GBDT and LR with high accuracy and strong robustness, were considered to be promising in risk classification of 1-5 cm gastric GISTs based on CT. Long diameter was found the most important feature for risk stratification.

Social Factors as Major Determinants of Rural Development Variation for Predicting Epidemic Vulnerability: A Lesson for the Future

There have been changes in social attitudes in recent years. These changes have been a consequence of a new societal view of the common good, which manifests itself in social responsibility for a clean and healthy environment. The outbreak and spread of the COVID-19 epidemic has highlighted the socio-spatial variation across regions and countries. The epidemic necessitated restrictive measures by state authorities. In the initial period in many countries, the actions of the authorities were identical throughout the country. This was mainly due to a lack of information about the differentiation of areas in relation to the epidemic risk. The aim of the research was to present a model for classifying rural areas taking into account vulnerability to epidemic threats. The model takes into account demographic, social, economic and spatial-environmental development factors. A total of 33 indicators based on public statistics that can be used to determine the area's vulnerability to epidemic threats were identified. The study showed that for Poland, 11 indicators are statistically significant to the developed classification model. The study found that social factors were vital in determining an area's vulnerability to epidemic threats. We include factors such as average number of persons per one apartment, village centers (number), events (number), number of people per facility (cultural center, community center, club, community hall), residents of nursing homes per 1000 inhabitants, and the number of children in pre-school education establishments per 1000 children aged 3-5 years. The research area was rural areas in Poland. The results of the classification and the methods used should be made available as a resource for crisis management. This will enable a better response to threats from other epidemics in the future, and will influence the remodeling of the environment and social behavior to reduce risks at this risk, which has a significant impact on sustainable development in rural areas.

Leveraging data analytics to understand the relationship between restaurants' safety violations and COVID-19 transmission

This paper leverages natural language processing, spatial analysis, and statistical analysis to examine the relationship between restaurants' safety violations and COVID-19 cases. We used location-based consumers' complaints data during the early stage of business reopening in Florida, USA. First, statistical analysis was conducted to examine the correlation between restaurants' safety violations and COVID-19 transmission. Second, a neural network-based deep learning model was developed to perform topic modeling based on consumers' complaints. Third, spatial modeling of the complaints' geographic distributions was performed to identify the hotspots of consumers' complaints and COVID-19 cases. The results reveal a positive relationship between consumers' complaints about restaurants' safety violations and COVID-19 cases. In particular, consumers' complaints about personal protection measures had the highest correlation with COVID-19 cases, followed by environmental safety measures. Our analytical methods and findings shed light on customers' behavioral shifts and hospitality businesses' adaptive practices during a pandemic.

Spatial modeling of zoonotic cutaneous leishmaniasis with regard to potential environmental factors using ANFIS and PCA-ANFIS methods

This study compares two adaptive neuro-fuzzy inference system (ANFIS) and principal component analysis (PCA)-ANFIS techniques for spatial modeling and forecasting of zoonotic cutaneous leishmaniasis (ZCL) cases in rural districts of Golestan province, Iran. We collected and prepared data on ZCL cases and climatic, topographic, vegetation, and human population factors. By applying the PCA algorithm, the parameters affecting the ZCL incidence were decomposed into principal components (PCs), and their dimensions were reduced. Then, PCs were used to train the ANFIS model. To evaluate the proposed approaches in model assessment phase, we used test data in 2016. In this phase, we showed that PCA-ANFIS model with values ​​of R² = 0.791, MAE = 0.681, RMSE = 0.904 compared to ANFIS model with values ​​of R² = 0.705, MAE = 0.827, RMSE = 1.073 has better performance in prediction of the ZCL cases. Actual and predicted maps of ZCL cases in 2016 by both models demonstrated that the high-risk regions of the disease are located in the northeastern, northern parts, and some central rural districts of Golestan province. Sensitivity analysis of the ANFIS model showed that population, vegetation, average wind speed, elevation, and average soil temperature, respectively, are the most significant factors in predicting the ZCL cases. The findings indicated the importance of machine learning (ML) techniques (ANFIS and PCA-ANFIS) in medical geography studies. By using these approaches, with less cost and shorter time, high-risk areas of diseases can be predicted, and the most effective factors on the spatial prediction of diseases can be identified. Public health policymakers can use these useful tools to control and prevent the disease and to allocate resources to disease-prone areas.

A review of GIS methodologies to analyze the dynamics of COVID-19 in the second half of 2020

COVID-19 has infected over 163 million people and has resulted in over 3.9 million deaths. Regarding the tools and strategies to research the ongoing pandemic, spatial analysis has been increasingly utilized to study the impacts of COVID-19. This article provides a review of 221 scientific articles that used spatial science to study the pandemic published from June 2020 to December 2020. The main objectives are: to identify the tools and techniques used by the authors; to review the subjects addressed and their disciplines; and to classify the studies based on their applications. This contribution will facilitate comparisons with the body of work published during the first half of 2020, revealing the evolution of the COVID-19 phenomenon through the lens of spatial analysis. Our results show that there was an increase in the use of both spatial statistical tools (e.g., geographically weighted regression, Bayesian models, spatial regression) applied to socioeconomic variables and analysis at finer spatial and temporal scales. We found an increase in remote sensing approaches, which are now widely applied in studies around the world. Lockdowns and associated changes in human mobility have been extensively examined using spatiotemporal techniques. Another dominant topic studied has been the relationship between pollution and COVID-19 dynamics, which enhance the impact of human activities on the pandemic's evolution. This represents a shift from the first half of 2020, when the research focused on climatic and weather factors. Overall, we have seen a vast increase in spatial tools and techniques to study COVID-19 transmission and the associated risk factors.

Spatial Modeling of COVID-19 Vaccine Hesitancy in the United States

Vaccine hesitancy refers to delay in acceptance or refusal of vaccines despite the availability of vaccine services. Despite the efforts of United States healthcare providers to vaccinate the bulk of its population, vaccine hesitancy is still a severe challenge that has led to the resurgence of COVID-19 cases to over 100,000 people during early August 2021. To our knowledge, there are limited nationwide studies that examined the spatial distribution of vaccination rates, mainly based on the social vulnerability index (SVI). In this study, we compiled a database of the percentage of fully vaccinated people at the county scale across the continental United States as of 29 July 2021, along with SVI data as potential significant covariates. We further employed multiscale geographically weighted regression to model spatial nonstationarity of vaccination rates. Our findings indicated that the model could explain over 79% of the variance of vaccination rate based on Per capita income and Minority (%) (with positive impacts), and Age 17 and younger (%), Mobile homes (%), and Uninsured people (%) (with negative effects). However, the impact of each covariate varied for different counties due to using separate optimal bandwidths. This timely study can serve as a geospatial reference to support public health decision-makers in forming region-specific policies in monitoring vaccination programs from a geographic perspective.

Burden of Respiratory Infection and Tuberculosis Among US States from 1990 to 2019

PURPOSE

To estimate the incidence, death, disability-adjusted life years (DALYs) and attributable risk factors for respiratory infection and tuberculosis (RIT) in the US from 1990 to 2019.

METHODS

Following the methodology framework and analytical strategies used in the Global Burden of Disease Study 2019, the incidence, death, DALYs and risk factors of RIT were examined by age, gender and states from 1990 to 2019 in the US. All estimates were calculated as counts, age-standardized rates per 100,000 people and percentage change, with 95% confidence intervals (CIs).

RESULTS

In 2019, the age-standardized incidence, death and DALY rates per 100,000 people of RIT were 339,703 (95% CI 303,184 to 382,354), 13.6 (95% CI 12.2 to 14.4) and 384.9 (95% CI 330.6 to 458.6), respectively. Among RIT causes, upper respiratory infection accounted for the large majority of RIT age-standardized incidence rate, while lower respiratory infection constituted the highest proportion of RIT age-standardized death and DALY rates. The age-standardized incidence, death and DALY rates of RIT in 2019 and their temporal trends since 1990 varied widely across states and socio-demographic index. Among all attributable risk factors, smoking was the leading one for age-standardized RIT deaths in 2019, followed by low temperature and alcohol use (the attributable fractions were 17.7%, 15.3% and 6.9%, respectively).

CONCLUSION

Our results suggest that RIT remained a major cause of health burden in the US, with large disparities persisting between US states. Intervention efforts for RIT hotspots, high-risk populations and modifiable risk factors are necessary.

Spatial statistical analysis of pre-existing mortalities of 20 diseases with COVID-19 mortalities in the continental United States

BACKGROUND

Although the United States is among the countries with the highest mortalities of COVID-19, inadequate geospatial studies have analyzed the disease mortalities across the nation.

METHODS

In this county-level study, we investigated age-adjusted co-mortalities of 20 diseases, including cardiovascular, cancer, drug and alcohol disorder, respiratory and infectious diseases with COVID-19 over the first ten months of epidemic. One-way analysis of variance was applied to the Local Moran's I classes (High-High and Low-Low clusters, and non-significant counties of COVID-19) to examine whether the mean mortality measures of covariates that fall into the classes are significantly different. Moreover, a mixed-effects multinomial logistic regression model was employed to estimate the effects of mortalities on COVID-19 classes.

RESULTS

Results showed that the distribution of COVID-19 case fatality ratio (CFR) and mortality rate co-occurrence of High-High clusters were mainly concentrated in Louisiana, Connecticut, and New Jersey. Also, positive associations were observed between High-High cluster of COVID-19 CFR and Asthma (OR = 4.584, 95 % Confidence Interval (CI): 2.583-8.137), Hepatitis (OR = 5.602, CI: 1.265-24.814) and Leukemia (OR = 2.172, CI: 1.518-3.106) mortality rates compared to the non-significant counties, respectively.

CONCLUSIONS

Our results indicated that counties with higher mortality of some cancers and respiratory diseases are more vulnerable to fall into clusters of HH COVID-19 CFR. Future vaccine allocation and more medical professionals and treatment equipment should be a priority to those High-High clusters.

German regional variation of acute and high oral corticosteroid use for asthma

OBJECTIVE: 

To improve understanding of real-world asthma treatment and inform physician education, we evaluated regional variation in asthma prevalence and oral corticosteroid (OCS) use across Germany.

METHODS: 

We developed a machine learning gradient-boosted tree model with IMS® Disease Analyzer electronic medical records, which cover 3% of German patients. This model had a 91% accuracy in predicting the presence of asthma and chronic obstructive pulmonary disease. We applied the model to the IMS^® Longitudinal Prescription database, with 82% national coverage, to classify patients receiving treatment for airflow obstruction from October 2017-September 2018 in 63 regions in Germany.

RESULTS: 

Of 2.4 million individuals under statutory health insurance predicted to have asthma, 13.7%, 18.7%, 36.5%, 29.4%, and 1.7% received treatment classified as Global Initiative for Asthma (GINA) Steps 1, 2, 3, 4, and 5, respectively. Approximately 7-15% of those at GINA Steps 1-4 and 35% at Step 5 treatment received ≥1 acute OCS prescription (duration <10 days). Of patients receiving GINA Steps 1-4 and Step 5 treatments, 1-3% and 86%, respectively, received ≥1 high-dosage OCS prescription. Cumulative OCS dosage and percentages of patients receiving OCS differed substantially across regions, and regions with lower OCS use had greater use of biologic therapies.

CONCLUSIONS: 

Both acute and high OCS use varied regionally across Germany, with overall use suggesting patients are considerable risk of adverse effects and long-term health consequences.

Supplemental data for this article can be accessed at publisher's website.