Grey Model Research Articles

Prediction of provincial Digital Economy Development Index based on grey combination forecasting model

PurposeIn many instances, the data exhibits periodic and trend characteristics. However, indices like the Digital Economy Development Index (DEDI), which pertains to science, technology, policy and economy, may occasionally display erratic behaviors due to external influences. Thus, to address the unique attributes of the digital economy, this study integrates the principle of information prioritization with nonlinear processing techniques to accurately forecast rapid and anomalous data.Design/methodology/approachThe proposed method utilizes the new information priority GM(1,1) model alongside an optimized BP neural network model achieved through the gradient descent technique (GD-BP). Initially, the provincial Digital Economic Development Index (DEDI) is derived using the entropy weight approach. Subsequently, the original GM(1,1) time response equation undergoes alteration of the initial value, and the time parameter is fine-tuned using Particle Swarm Optimization (PSO). Next, the GD-BP model addresses the residual error. Ultimately, the prediction outcome of the grey combination forecasting model (GCFM) is derived by merging the findings from both the NIPGM(1,1) model and the GD-BP approach.FindingsUsing the DEDI of Jiangsu Province as a case study, researchers demonstrate the effectiveness of the grey combination forecasting model. This model achieves a mean absolute percentage error of 0.33%, outperforming other forecasting methods.Research limitations/implicationsFirst of all, due to the limited data access, it is impossible to obtain a more comprehensive dataset related to the DEDI of Jiangsu Province. Secondly, according to the test results of the GCFM from 2011 to 2020 and the forecasting results from 2021 to 2023, it can be seen that the results of the GCFM are consistent with the actual development situation, but it cannot guarantee the correctness of the long-term forecasting, so the combination forecasting model is only suitable for short-term forecasting.Originality/valueThis article proposes a grey combination prediction model based on the principles of new information priority and nonlinear processing.

Carbon Emission Optimization for Assembled Buildings Using Interval Grey GERT Modelling and Modify NSGA-III Algorithm in China

Carbon Emission Optimization for Assembled Buildings Using Interval Grey GERT Modelling and Modify NSGA-III Algorithm in China

Study on the Development of Higher Education in China Based on Economic Growth and Technological Advancement

This paper investigates the dynamic relationship between higher education, economic growth, and technological advancement in China. As the country has rapidly expanded its higher education system, challenges such as educational disparities, quality inconsistencies, and the limitations of an exam-oriented framework have become increasingly evident. Utilizing a grey prediction model, the study examines how economic and technological developments have both enhanced and complicated the landscape of higher education, exacerbating urban-rural divides and limiting opportunities for critical thinking and creativity. To address these challenges, the paper proposes strategic interventions, including curriculum reform, increased access to higher education, and the adoption of active learning methodologies. These strategies aim to foster a more equitable, innovative, and responsive educational system that aligns with China's long-term developmental objectives. The findings provide valuable insights for policymakers seeking to navigate the complexities of modernizing China's higher education while ensuring broad-based, sustainable growth.

Distinct comorbidity phenotypes among post-9/11 Veterans with epilepsy are linked to diverging outcomes and mortality risks.

To investigate phenotypes of comorbidity before and after an epilepsy diagnosis in a national cohort of post-9/11 Service Members and Veterans and explore phenotypic associations with mortality. Among a longitudinal cohort of Service Members and Veterans receiving care in the Veterans Health Administration (VHA) from 2002 to 2018, annual diagnoses for 26 conditions associated with epilepsy were collected over 5 years, ranging from 2 years prior to 2 years after the year of first epilepsy diagnosis. Latent class analysis (LCA) was used to identify probabilistic comorbidity phenotypes with distinct health trajectories. Descriptive statistics were used to describe the characteristics of each phenotype. Fine and Gray cause-specific survival models were used to measure mortality outcomes for each phenotype up to 2021. Six distinct phenotypes were identified: (1) relatively healthy, (2) post-traumatic stress disorder, (3) anxiety and depression, (4) chronic disease, (5) bipolar/substance use disorder, and (6) polytrauma. Accidents were the most common cause of death overall, followed by suicide/mental health and cancer, respectively. Each phenotype exhibited unique associations with mortality and cause of death, highlighting the differential impact of comorbidity patterns on patient outcomes. By delineating clinically meaningful epilepsy comorbidity phenotypes, this study offers a framework for clinicians to tailor interventions. Moreover, these data support systems of care that facilitate treatment of epilepsy and comorbidities within an interdisciplinary health team that allows continuity of care. Targeting treatment toward patients with epilepsy who present with specific heightened risks could help mitigate adverse outcomes and enhance overall patient care.

A dynamic multivariate partial grey model based on the traffic flow parameter equation and its application

A dynamic multivariate partial grey model based on the traffic flow parameter equation and its application

The effect of median household income on the prognosis of lung adenocarcinoma: A SEER analysis

Background: Lung cancer is the main cause of death related to malignant tumors. Since cause-specific mortality can guide clinical decision-making, this study employed the Fine–Gray model based on the Surveillance, Epidemiology, and End Results (SEER) database to identify significantly how socio-economic status influences initial treatment decisions and survival outcomes in patients with lung adenocarcinoma. Objective: The aim of this study was to identify the predictors of lung adenocarcinoma. Methods: The U.S. Surveillance, Epidemiology, and End Results (SEER) database was used to identify patients diagnosed with lung adenocarcinoma between 2000 and 2018. Seventeen thousand four hundred forty-one patients with lung adenocarcinoma were subdivided in four socio-economic tertiles, based on median household income. Cox regression modeling explored the relationship between race, surgery, grading, and median household income on survival outcomes. The study also assessed patient demographic characteristics, age at diagnosis, and surgical interventions. Results: Among 17 441 patients with primary lung adenocarcinoma, the age distribution was as follows: less than 45 years (n=202, 1.16%), between 45 and 54 years (n=1121, 6.43%), 55 and 64 years (n=4252, 24.38%), 65 and 74 years (n=6357, 36.45%), 75 and 84 years (n=4426, 25.38%), and more than 84 years (n=1083, 6.2%). The adjusted hazard ratio (aHR) with 95% CI for ages 65–74 years, 75–84 years, and older than or equal to 85 years were 0.25 (0.11,1.29), 0.40 (0.11,1.50), and 0.72 (0.11,2.05), respectively. Multifactorial Cox regression indicated that the aHR for tumor metastasis was 0.93 (0.03, 2.54), and for patients who did not undergo surgery, it was 1.46 (0.03, 4.31). Grade IV patients exhibited the lowest survival rate [0.66(0.11, 1.93)]. A notable correlation existed between median household income and survival, with distinctly lower survival rates observed in low-income groups. Conclusion: Older patients, especially those who did not undergo surgery and had a higher tumor grade, had significantly reduced survival. Moreover, survival rates for black patients in lower-income brackets were worse than those for white patients in the same financial category. Implications for practice: The quality of life of lung cancer patients is affected by low-income family.

Analysis of tourist flow prediction model of rural tourism on the edge of big cities in China

Introduction: With the adoption of the rural rehabilitation strategy in recent years, China’s rural tourist industry has entered a golden age of growth. Due to the lack of management and decision-support systems, many rural tourist attractions in China experience a “tourist overload” problem during minor holidays or Golden Week, an extended vacation of seven or more consecutive days in mainland China formed by transferring holidays during a specific holiday period. This poses a severe challenge to tourist attractions and relevant management departments. Objective: This study aims to summarize the elements influencing passenger flow by examining the features of rural tourist attractions outside China’s largest cities. Additionally, the study will investigate the variations in the flow of tourists. Method: Grey Model (1,1) is a first-order, single-variable differential equation model used for forecasting trends in data with exponential growth or decline, particularly when dealing with small and incomplete datasets. Four prediction algorithms—the conventional GM(1,1) model, residual time series GM(1,1) model, single-element input BP neural network model, and multi-element input BP network model—were used to anticipate and assess the passenger flow of scenic sites. Result: The multi-input BP neural network model and residual time series GM(1,1) model have significantly higher prediction accuracy than the conventional GM(1,1) model and unit-input BP neural network model. A multi-input BP neural network model and the residual time series GM(1,1) model were used in tandem to develop a short-term passenger flow warning model for rural tourism in China’s outskirts. Conclusion: This model can guide tourists to staggered trips and alleviate the problem of uneven allocation of tourism resources.

Beyond Traditional Forecasting: Machine Learning and Adaptive Algorithms in EV Sales Predictions

The Electric Vehicle market is growing fast, creating new challenges for sales predictions and market dynamics. Traditional forecasting methods like linear regression and expert judgment struggle with the market's non-linear and complex nature. To improve accuracy, new models such as machine learning and Adaptive Optimized Grey Models are being applied to predict EV sales, which is discussed in this review paper. These models dynamically adjust parameters and account for external factors, offering better precision. AOGMs, in particular, utilize data preprocessing and buffering to minimize external disturbances and enhance prediction smoothness. Techniques like the Particle Swarm Optimization help these models more effectively capture the EV sales trends. Additionally, adaptive grey models are not only essential for forecasting sales but also for identifying market risks like overcapacity or supply shortages. These advanced methods provide a more robust framework for understanding and anticipating changes in the rapidly evolving the EV market.

A combined mechanistic and empirical models for tailored microwave-assisted extraction of rice bran phenolic

ABSTRACT Rice bran is a promising source of functional food due to its high phenolic content. Microwave-assisted extraction (MAE) has become a widely used method for optimizing phenolic extraction from rice bran. However, understanding the phenolic extraction process using the MAE method is challenging due to the system’s complexity. Modeling and simulation are powerful tools to provide a better understanding of how the system behaves and can be optimized. This study proposes a novel approach that combines mechanistic and empirical models (gray box) to predict the phenolic content in rice bran extraction using MAE. It involved phenolic extraction using different ethanol concentrations (40%, 60%, 80%), microwave application time (1 min, 2 min, 3 min), and microwave power (90 W, 180 W, 270 W). The results showed a good agreement between experimental and computed extraction yields, as indicated by the reported statistical parameters and gray box model. The optimum conditions were reached at 60% ethanol concentration, 2-min microwave application time, and 180 W microwave power with the phenolic content at 0.534 mg/mL (model) and 0.561 mg/mL or 25.06 mg/g (experiment). Following the study, this combined approach can potentially reduce the time and cost required for optimization studies of phenolic extraction.

An innovative method for short-term forecasting of blockchain cryptocurrency price

Cryptocurrency market sentiment is relatively unstable, which makes cryptocurrency price an attribute of high volatility. Accurate forecasting methods help to clarify the volatility trend of the cryptocurrency price, thereby reducing the investment risk of participants in the cryptocurrency market. Therefore, this research proposed a new method for short-term forecasting of the cryptocurrency price based on a small sample. This study took three typical blockchain cryptocurrencies (Bitcoin, Ethereum, Litecoin) as experimental objects, chose data intervals with different volatility trends in the U.S. stock indices between 2022 and 2023 as sample data, and used grey correlation analysis to select core affecting variables. Furthermore, this study built a grey multivariate convolution model with prioritized accumulating novel information for conducting prediction experiments on blockchain cryptocurrency price. The research findings demonstrate that the proposed model achieves high prediction accuracy in all experiments, and the model accuracy is superior to the comparison models. This study proposes a scientific prediction approach for blockchain cryptocurrency price, which can guide financial investors in developing and analyzing quantitative financial trading strategies to a certain extent. Meanwhile, this study provides a specific reference for relevant government departments to strengthen cryptocurrency regulation, prevent financial risks, and maintain financial stability.

Prediction of Energy-Related Carbon Emissions in East China Using a Spatial Reverse-Accumulation Discrete Grey Model

In order to better analyze and predict energy-related carbon emissions in East China to address climate change, this paper enhances the predictive capabilities of grey models in spatial joint prediction by creating the reverse-accumulation spatial discrete grey model RSDGM (1,1,m) and accumulation spatial discrete grey breakpoint model RSDGBM (1,1,m,t), which took the impact of system shocks into consideration. The efficiency of the models is confirmed by calculating the energy-related carbon emissions in East China from 2010 to 2022. Future emissions are predicted, and the spatial spillover effect of emissions in East China is discussed. The conclusions are as follows: (1) The RSDGM (1,1,m) theoretically avoids errors in background values and parameter calculations, reducing computational complexity. Empirically, the model exhibits high performance and reflects the priority of new information in spatial joint analysis. (2) The RSDGBM (1,1,m,t) captures the impact of shocks on system development, improving the reliability of carbon emissions prediction. (3) Jiangsu and Shandong are positively affected by spatial factors in terms of carbon emissions, while Shanghai and Zhejiang are negatively affected. (4) It is estimated that carbon emissions in East China will increase by approximately 23.8% in 2030 compared to the level in 2022, with the levels in Zhejiang and Fujian expected to increase by 45.2% and 39.7%, respectively; additionally, the level in Shanghai is projected to decrease. Overall, East China still faces significant pressure to reduce emissions.

A New Grey Prediction Model and Its Application in Renewable Energy Consumption

Renewable energy is an energy resource that can be used continuously. At present, international oil prices continue to rise, the problem of global climate change is becoming increasingly prominent, and renewable energy and clean energy have ushered in a new round of development opportunities. Based on the new gray prediction model, this paper forecasts the consumption of renewable energy and further analyzes the sustainable development of renewable energy. In this paper, the combinatorial optimization method of cumulative order, background value coefficient, and initial conditions, parameter optimization combination, parameter combinatorial optimization process of the gray prediction model, and parameter optimization mechanism based on the PSO algorithm are proposed, and the reduction error analysis is carried out. The consumption of wind power and photovoltaic renewable energy is forecasted, and three different forecasting methods as exponential smoothing method, time series analysis method, and new gray forecasting method are compared, and the wind speed, irradiation intensity, and load are forecasted by these three different forecasting methods. Compared with the time series analysis method and the exponential smoothing method, the RMSE of the new grey prediction method is reduced by 127.12% and 160.59%, and the error rate is reduced by 3.16% and 4%, respectively. Based on the consumption forecast of renewable energy, this paper analyzes its sustainability from three directions economy, resource supply, and environment, and finally gives energy policy recommendations.

Short-term hybrid prognostics of fuel cells: A comparative and improvement study

Accurately predicting the performance degradation trend of fuel cells helps take measures in advance and prolong the stack's service life, leading to a novel hybrid forecasting approach. The first grey model prediction method based on residual exponential smoothing optimization (ES-R-GM) can capture the voltage deterioration trend. We explore the integration of two different ES techniques, specifically the double ES (ES2) and cubic ES (ES3), to investigate their effect on enhancing the predictive accuracy of the GM model. The second method of the adaptive network fuzzy inference system (ANFIS) can characterize local nonlinear behavior. We utilize the simulated annealing (SA) algorithm to optimize ANFIS results under different fuzzy rule selection strategies. The outcomes of the two prediction methods mentioned above are combined to create a hybrid prediction using the data fusion method and the moving window technique. Various hybrid methods are evaluated under general conditions and further detailed optimization. The data collected from the experimental platform confirms the suggested hybrid framework. The results show that the hybrid ES3-R-GM + ANFIS-SC method outperformed the single models in final prediction accuracy and can effectively track both global trends and local changes. Simultaneously, it takes less time to calculate than the literature. Moreover, when applied to consistent public datasets, the hybrid approach maintains its robustness and accuracy compared with other hybrid prognostic methods.

Spatiotemporal Variations in Gross Ecosystem Product and Its Relationship with Economic Growth in Ecologically Vulnerable Watershed Areas: A Case Study of Yongding River Basin

Ecosystem service value is crucial for balancing economic growth and ecological preservation in ecologically vulnerable watershed areas. Although Gross Ecosystem Product (GEP) has received significant attention, most existing studies have focused on how to measure it. Few studies have explored spatiotemporal variations in GEP and how land-use changes affect these variations regarding ecological restoration at the river basin level. Additionally, while many studies have examined the relationship between ecosystem service value and economic growth, there is little research on how components of GEP influence economic growth. Analyzing the spatiotemporal structure of GEP and its components could offer new insights into optimizing ecological restoration strategies and promoting sustainable development in vulnerable watershed regions. In this study, we used ArcGIS, InVEST, SPSS, and Python to analyze spatiotemporal variations in GEP in the Yongding River Basin within the Beijing–Tianjin–Hebei Economic Region from 1995 to 2020. Moran’s Index and variance decomposition were applied to analyze the spatiotemporal structure. The grey prediction model forecasted GEP trends from 2025 to 2035. The random forest model was used to assess land-use changes’ impacts on GEP. Paired T-tests were used to compare GEP and GDP, and a dynamic panel model was used to examine how ecosystem service value factors influenced economic growth. The results show the following: (1) Regarding values, GEP accounting and variance decomposition results indicated that ecosystem cultural service value (ECV) and ecosystem regulating service value (ERV) each contributed about half of the total GEP. Ecosystem provisioning service value (EPV) showed an upward trend with fluctuations. Regarding the spatial distribution, Moran’s I analysis showed significant positive spatial correlations for EPV and ERV. The grey prediction model results indicated significant growth in GEP from 2025 to 2035 under current ecological restoration policies, especially for ERV and ECV. (2) In terms of the influence of land-use changes, random forest analysis showed that the forest land area was consistently the most influential factor across GEP, EPV, and ERV. Unused land area was identified as the most significant factor for ECV. (3) Before 2010, GEP was larger than GDP, with significant differences between 1995 and 2000. From 2010 onwards, GDP surpassed GEP, but the differences were not statistically significant. Dynamic panel regression further showed that the water conservation value significantly boosted GDP, whereas the water purification value significantly reduced it. This study highlights the importance of integrating GEP into ecological restoration and economic development to ensure the sustainability of ecologically vulnerable watershed areas.

Regional agricultural drought vulnerability prediction based on interpretable Random Forest.

Drought is one of the leading natural disasters in the world. It severely restricts the normal functioning of societies and economies. The purpose of this paper is to predict the drought vulnerability index in Henan Province so that timely measures can be taken to mitigate the impact of drought and to ensure the continued stability of people's lives and social development. First, the drought vulnerability index of Henan Province for 2010-2022 was calculated based on relevant literature and the regional situation of Henan Province. In addition, the main factors of sensibility and resilience are selected, the prediction indicator system is established, and the grey correlation analysis model is applied to select the top 10 main factors of drought vulnerability. Second, the Random Forest model was trained and simulated, and the drought vulnerability index of Henan Province was predicted for 2023-2025. Finally, the SHapley Additive exPlanations model was used to explain the specific effects of the influences of the Random Forest model. The interpretable Random Forest model not only accurately predicts the drought vulnerability index in Henan Province but also alleviates concerns about the "black box" problem of indirect interpretation of the Random Forest model. The results of the study show that the Average Percentage Error of the drought vulnerability index in Henan Province from 2010 to 2022 simulated by the Random Forest model is as low as 3.0457%, and the simulation accuracy is as high as 96.9643%. Drought vulnerability prediction for Henan Province was also made for the period 2023-2025, showing an increasing trend. Relevant sectors should enhance awareness and early warning of drought, improve resilience, and take effective measures to reduce the socio-economic impact of drought.

Height status matters for risk of mortality in critically ill children

BackgroundAnthropometric measurements are crucial in pediatric critical care, but the impact of height on ICU outcomes is underexplored despite a substantial number of short-for-age children in ICUs. Previous studies suggest that short stature increases the risk of poor clinical outcomes. This study examines the relationship between short stature and ICU outcomes.MethodsWe conducted a retrospective cohort study using a Japanese nationwide database (the Japanese Intensive Care Patient Database; JIPAD), which included pediatric patients under 16 years admitted to ICUs from April 2015 to March 2020. Height standard deviation scores (SD scores) were calculated based on age and sex. Short-stature patients were defined as height SD score < − 2. The primary outcome was all-cause ICU mortality, and the secondary outcome was the length of stay in ICU.ResultsOut of 6,377 pediatric patients, 27.2% were classified as having short stature. The ICU mortality rate was significantly higher in the short-stature group compared to the normal-height group (3.6% vs. 1.4%, p < 0.01). Multivariable logistic regression showed that short stature was independently associated with increased ICU mortality (OR = 2.73, 95% CI 1.81–4.11). Additionally, the Fine–Gray subdistribution hazards model indicated that short stature was associated with a lower chance of ICU discharge for each additional day (HR 0.85, 95% CI 0.81–0.90, p < 0.01).ConclusionsShort stature is a significant risk factor for increased ICU mortality and prolonged ICU stay in critically ill children. Height should be considered in risk assessments and management strategies in pediatric intensive care to improve outcomes.

A novel dynamic risk score to predict venous thromboembolism (VTE) recurrences after 3 months of anticoagulation in patients with incident VTE and without active cancer

Abstract Background Predicting the risk of recurrent venous thromboembolism (rVTE) during and following anticoagulation is complex. Current models lack the ability to adapt to changing patient conditions over time and fail to account for the direct impact of anticoagulation in real-world clinical settings. Purpose To develop a dynamic risk prediction model for rVTE to help clinicians decide on the duration of anticoagulant therapy in conjunction with a model for bleeding (also submitted to the conference). Methods UK Clinical Practice Research Datalink data (2001-2020) was used to generate a retrospective cohort with first VTE who had received 3 months of anticoagulation. Patient episodes of rVTE were evaluated. Covariates were collected at baseline and subsequently as time-varying data to capture changes post-VTE. Hazards with 95% confidence intervals (CI) were estimated and covariates included in a Fine &amp; Gray model used as predictors of rVTE. An additive (logarithmic) scoring scheme was developed from subdistribution hazard ratios, discrimination (expressed by the C-statistic) estimated from 10-fold cross-validation. Results 51,465 patients with a first VTE were included; 4041 rVTE were identified in 200,698 person-years of observation. Incidence rates for rVTE were 2.01 per 100 person-years. Nineteen independent predictors of rVTE (recorded before or after the VTE diagnosis) were included in the model, 13 associated with increased risk and 6 with decreased risk, Table 1. Patients recognised as lower-risk, medium and higher-risk (10.5%, 37% and 52.5% of the population, respectively, at 90 days after first VTE, assuming no anticoagulation treatment) had an annualised rVTE incidence rate of 2.20, 3.84, and 7.44 per 100 person-years. The C-statistic for rVTE was 0.65 (95%CI, 0.64-0.66). The points scored can be added or subtracted to predict risk (Table 2). Conclusions Our dynamic risk score effectively identifies patients at risk of rVTE three months post their initial VTE diagnosis. It allows for continuous monitoring of risk, and modelling of treatment impact, providing clinicians with a pragmatic tool to help determine treatment duration, and adapt their strategies in response to evolving patient conditions.Table 1:Predictors for VTE recurrenceTable 2:VTE recurrence risk score

A novel dynamic risk score to predict Clinically Significant Bleeding (CSB) after 3 months of anticoagulation in patients with incident Venous ThromboEmbolism (VTE) and without active cancer

Abstract Background Predicting the risk of clinically significant bleeding (CSB), bleeding requiring hospitalisation or fatal bleeding, during and following anticoagulation is complex. Current models lack the ability to adapt to changing patient conditions over time and fail to account for the direct impact of anticoagulation in real-world clinical settings. Purpose To develop a dynamic risk prediction model for CSB to help clinicians decide on the duration of anticoagulant therapy in conjunction with a model for recurrences of VTE (also submitted to the conference). Methods UK Clinical Practice Research Datalink data (2001-2020) was used to generate a retrospective cohort with first VTE who had received 3 months of anticoagulation. Patient episodes of CSB and recurrent VTE were evaluated. Covariates were collected at baseline and subsequently as time-varying data to capture changes post-VTE. Hazards with 95% confidence intervals (CI) were estimated and covariates included in a Fine &amp; Gray model used as predictors of CSB. An additive (logarithmic) scoring scheme was developed from subdistribution hazard ratios, discrimination (expressed by the C-statistic) estimated from 10-fold cross-validation. Results 51,621 patients with a first VTE were included; 3307 CSB were identified in 206,292 person-years of observation. Incidence rates for CSB were 1.60 per 100 person-years. 18 independent predictors of CSB (recorded before or after the VTE diagnosis) were included in the model, Table 1. Patients recognised as lower-risk, medium and higher-risk (67.5 %, 15.9 % and 16.7 % of the population, respectively, at 90 days after first VTE, assuming no anticoagulation treatment) had an annualised incidence rate of 0.84, 2.94, and 9.45 per 100 person-years. The C-statistic for CSB was 0.78 (95%CI, 0.77-0.79). The points scored can be added to predict risk (Table 2). Conclusions Our dynamic score effectively identifies patients at risk of CSB three months post their initial VTE diagnosis. It allows for continuous monitoring of risk, and modelling of treatment impact, providing clinicians with a pragmatic tool to help determine treatment duration, and adapt their strategies in response to evolving patient conditions.Table 1:Predictors for CSBTable 2:CSB risk score

Impact of resumption of Oral AntiCoagulation therapy (OAC) after GastroIntestinal Bleeding (GIB) on clinical outcomes in patients with Non-Valvular Atrial Fibrillation (NVAF)

Abstract Purpose Optimal resumption of oral anticoagulation therapy (OAC) after gastrointestinal bleeding (GIB)has not been well established. This study aimed to compare clinical outcomes between patients who resumed or discontinued OAC after index GIB. Methods We retrospectively enrolled consecutive patients with non-valvular atrial fibrillation (NVAF) and prior GIB from January 2018 to July 2022 in 16 public hospitals in Hong Kong. Resumption of OAC was substantiated by new prescription after index GIB on electronic medical records. Patients were divided into 5 groups: (i) OAC naïve (as reference group); (ii) resume direct oral anticoagulant (DOAC); (iii) resume Vitamin K antagonist (VKA); (iv) discontinue DOAC; (v) discontinue VKA. Endpoints included recurrent GIB, ischemic stroke and all-cause death. Risks of target events were estimated for all groups and compared with individuals without OAC, using adjusted sub-distribution hazard ratios (aSHR) derived from Fine and Gray regression models, accounting for death as competing risk for recurrent GIB and stroke, and COX proportional regression for all-cause mortality, adjusting for components of CHA2DS2VASC, HASBLED scores, causes of index GIB (i.e., GI ulcer, diverticular disease, angiodysplasia or undermined) and endoscopy intervention. Among patients who resumed OAC, outcomes were compared between different DOAC agents (i.e, apixaban, dabigatran, edoxaban, rivaroxaban), using VKA as reference. Results Of 2,350 patients identified (mean age 80.5±10.4, female 54.5%), 44.0% (n=1,034) were OAC naïve, 4.4% (n=103) discontinued VKA, 11.7% (n=275) discontinued DOAC, 11.2% (n=263) resumed VKA, and 28.7% (n=675) resumed DOAC. During a median follow-up duration of 1.2(IQR:0.2-2.5) years, discontinuation of DOAC (aSHR=0.5(0.29-0.9) was associated with lower risk of recurrent GIB, whereas resumption of VKA (aSHR=1.8(1.0-3.1)) was associated with the highest risk of recurrent GIB (Figure 1a). Among DOACs, apixaban (aSHR=0.5(0.3-0.9)) was associated with the lowest risk of recurrent GIB (Figure 1b). Risk of ischemic stroke across groups were similar. Patients who discontinued DOAC (aHR=1.4(1.2-1.6)) were at higher risk of all-cause mortality, whereas resumption of VKA (aHR=0.4(0.3-0.6)) and DOAC (aHR=0.4(0.4-0.5)) were both associated with decreased mortality (Figure 1c). Conclusion In this real-world dataset, resumption of VKA in NVAF patients after GIB was associated with high risk of recurrent GIB and lowest risk with apixaban. OAC resumption either VKA or DOAC was associated with reduced mortality.

Resumption of anticoagulation therapy among Non-Valvular Atrial Fibrillation (NVAF) patients with prior intracranial haemorrhage

Abstract Purpose Resumption of anticoagulation therapy (OAC) in patients with non-valvular atrial fibrillation (NVAF) suffering from intracranial haemorrhage (ICH) is challenging in clinical practice. The aim of this study was to evaluate impact of resumption of OAC on clinical outcomes in NVAF patients with history of ICH. Methods Consecutive patients with NVAF with or without OAC who survived from an index ICH between January 2018 and July 2022 in 16 public hospitals in Hong Kong were identified and analyzed retrospectively. Resumption of OAC was substantiated by new prescription obtained from electronic medical records. Patients were divided into 5 groups: (i) OAC naïve (as reference group); (ii) resume direct oral anticoagulant (DOAC); (iii) resume Vitamin K antagonist (VKA); (iv) discontinue DOAC; (v) discontinue VKA. Risks of ischemic stroke, recurrent ICH and all-cause mortality were estimated for all groups and compared with individuals without OAC, using adjusted sub-distribution hazard ratios (aSHR) derived from Fine and Gray regression models, accounting for death as competing risk, adjusting for components of CHA2DS2VASC and HASBLED scores. Sub-analysis was conducted to compare outcomes between different DOAC agents (i.e, apixaban, dabigatran, edoxaban, rivaroxaban), using VKA as reference among patients who resumed DOAC or VKA. Results Of 982 patients who met inclusion criteria (mean age 77.7±10.9, female 40.9%), 39.1% (n=384) were OAC naive, 32.2% (n=316) resumed DOAC, 12.6% (n=124) resumed VKA, 9.0% (n=88) discontinued DOAC and 7.1% (n=70) discontinued VKA. During a median follow-up period of 2.0 (IQR: 0.7-3.5) years, no statistically significant difference in risk of ischemic stroke was observed among 5 groups. Discontinuation of DOAC was associated with markedly higher risk of ischemic stroke (aSHR=8.32 (1.8-39.3)) only among AF patients with CHA2DS2VASC score ≥ 4, Risk of recurrent ICH was higher in patients who resumed VKA (aSHR=1.6(1.0-2.5)). Sub-analysis demonstrated resumption of apixaban was associated with lowest risk of recurrent ICH (aSHR=0.6(0.4-0.9)). Compared to OAC naïve patients, those who resumed DOAC (aSHR=0.5(0.4-0.7)) had lowest risk of all-cause mortality, whereas discontinuation DOAC (aSHR=1.9(0.4-2.5) and VKA (aSHR=1.6(0.2-2.2)) were both associated with increased mortality. Conclusion Our real-world data revealed among NVAF patients with prior ICH long term discontinuation of OAC was associated with increased risk of ischemic stroke. Recurrent ICH was highest among patients who resumed VKA and lowest with apixaban.