Early detection of disease outbreaks and non-outbreaks using incidence data (preprint)

Forecasting the occurrence and absence of novel disease outbreaks is essential for disease management. Here, we develop a general model, with no real-world training data, that accurately forecasts outbreaks and non-outbreaks. We propose a novel framework, using a feature-based time series classification method to forecast outbreaks and non-outbreaks. We tested our methods on synthetic data from a Susceptible-Infected-Recovered model for slowly changing, noisy disease dynamics. Outbreak sequences give a transcritical bifurcation within a specified future time window, whereas non-outbreak (null bifurcation) sequences do not. We identified incipient differences in time series of infectives leading to future outbreaks and non-outbreaks. These differences are reflected in 22 statistical features and 5 early warning signal indicators. Classifier performance, given by the area under the receiver-operating curve, ranged from 0.99 for large expanding windows of training data to 0.7 for small rolling windows. Real-world performances of classifiers were tested on two empirical datasets, COVID-19 data from Singapore and SARS data from Hong Kong, with two classifiers exhibiting high accuracy. In summary, we showed that there are statistical features that distinguish outbreak and non-outbreak sequences long before outbreaks occur. We could detect these differences in synthetic and real-world data sets, well before potential outbreaks occur.

Advancing Real-time Pandemic Forecasting Using Large Language Models: A COVID-19 Case Study (preprint)

Forecasting the short-term spread of an ongoing disease outbreak is a formidable challenge due to the complexity of contributing factors, some of which can be characterized through interlinked, multi-modality variables such as epidemiological time series data, viral biology, population demographics, and the intersection of public policy and human behavior. Existing forecasting model frameworks struggle with the multifaceted nature of relevant data and robust results translation, which hinders their performances and the provision of actionable insights for public health decision-makers. Our work introduces PandemicLLM, a novel framework with multi-modal Large Language Models (LLMs) that reformulates real-time forecasting of disease spread as a text reasoning problem, with the ability to incorporate real-time, complex, non-numerical information that previously unattainable in traditional forecasting models. This approach, through a unique AI-human cooperative prompt design and time series representation learning, encodes multi-modal data for LLMs. The model is applied to the COVID-19 pandemic, and trained to utilize textual public health policies, genomic surveillance, spatial, and epidemiological time series data, and is subsequently tested across all 50 states of the U.S. Empirically, PandemicLLM is shown to be a high-performing pandemic forecasting framework that effectively captures the impact of emerging variants and can provide timely and accurate predictions. The proposed PandemicLLM opens avenues for incorporating various pandemic-related data in heterogeneous formats and exhibits performance benefits over existing models. This study illuminates the potential of adapting LLMs and representation learning to enhance pandemic forecasting, illustrating how AI innovations can strengthen pandemic responses and crisis management in the future.

Event Detection from Social Media for Epidemic Prediction (preprint)

Social media is an easy-to-access platform providing timely updates about societal trends and events. Discussions regarding epidemic-related events such as infections, symptoms, and social interactions can be crucial for informing policymaking during epidemic outbreaks. In our work, we pioneer exploiting Event Detection (ED) for better preparedness and early warnings of any upcoming epidemic by developing a framework to extract and analyze epidemic-related events from social media posts. To this end, we curate an epidemic event ontology comprising seven disease-agnostic event types and construct a Twitter dataset SPEED with human-annotated events focused on the COVID-19 pandemic. Experimentation reveals how ED models trained on COVID-based SPEED can effectively detect epidemic events for three unseen epidemics of Monkeypox, Zika, and Dengue; while models trained on existing ED datasets fail miserably. Furthermore, we show that reporting sharp increases in the extracted events by our framework can provide warnings 4-9 weeks earlier than the WHO epidemic declaration for Monkeypox. This utility of our framework lays the foundations for better preparedness against emerging epidemics.

Decisional procrastination and problematic mobile phone use: the mediation of anxiety and the moderation of 2019-nCoV traumatic experience (preprint)

Background Decisional procrastination (DP) has an important effect on problematic mobile phone use (PMPU); however, the potential mechanisism and boundary conditions between DP and PMPU remianed to be further explored. This essay studied the mediation of anxiety between DP and PMPU and whether 2019-nCoV traumatic experience moderated the mediation process.Methods A questionnaire was used in this study. A valid sample of 798 college students reported levels of decisional procrastination, problematic mobile phone use, anxiety, and 2019-nCoV traumatic experience.Results The results indicated that decisional procrastination is positively associated with problematic mobile phone use among college students. Anxiety served as a partial mediator in the association between decisional procrastination and problematic mobile phone use. 2019-nCoV traumatic experience would positively moderated the mediating effects of anxiety between DP and PMPU. A higher degree of 2019-nCoV traumatic experience would strengthen the mediation effects of DP to PMPU through anxiety.Conclusions This study deepens our understanding of how DP affects college students' PMPU. It was found in the study that 2019-nCoV traumatic experience positively moderated the mediation effects of anxiety between DP and PMPU. These findings provide universities with a theoretical foundation for preventing PMPU among college students.

Comparing the gestational diabetes mellitus prevalence and oral glucose tolerance test characteristics before and after the COVID-19 outbreak in Hongshan District of Wuhan, China, 2018- 2021 (preprint)

Background: The Coronavirus Disease 2019 (COVID-19) pandemic has significantly impacted the management and prevalence of gestational diabetes mellitus (GDM) among pregnant women worldwide. This study aimed to investigate the effects of the pandemic on GDM prevalence and oral glucose tolerance test (OGTT) characteristics in Hongshan District, Wuhan, China. Methods: We retrospectively analyzed data from 91,932 pregnant women screened for GDM before (January 1, 2018, to December 31, 2019) and after (January 1, 2020, to December 31, 2021) the onset of the COVID-19 pandemic. The study focused on changes in GDM prevalence, OGTT positive rates and glucose value distribution, and the diagnostic performance of OGTT. Results: The prevalence of GDM increased significantly from 14.5% (95% CI, 14.2-14.8%) pre-pandemic to 21.9% (95% CI, 21.5-22.4%) post-pandemic. A notable rise in OGTT positive rates was observed across all time points, with the most significant increase at the 0-hour mark. Regression analysis indicated a significant risk increase for GDM during the pandemic, even after adjusting for age. Diagnostic accuracy of the 0-hour OGTT improved in the pandemic era, with the area under the curve (AUC) rising from 0.78 to 0.79 and sensitivity from 0.56 to 0.58. Median OGTT values at all time points significantly increased post-pandemic, even after adjusting for age, indicating a shift in glucose metabolism among the study population. Conclusion: The COVID-19 pandemic has led to a significant increase in the prevalence of GDM among pregnant women in Hongshan District, Wuhan. This is evidenced by the elevated rates of positive OGTT and altered median glucose values, indicating a shift in glucose metabolism. These findings underscore the profound impact of the pandemic on maternal and neonatal health. They emphasize the imperative for continuous monitoring and the development of updated, localized diagnostic criteria for OGTT to enhance the identification and treatment of GDM during and after global health crises.

Multi-scenario surveillance of respiratory viruses in aerosols with a sub-single molecule spatial resolution (preprint)

Highly sensitive airborne virus monitoring is critical for preventing and containing epidemics. However, the detection of airborne viruses at ultra-low concentrations remains challenging due to the lack of ultra-sensitive methods and easy-to-deployment equipment. Here, we present an integrated microfluidic cartridge that can accurately detect SARS-CoV-2 and various respiratory viruses with a sensitivity of 10 copies/mL. When seamlessly integrated with a high-flow aerosol sampler, our microdevice can achieve a sub-single molecule spatial resolution of 0.83 copies/m3 for airborne virus surveillance. We then designed a series of virus-in-aerosols monitoring systems (RIAMs), including versions of a multi-site sampling RIAMs (M-RIAMs), a stationary real-time RIAMs (S-RIAMs), and a roaming real-time RIAMs (R-RIAMs) for different application scenarios. Using M-RIAMs, we performed a comprehensive evaluation of 210 environmental samples from COVID-19 patient wards, including 30 aerosol samples. The highest positive detection rate of aerosol samples (60%) proved the aerosol-based SARS-CoV-2 monitoring represents an effective method for spatial risk assessment. The detection of 78 aerosol samples in real-world settings via S-RIAMs confirmed its reliability for ultra-sensitive and continuous airborne virus monitoring. Therefore, RIAMs shows the potential as an effective solution for mitigating the risk of airborne virus transmission.

An early warning indicator trained on stochastic disease-spreading models with different noises (preprint)

The timely detection of disease outbreaks through reliable early warning signals (EWSs) is indispensable for effective public health mitigation strategies. Nevertheless, the intricate dynamics of real-world disease spread, often influenced by diverse sources of noise and limited data in the early stages of outbreaks, pose a significant challenge in developing reliable EWSs, as the performance of existing indicators varies with extrinsic and intrinsic noises. Here, we address the challenge of modeling disease when the measurements are corrupted by additive white noise, multiplicative environmental noise, and demographic noise into a standard epidemic mathematical model. To navigate the complexities introduced by these noise sources, we employ a deep learning algorithm that provides EWS in infectious disease outbreak by training on noise-induced disease-spreading models. The indicator's effectiveness is demonstrated through its application to real-world COVID-19 cases in Edmonton and simulated time series derived from diverse disease spread models affected by noise. Notably, the indicator captures an impending transition in a time series of disease outbreaks and outperforms existing indicators. This study contributes to advancing early warning capabilities by addressing the intricate dynamics inherent in real-world disease spread, presenting a promising avenue for enhancing public health preparedness and response efforts.

Domain Adaptation Using Pseudo Labels for COVID-19 Detection (preprint)

In response to the need for rapid and accurate COVID-19 diagnosis during the global pandemic, we present a two-stage framework that leverages pseudo labels for domain adaptation to enhance the detection of COVID-19 from CT scans. By utilizing annotated data from one domain and non-annotated data from another, the model overcomes the challenge of data scarcity and variability, common in emergent health crises. The innovative approach of generating pseudo labels enables the model to iteratively refine its learning process, thereby improving its accuracy and adaptability across different hospitals and medical centres. Experimental results on COV19-CT-DB database showcase the model's potential to achieve high diagnostic precision, significantly contributing to efficient patient management and alleviating the strain on healthcare systems. Our method achieves 0.92 Macro F1 Score on the validation set of Covid-19 domain adaptation challenge.

Advancing COVID-19 Detection in 3D CT Scans (preprint)

To make a more accurate diagnosis of COVID-19, we propose a straightforward yet effective model. Firstly, we analyse the characteristics of 3D CT scans and remove the non-lung parts, facilitating the model to focus on lesion-related areas and reducing computational cost. We use ResNeSt50 as the strong feature extractor, initializing it with pretrained weights which have COVID-19-specific prior knowledge. Our model achieves a Macro F1 Score of 0.94 on the validation set of the 4th COV19D Competition Challenge $\mathrm{I}$, surpassing the baseline by 16%. This indicates its effectiveness in distinguishing between COVID-19 and non-COVID-19 cases, making it a robust method for COVID-19 detection.

Real-word immunogenicity of a recombinant subunit COVID-19 vaccine and effectiveness against omicron infection: a prospective, multi-centre, longitudinal cohort study in Alzheimer's disease patients (preprint)

Introduction Vaccination is an essential strategy against COVID-19 in the current era of emerging variants. This study evaluates the real-world immunogenicity and effectiveness of the recombinant subunit COVID-19 vaccine (Zifivax) in Alzheimer's disease (AD) patients.Methods 249 AD patients were enrolled in a multicentre, longitudinal cohort study. Levels of RBD-IgG, neutralization antibody activity, and cytokines were identified to evaluate the immune responses. Clinical outcomes were assessed within one month following Omicron infection..Results Following three doses, the vaccine induced a robust immune response, elevating neutralizing antibodies and activating T-cells. AD patients exhibited significantly higher humoral immune responses compared to unvaccinated counterparts. Following Omicron infection, unvaccinated patients experienced higher levels of Th1/Th2-type cytokines than vaccinated individuals. Vaccination correlated with increased survival rates and extended survival times after infection..Discussion The findings highlight the vaccine's efficacy in reducing severe illness, and preventing death in AD patients facing Omicron infection.

The impact of COVID-19 pandemic restrictions lifting on the health of the elderly :evidence from a quasi- experiment (preprint)

The lifting of COVID-19 restrictions has brought about positive changes, yet the adaptation of the elderly in the post-pandemic era still faces challenges. The purpose of this study is to investigate the health changes in the elderly after the lifting of COVID-19 restrictions through a quasi-natural Experiment design, to unravel the effects of the lifting of the pandemic restrictions. The article is based on the data of the elderly in China and South Korea from 2020 to 2022 and employs the PSM-DID method for empirical testing to examine the impact of the lifting of pandemic restrictions on the health of the elderly. The results show that: (1) The lifting of the pandemic restrictions significantly improved the physical health of the elderly. (2) The lifting of the pandemic restrictions effectively improved the mental health of the elderly and significantly reduced their depression scores. (3) Heterogeneity tests indicate that the lifting of the pandemic restrictions had a more treatment effect on improving the health of elderly groups that are female, younger, lower-income, and suffering from chronic diseases. The gradual recovery of health in the elderly in the post-pandemic era is an important phenomenon, but more research is needed on the potential health impacts of pandemic lockdown measures to provide information for the fields of public health and elderly health.

End-expiratory lung volumes as a potential indicator for COVID-19 associated acute respiratory distress syndrome: a retrospective study (preprint)

Background End-expiratory lung volume (EELV) has been observed to decrease in acute respiratory distress syndrome (ARDS). Yet, research investigating EELV in patients with COVID-19 associated ARDS (CARDS) remains limited. It is unclear EELV serve as a potential metric for monitoring disease progression and identifying patients with ARDS at increased risk of adverse outcomes.Study Design and Methods: This retrospective study included mechanically ventilated patients with CARDS during the initial phase of epidemic control in Shanghai. EELV was measured within 48 hours post-intubation, followed by regular assessments every 3–4 days. Chest CT scans, performed within a 24-hour window around each EELV measurement, were analyzed using AI software. Differences in patient demographics, clinical data, respiratory mechanics, EELV, and chest CT findings were assessed using linear mixed models (LMM).Results Out of the 38 enrolled patients, 26.3% survived until discharge from the ICU. In the survivor group, EELV, EELV/PBW and EELV/preFRC were significantly higher than those in the non-survivor group (survivor group vs non-survivor group: EELV: 1455 vs 1162 ml, P = 0.049; EELV/PBW: 24.1 vs 18.5 ml/kg, P = 0.011; EELV/preFRC: 0.45 vs 0.34, P = 0.005). Follow-up assessments showed a sustained elevation of EELV/PBW and EELV/preFRC among the survivors. Additionally, EELV exhibited a positive correlation with total lung volume and residual lung volume, while demonstrating a negative correlation with lesion volume determined through chest CT scans analyzed using AI software.Conclusion EELV is a useful indicator for assessing disease severity and monitoring the prognosis of patients with CARDS.

How Contingency Adjusts CSR in Tourism Industry: A Quasi Experiment in China (preprint)

Although numerous organizational researchers have acknowledged that COVID-19 shocks reduced the tourism industry’s financial performance, relevant literature remains scarce. Do tourism firms reduce corporate social responsibility (CSR) investments to decrease costs? The answer is unclear. This study fills the gap between stakeholder and cost stickiness theories. Based on a quasi-natural experiment of listed Chinese tourism companies from 2017 to 2021, the study finds that the COVID-19 shock caused tourism firms to increase strategic and decrease responsive CSR. In addition, tourism firms that adopted cost leadership strategies trimmed responsive CSR more than strategic CSR. Tourism firms with differentiated leadership strategies increased strategic and decreased responsive CSR. Tourism firms with higher levels of political connections increased responsive CSR, while tourism firms with higher organizational resilience increased strategic CSR. At the theoretical level, this study reveals the theoretical mechanism of the COVID-19 epidemic’s shock on tourism firms' adjustment of CSR from the perspective of cost stickiness. On a practical level, it helps inform tourism firms’ decision-making regarding CSR adjustments for sustainable development when they face widespread crisis scenarios.

Simultaneous Detection of SARS-CoV-2 and Six Other Human Coronaviruses by Multiplex PCR and MALDI-TOF MS (preprint)

The outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is challenging the health systems worldwide, and large population testing is a vital step to control this pandemic. Here, we developed a new method (named HCoV-MS), which combines multiplex PCR with matrix-assisted laser desorption/ionization-time of flight mass spectrometry to simultaneously detect and differentiate seven human coronaviruses (HCoVs). The HCoV-MS method had good specificity and sensitivity, with a detection limit of 1-5 copies/reaction. To validate the HCoV-MS method, we tested 151 clinical samples, and the results showed good concordance with real-time PCR. In addition, 41 D614G variants were identified, which were consistent with the sequencing results. This method was also used in EQAE-SARS-COV in 2020, and all the samples were accurately identified. Taken together, HCoV-MS could be used as an effective method for large-scale detection. It was also capable of detecting key single nucleotide polymorphism about variants.

Discovery of orally active SARS-CoV-2 papain-like protease (PLpro) inhibitors (preprint)

Vaccines and first-generation antiviral therapeutics have provided important protection against coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, there remains a need for additional therapeutic options that provide enhanced efficacy and protection against potential viral resistance. The SARS-CoV-2 papain-like protease (PLpro) is one of two essential cysteine proteases involved in viral replication. While inhibitors of the SARS-CoV-2 main protease (Mpro) have demonstrated clinical efficacy, known PLpro inhibitors have to date lacked the inhibitory potency and requisite pharmacokinetics to demonstrate that targeting PLpro translates to in vivo efficacy in a preclinical setting. Herein, we report the machine learning-driven discovery of potent, selective, and orally available SARS-CoV-2 PLpro inhibitors, with lead compound PF-07957472 (4) providing robust efficacy in a mouse-adapted model of COVID-19 infection.