Exploration of spatiotemporal heterogeneity and socio-demographic determinants on COVID-19 incidence rates in Sarawak, Malaysia

BACKGROUND

In Sarawak, 252 300 coronavirus disease 2019 (COVID-19) cases have been recorded with 1 619 fatalities in 2021, compared to only 1 117 cases in 2020. Since Sarawak is geographically separated from Peninsular Malaysia and half of its population resides in rural districts where medical resources are limited, the analysis of spatiotemporal heterogeneity of disease incidence rates and their relationship with socio-demographic factors are crucial in understanding the spread of the disease in Sarawak.

METHODS

The spatial dependence of district-wise incidence rates is investigated using spatial autocorrelation analysis with two orders of contiguity weights for various pandemic waves. Nine determinants are chosen from 14 covariates of socio-demographic factors via elastic net regression and recursive partitioning. The relationships between incidence rates and socio-demographic factors are examined using ordinary least squares, spatial lag and spatial error models, and geographically weighted regression.

RESULTS

In the first 8 months of 2021, COVID-19 severely affected Sarawak's central region, which was followed by the southern region in the next 2 months. In the third wave, based on second-order spatial weights, the incidence rate in a district is most strongly influenced by its neighboring districts' rate, although the variance of incidence rates is best explained by local regression coefficient estimates of socio-demographic factors in the first wave. It is discovered that the percentage of households with garbage collection facilities, population density and the proportion of male in the population are positively associated with the increase in COVID-19 incidence rates.

CONCLUSION

This research provides useful insights for the State Government and public health authorities to critically incorporate socio-demographic characteristics of local communities into evidence-based decision-making for altering disease monitoring and response plans. Policymakers can make well-informed judgments and implement targeted interventions by having an in-depth understanding of the spatial patterns and relationships between COVID-19 incidence rates and socio-demographic characteristics. This will effectively help in mitigating the spread of the disease.

A methodology framework for bipartite network modeling

The graph-theoretic based studies employing bipartite network approach mostly focus on surveying the statistical properties of the structure and behavior of the network systems under the domain of complex network analysis. They aim to provide the big-picture-view insights of a networked system by looking into the dynamic interaction and relationship among the vertices. Nonetheless, incorporating the features of individual vertex and capturing the dynamic interaction of the heterogeneous local rules governing each of them in the studies is lacking. The methodology in achieving this could hardly be found. Consequently, this study intends to propose a methodology framework that considers the influence of heterogeneous features of each node to the overall network behavior in modeling real-world bipartite network system. The proposed framework consists of three main stages with principal processes detailed in each stage, and three libraries of techniques to guide the modeling activities. It is iterative and process-oriented in nature and allows future network expansion. Two case studies from the domain of communicable disease in epidemiology and habitat suitability in ecology employing this framework are also presented. The results obtained suggest that the methodology could serve as a generic framework in advancing the current state of the art of bipartite network approach.

Development and user testing study of MozzHub: a bipartite network-based dengue hotspot detector

Traditionally, dengue is controlled by fogging, and the prime location for the control measure is at the patient's residence. However, when Malaysia was hit by the first wave of the Coronavirus disease (COVID-19), and the government-imposed movement control order, dengue cases have decreased by more than 30% from the previous year. This implies that residential areas may not be the prime locations for dengue-infected mosquitoes. The existing early warning system was focused on temporal prediction wherein the lack of consideration for spatial component at the microlevel and human mobility were not considered. Thus, we developed MozzHub, which is a web-based application system based on the bipartite network-based dengue model that is focused on identifying the source of dengue infection at a small spatial level (400 m) by integrating human mobility and environmental predictors. The model was earlier developed and validated; therefore, this study presents the design and implementation of the MozzHub system and the results of a preliminary pilot test and user acceptance of MozzHub in six district health offices in Malaysia. It was found that the MozzHub system is well received by the sample of end-users as it was demonstrated as a useful (77.4%), easy-to-operate system (80.6%), and has achieved adequate client satisfaction for its use (74.2%).

Assessing fitness-for-purpose and comparing the suitability of COVID-19 multi-country models for local contexts and users

Background: Mathematical models have been used throughout the COVID-19 pandemic to inform policymaking decisions. The COVID-19 Multi-Model Comparison Collaboration (CMCC) was established to provide country governments, particularly low- and middle-income countries (LMICs), and other model users with an overview of the aims, capabilities and limits of the main multi-country COVID-19 models to optimise their usefulness in the COVID-19 response. Methods: Seven models were identified that satisfied the inclusion criteria for the model comparison and had creators that were willing to participate in this analysis. A questionnaire, extraction tables and interview structure were developed to be used for each model, these tools had the aim of capturing the model characteristics deemed of greatest importance based on discussions with the Policy Group. The questionnaires were first completed by the CMCC Technical group using publicly available information, before further clarification and verification was obtained during interviews with the model developers. The fitness-for-purpose flow chart for assessing the appropriateness for use of different COVID-19 models was developed jointly by the CMCC Technical Group and Policy Group. Results: A flow chart of key questions to assess the fitness-for-purpose of commonly used COVID-19 epidemiological models was developed, with focus placed on their use in LMICs. Furthermore, each model was summarised with a description of the main characteristics, as well as the level of engagement and expertise required to use or adapt these models to LMIC settings. Conclusions: This work formalises a process for engagement with models, which is often done on an ad-hoc basis, with recommendations for both policymakers and model developers and should improve modelling use in policy decision making.

Modelling the Effectiveness of Epidemic Control Measures in Preventing the Transmission of COVID-19 in Malaysia

Malaysia is currently facing an outbreak of COVID-19. We aim to present the first study in Malaysia to report the reproduction numbers and develop a mathematical model forecasting COVID-19 transmission by including isolation, quarantine, and movement control measures. We utilized a susceptible, exposed, infectious, and recovered (SEIR) model by incorporating isolation, quarantine, and movement control order (MCO) taken in Malaysia. The simulations were fitted into the Malaysian COVID-19 active case numbers, allowing approximation of parameters consisting of probability of transmission per contact (β), average number of contacts per day per case (ζ), and proportion of close-contact traced per day (q). The effective reproduction number (Rt) was also determined through this model. Our model calibration estimated that (β), (ζ), and (q) were 0.052, 25 persons, and 0.23, respectively. The (Rt) was estimated to be 1.68. MCO measures reduce the peak number of active COVID-19 cases by 99.1% and reduce (ζ) from 25 (pre-MCO) to 7 (during MCO). The flattening of the epidemic curve was also observed with the implementation of these control measures. We conclude that isolation, quarantine, and MCO measures are essential to break the transmission of COVID-19 in Malaysia.

A Preliminary Study on the Cost Implications and Validity of Executing a Control Strategy on Individual Appliance Usage

This study evaluates a number of power-saving measures that were applied on individual electric appliances to investigate their contributions to energy savings of the electricity network. In order to carry out the study, power-saving measures were applied on all appliances under study for certain period periods at the Universiti Malaysia Sarawak (UNIMAS). The appliances are the air conditioner (AC), computers, lightings and closed-circuit television (CCTV). As a means of validating the accuracy of models developed for electricity costs, a comparison of was done between measurements taken from the electricity network and those taken from PowerLogic PM5350 power meter (PM5350) installed for the purpose of this research. The results from model analysis show significant cost savings of 39.9%, 20.3%, 8% and 0.6% when control strategies were applied to AC, lightings, computers and CCTV

A Predictive Framework for Electricity Consumption

This study investigates the performance of regression model, Kalman filter adaptation algorithm and artificial neural network to assess their qualities for predictions. It develops predictive algorithms based on price, temperature and humidity as multiple variables affecting time-varying aspect of electricity consumption. In order to meet energy demand through the use of electricity as an energy source for daily activities in buildings such as air conditioning, lighting, computers and cooking stoves., adequate allocation of energy resources and planning should be done, including predicting for electricity consumption. The process involves collecting data from the power grid of Faculty of Computer Science and Information Technology building, Universiti Malaysia Sarawak. The forecasting techniques were tested on the data collected, and the dataset consists of electricity consumption readings, with electricity price, humidity and temperature included in the forecasting model. The performances of regression model, artificial neural network and Kalman algorithm were tested using statistical evaluation parameters, root mean squared error (RMSE) and mean absolute percentage error (MAPE); while the parameter, standard deviation, was used to check the validity of models. This study identified Kalman algorithm as the most effective method of predicting consumption data compared to regression model, and artificial neural network.

HUMAN BEHAVIORAL CHANGES AND ITS IMPACT IN DISEASE MODELING

One of the threats of the world health is the infectious diseases. This leads to the raise of concern of the policymakers and disease researchers. Vaccination program is one of the methods to prevent the vaccine-preventable diseases and hence help to eradicate the diseases. The impact of the preventive actions is related to the human behavioral changes. Fear of the diseases will increase one’s incentive in taking the preventive actions to avoid the diseases. As human behavioral changes affecting the impact of the preventive actions, the individual-based model is constructed to incorporate the behavioral changes in disease modeling. The agents in the individual-based model are allowed to move randomly and interact with each other in the environment. The interactions will cause the disease viruses as well as the fearfulness to be spread in the population. In addition, the individual-based model can have different environment setups to distinguish the urban and rural areas. The results shown in this paper are divided into two subsections, which are the justification of using uniform distribution as random number generator, and the variation of disease spread dynamics in urban and rural areas. Based on the results, the uniform distribution is found to be sufficient in generating the random numbers in this model as there is no extreme outlier reported in the experiment. We have hypothesized the individuals in urban area to have higher level of fearfulness compared to those in rural area. However, the preliminary results of the survey conducted show a disagreement with the hypothesis. Nevertheless, the data collected still show two distinct classes of behavior. Thus, the distinction does not fall into the samples taken from rural or urban areas but perhaps more on the demographic factors. Therefore, the survey has to be study again and demographic factors have to be included in the survey as we could not distinguish the level of fearfulness by areas.