Coronavirus disease (COVID-19) is a novel pandemic that affects every other country in the world. Various countries have adopted control measures involving restriction of movement. Several studies have used mathematical modelling to predict the dynamic of this pandemic. Forecasting techniques can be used to predict the incidence cases for the short term. The study aims to forecast the COVID-19 incidence using the Auto Regressive Integrated Moving Average (ARIMA) method. MATERIALS AND METHODS:Using publicly available data, we performed a forecast of Malaysia COVID-19 new cases using Expert Modeler Method in SPSS and ARIMA model in R to predict COVID-19 cases in Malaysia. We compare 3 different time frames based on different Movement Control Order (MCO) period. We compare the model fit and prediction across models.RESULTS:All models show static cases for each MCO 7-day prediction. For prediction until 12 May, the third MCO time frame shows the best model fit for both techniques. Both software shows a stationary trend of cases of below 100. CONCLUSION:These MCO models have shown to stabilize the rate of new cases. Further sub analysis and quality of data is needed to improve the accuracy of the model.
Many studies have shown the effectiveness of educational modules on increasing colorectal cancer screening uptake at individual level but not adjusted for potential clustering effect such as workplace. Longitudinal studies on workplace colorectal cancer screening require a series of analysis under different conditions due to heterogeneity of workplace population. To achieve this, a sensitivity analysis based on Generalized Estimating Equations was conducted to determine the robustness of the predictive performance of health education module in increasing screening uptake. Materials and Method: A parallel, single blind, cluster randomized trial was conducted among 15 organizations in Kuantan, Pahang. Intervention group received a complex Health Education Module comprising of group education, practical session on fecal occult blood test usage and WhatsApp group follow-up, while control group received standard colorectal cancer screening brochure. Sensitivity analyses using intention to treat analysis with interaction term, compatibility term, behavioral intention term and key assumption term were performed. Data were imputed and analysed using generalized estimating equation with IBM SPSS version 23. Pooled adjusted odds ratio was calculated using random effect model with inverse variance weighting using RevMan version 3.5. Results: A total of 166 participants from 15 organizations were recruited in the study. Intervention and control group were comparable at baseline (P>0.05). Health Education Module given in intervention group significantly increased the uptake of FOBT by nearly 5 times compared to control group in sensitivity analyses (pooled adjusted OR=4.60, 95% CI=2.65-7.99, I2=47%, P