Olubukola O Olusola-Makinde1, Olusola S Makinde2. 1. Department of Microbiology, Federal University of Technology, Akure, Akure, Ondo, Nigeria. 2. Department of Statistics, Federal University of Technology, Akure, Akure, Ondo, Nigeria.
Abstract
BACKGROUND: Coronavirus Disease 2019 (COVID-19) has been surging globally. Risk strata in medical attention are of dynamic significance for apposite assessment and supply distribution. Presently, no known cultured contrivance is available to fill this gap of this pandemic. The aim of this study is to develop a predictive model based on vector autoregressive moving average (VARMA) model of various orders for gender based daily COVID-19 incidence in Nigeria. This study also aims to proffer empirical evidence that compares incidence between male and female for COVID-19 risk factors. METHODS: Wilcoxon signed-rank test is employed to investigate the significance of the difference in the gender distributions of the daily incidence. A VARMA model of various orders is formulated for the gender based daily COVID-19 incidence in Nigeria. The optimal VARMA model is identified using Bayesian information criterion. Also, a predictive model based on univariate autoregressive moving average model is formulated for the daily death cases in Nigeria. Fold change is estimated based on crude case-fatality risk to investigate whether there is massive underreporting and under-testing of COVID-19 cases in Nigeria. RESULTS: Daily incidence is higher in males on most days from 11 April 2020 to 12 September 2020. Result of Wilcoxon signed-rank test shows that incidence among male is significantly higher than female (p-value < 2.22 × 10-16). White neural network test shows that daily female incidence is not linear in mean (p-value = 0.00058746) while daily male incidence is linear in mean (p-value = 0.4257). McLeod-Li test shows that there is autoregressive conditional heteroscedasticity in the female incidence (Maximum p-value = 1.4277 × 10-5) and male incidence (Maximum p-value = 9.0816 × 10-14) at 5% level of significance. Ljung-Box test (Tsay, 2014) shows that the daily incidence cases are not random (p-value=0.0000). The optimal VARMA model for male and female daily incidence is VARMA (0,1). The optimal model for the Nigeria's daily COVID-19 death cases is identified to be ARIMA (0,1,1). There is no evidence of massive underreporting and under-testing of COVID-19 cases in Nigeria. CONCLUSIONS: Comparison of the observed incidence with fitted data by gender shows that the optimal VARMA and ARIMA models fit the data well. Findings highlight the significant roles of gender on daily COVID-19 incidence in Nigeria.
BACKGROUND: Coronavirus Disease 2019 (COVID-19) has been surging globally. Risk strata in medical attention are of dynamic significance for apposite assessment and supply distribution. Presently, no known cultured contrivance is available to fill this gap of this pandemic. The aim of this study is to develop a predictive model based on vector autoregressive moving average (VARMA) model of various orders for gender based daily COVID-19 incidence in Nigeria. This study also aims to proffer empirical evidence that compares incidence between male and female for COVID-19 risk factors. METHODS: Wilcoxon signed-rank test is employed to investigate the significance of the difference in the gender distributions of the daily incidence. A VARMA model of various orders is formulated for the gender based daily COVID-19 incidence in Nigeria. The optimal VARMA model is identified using Bayesian information criterion. Also, a predictive model based on univariate autoregressive moving average model is formulated for the daily death cases in Nigeria. Fold change is estimated based on crude case-fatality risk to investigate whether there is massive underreporting and under-testing of COVID-19 cases in Nigeria. RESULTS: Daily incidence is higher in males on most days from 11 April 2020 to 12 September 2020. Result of Wilcoxon signed-rank test shows that incidence among male is significantly higher than female (p-value < 2.22 × 10-16). White neural network test shows that daily female incidence is not linear in mean (p-value = 0.00058746) while daily male incidence is linear in mean (p-value = 0.4257). McLeod-Li test shows that there is autoregressive conditional heteroscedasticity in the female incidence (Maximum p-value = 1.4277 × 10-5) and male incidence (Maximum p-value = 9.0816 × 10-14) at 5% level of significance. Ljung-Box test (Tsay, 2014) shows that the daily incidence cases are not random (p-value=0.0000). The optimal VARMA model for male and female daily incidence is VARMA (0,1). The optimal model for the Nigeria's daily COVID-19 death cases is identified to be ARIMA (0,1,1). There is no evidence of massive underreporting and under-testing of COVID-19 cases in Nigeria. CONCLUSIONS: Comparison of the observed incidence with fitted data by gender shows that the optimal VARMA and ARIMA models fit the data well. Findings highlight the significant roles of gender on daily COVID-19 incidence in Nigeria.
Authors: Christian Drosten; Stephan Günther; Wolfgang Preiser; Sylvie van der Werf; Hans-Reinhard Brodt; Stephan Becker; Holger Rabenau; Marcus Panning; Larissa Kolesnikova; Ron A M Fouchier; Annemarie Berger; Ana-Maria Burguière; Jindrich Cinatl; Markus Eickmann; Nicolas Escriou; Klaus Grywna; Stefanie Kramme; Jean-Claude Manuguerra; Stefanie Müller; Volker Rickerts; Martin Stürmer; Simon Vieth; Hans-Dieter Klenk; Albert D M E Osterhaus; Herbert Schmitz; Hans Wilhelm Doerr Journal: N Engl J Med Date: 2003-04-10 Impact factor: 91.245
Authors: Gbenga J Abiodun; Olusola S Makinde; Abiodun M Adeola; Kevin Y Njabo; Peter J Witbooi; Ramses Djidjou-Demasse; Joel O Botai Journal: Int J Environ Res Public Health Date: 2019-06-05 Impact factor: 3.390
Authors: Catherine Gebhard; Vera Regitz-Zagrosek; Hannelore K Neuhauser; Rosemary Morgan; Sabra L Klein Journal: Biol Sex Differ Date: 2020-05-25 Impact factor: 5.027
Authors: Folajimi O Shorunke; Emmanuel C Okolocha; Grace S Kia; Aishat Usman; Oluseyi Akano; Emmanuel J Awosanya Journal: One Health Outlook Date: 2022-03-24