Saudi Arabian Administrative Areas and Neighborhoods in COVID-19 Infections: An Application of 3 X 3 Model

Saudi Arabia has been seriously affected by COVID-19 across various administrative areas. Not only the prominent cities but also upcoming future cities and small townships were affected. This research aims at an analysis of COVID-19 data published by the Ministry of Health of Saudi Arabia to understand effects of broader administrative areas and neighborhoods and its interaction on spread of the epidemic. This research applies a generalized linear model (3 X 3) of administrative areas (major, middle sized and others) and neighborhoods (large, medium sized, and others) on COVID-19 infected cases classifying on a monthly basis from March to November, 2020. A total of 213 neighborhoods of various categories have been affected in the country with variousfrequencies and changes, based on local demographics. More than the broader administrative areas smaller neighborhoods receive signicance and thus the interaction of variables in producing the number of cases: giving lessons for policies, programs and plans of development.


Introduction
COVID-19 spread rapidly in the Kingdom of Saudi Arabia where a population of 34.3 million live in an area of 2.2 square kilometers. The initial rapid rise, peak, and erratic behavior of the epidemic with geographic variations depend heavily on super-spreaders and heterogeneous population characteristics (Eilersen and Sneppen, 2021). Highly populated industrial cities, the epic centres demand epidemic compartmental models to policy interventions distinguishing early implementation, later interventions, and mild interventions to pave way for post-pandemic urban growth strategies to boosting urban potential (Tian et al., 2021;Guaralda, et al., 2020). Such urban units are commonly found everywhere including Saudi Arabia, where spatial architecture, urban inequalities, and local governance patterns differ, which add up to the precarious conditions challenging protection measures (Ziccardi, 2020). Besides, regional comparisons on COVID-19 including mortality are determined by health services delivery, community-level healthcare, testing approaches, and characteristics of surveillance systems enabling opportunities and dynamics of balancing regional development (Signorelli et al., 2020;Guaralda, et al., 2020).
Despite the precautionary and preventive measures adopted in Saudi Arabia, the epidemic spread faster than anticipated, thus crafted political and economic strategies with proactive precautionary measures, thereby atten the epidemic curve by increasing recovery and thus achieving comparatively low case fatality (Ministry of Health. 2020; El Sayed, 2020; Jdaitawi et al., 2020; Al-Otaibi, 2020). Besides, the Kingdom continues decisive bold measures to safeguard population at a great socioeconomic cost geared with swift community action, hospital preparedness, and mitigation measures (Yezli and Khan, 2020; Obeid et al., 2020; Barry et al., 2020 1 ). As a result, the metropolises and future cities are seriously affected despite availability of essential health services (Jdaitawi et al., 2020). Thus, the rapid urbanization, particularly around capital cities undergo rethinking beyond conventional, neoliberal growth strategies to reduce risks of urban growth within national, regional, and local planning, design, and development strategies (Guaralda, et al.,2020).
Cities with high concentration of population and economic activities are the seriously affected locations, hotspots of COVID-19 infections. Thus, dynamics of the pandemic in urban areas are important in establishing the impact of COVID-19, as stated by Shari and Khavarian-Garmsir, (2020). Moreover, Saudi Arabia is a large country in terms of geographic area having vivid climatic and ecological conditions, climatic sensitivity, socioeconomic conditions, seasonal patterns, and population factors play havoc in transmission rate, disease burden, and control policies, as stated by Metelmann Table 1 for the number of affected neighborhoods in each category of region and of the neighborhoods monthwise): thus forming a 3 X 3 analysis model.

Results And Discussions
Considering the theme of this research, data analyzed by means of 3 X 3 ANOVA (univariate general linear model), one-way ANOVA, and mean number of infected persons at various neighborhoods classi ed by the above mentioned criteria of administrative areas and neighborhoods. There is a signi cant F value in case of neighborhoods, on all months, but not in case of administrative areas ( Table  2). As this analysis considered 197 neighborhoods (divided into large cities, medium sized cities, and others including small towns and villages) located at various parts of 13 administrative areas (divided into major, middle sized and other smaller), it could be understood that more than the broader administrative boundaries the locational characteristics matter in case of COVID-19 epidemic. Although, the major administrative areas accounted for the majority of infected cases reported in the country, as a characteristic it plays a role lesser than the more homogenous neighborhoods which is proxy of grassroot level demographics including geography, livelihoods, and proportion of expatriate population, where super-spreaders operate on heterogeneous population, as stated by Eilersen and Sneppen, (2021).  The F value of neighborhood was found to be the highest in May (395.2). Higher F values are reported in November, October, September and August too, which are considered to be months of intense spread of COVID-19. On the other hand, the F value for administrative area remained low between 0.1 and 5.0 throughout the period: insigni cant till May. This pattern was found to be re ected in the interaction also. That means, the two variables operate together to produce differentials on COVID-19 infections. There are reasonable R 2 values, above 0.60 and adjusted R 2 above 0.50 making the tests logical.
There Region. These large cities accounted for a large share of the cases but it declined rapidly over the period (Fig. 1). These metropolises with high community and business activities have higher mitigation potential as part of emergency preparedness associated with health delivery; community outreach; diagnostic and surveillance systems etc., as stated by Signorelli, et al., (2020). During the early months of infection, that is, March and April these major cities had higher proportions above 80 percent but which decreased to 48 percent by the month of November. It is to be examined for the effect of intervention strategy adopted, early or late, that impact effectiveness, besides, the characteristics such as spatial architecture and governance pattern (Tian et al., 2021;Ziccardi, 2020). Saudi Arabia has taken a lead in implementing precautionary preventive measures anticipating the danger of epidemic throughout the country (Ministry of health, 2020). Despite, emergencies and disease spread to all parts with varying intensities. Thus, explaining opportunities and dynamics in regional towns are essential, as stated by Guaralda et al., Here, the impact of major urban components such as environmental quality; socioeconomic impacts; management and governance; and transportation and urban design on COVID-19 spread receives importance (Shari and Khavarian-Garmsir, 2020). On the contrary, other 171 smaller neighborhoods also experienced a rapid increase from 2.1 percent in March to 21.0 percent in November. In short, while the larger cities located at major regions had a month wise decline in reported cases, a faster increase in the upcoming and slower increase in smaller neighborhoods experienced. Probably, this trend is re ected in the interaction results.
Separate effects of administrative area and neighborhood were investigated by means of One-Way ANOVA performed on month-wise COVID-19 cases with administrative areas as well as neighborhoods. The former one is not found to be signi cant at 0.05 level from May onwards whereas the later variable has signi cance at 0.00 level throughout the period since March (Table 3). These results are indicative of the above that more than the broader regions, the smaller geographic units play prominent roles in creating the spread of COVID-19. During the initial stages of infection (March and April), administrative areas did not play signi cant roles, but slowly their roles became clearer and by September, it started playing the prominence. This explains geographic variations in spread of infection along the population heterogeneity.  Thus, there arises a need for rethinking beyond conventional growth strategies of cities in line with growth models and urbanization in the emergency preparedness and epidemic spread accounting for planning, design, and development strategies (Guaralda et al., 2020). One way, COVID − 19 enlightened the development community with the planners and policy makers on transformative actions towards creating resilient and sustainable cities (Shari and Khavarian-Garmsir, 2020.). It is also important for the authorities to be vigilant and evidence informed as a preparation for immediate disease prevention measures and policies (Metelmann, et al., 2021).

Conclusions And Recommendations
The popular concept giving 'importance to grass-root level action plans and interventions give output' proves right in this context of COVID-19 epidemic in Saudi Arabia. The larger administrative areas that divide the country into 13 for the purpose of administration and public policies have lesser effect healthcare interventions focusing on COVID-19 epidemic control. The number of infected cases changed in a haphazard manner, with the factors uncontrolled. But, with the increasing number of infected cases started following a pattern inside administrative areas; thereby gaining signi cance. On the contrary, the smaller neighborhoods have a more important role in changes in the number of infected cases per day.
Moreover, the interaction effect of these two variables received signi cance since July, 2020. Thus, there is an interactive rout for the neighborhoods COVID-19 infections via administrative area dynamics. Thus, the developmental plans, programs, and policies accounting the grass root level demographic dynamics proves to be of value especially in the context of epidemics and other emergencies.

Declarations
Ethics approval and consent to participate As this manuscript is based upon published data, ethics approval and consent to participate are not applicable

Consent for publication
As this manuscript is based upon published data, consent for publication is not applicable. However, the authors express their consent to publish this manuscript in your esteemed journal Proportion of cases by month in various categories of neighborhoods