SARS-CoV-2 Surveillance in the Middle East and North Africa: Longitudinal Trend Analysis

Introduction

Background

SARS-CoV-2, the novel coronavirus that causes the disease COVID-19, first presented in December 2019 in Wuhan City, China, and was declared a public health emergency of international concern on January 30, 2020, as it spread quickly around the globe through human-to-human transmission [1-4] . According to the World Health Organization (WHO), the first reports of COVID-19 in the Middle East and North Africa (MENA) included 8 cases in the United Arab Emirates between January 29, 2020, and February 9, 2020 [5] (Figure 1). It has since spread to every country in the region [6]. MENA is divided by large wealth disparities and regional conflicts, increasing risks to the COVID-19 pandemic that has no political or religious agenda [6]. Despite varying levels of health system preparedness in MENA countries, early implementation of strict containment measures helped to limit the spread of SARS-CoV-2. As of October 28, the WHO reported 43,540,739 confirmed cases of COVID-19, including 1,160,650 deaths worldwide [7]. There have been 2,982,597 confirmed cases in MENA, resulting in 75,737 deaths [7].

Many countries use a combination of containment and mitigation strategies, including isolation of cases, contact tracing, social distancing, border closures, masking, hand and surface hygiene, and travel restrictions [8]. In the absence of a vaccine, worldwide systemic public health surveillance that can reliably track cases and identify where outbreaks will occur is needed to inform disease control policy for COVID-19 prevention [4,9]. Global SARS-CoV-2 surveillance necessitates dividing the globe into separately surveyed regions. The World Bank, an international financial institution that aims to reduce poverty and increase sustainable prosperity in nations throughout the globe, divides the world into regions based on shared geographic, cultural, and historical qualities. The MENA region comprises Bahrain, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Oman, Qatar, Saudi Arabia, Syria, the United Arab Emirates, Yemen, Algeria, Djibouti, Egypt, Libya, Morocco, and Tunisia [10].

Responses to COVID-19 in MENA have ranged from restrictive temporary lockdowns to denial and lack of organization [6,11]. Lessons learned from the severe acute respiratory syndrome (SARS), the Middle East respiratory syndrome (MERS), and the 2009 H1N1 outbreaks inform the current pandemic [12]. For example, in Saudi Arabia, the Ministry of Health established the Command and Control System and the Saudi Center for Disease Control and Prevention shortly after the rise of MERS [12]. Some countries repurposed surveillance resources to encourage COVID-19 containment. This is best illustrated in the United Arab Emirates where the “Oyoon” (“eyes” in Arabic) surveillance camera program, originally developed to track and prevent crime, was repurposed to check the temperatures of those passing by and to ensure people are social distancing [13].

‎View this figure

Economics

Countries in MENA face a dual crisis from the COVID-19 pandemic and the collapse of oil prices [14]. These concurrent crises are exacerbated by structural economic challenges including large and inefficient public sectors, uncompetitive business environments, high youth and female unemployment, governance challenges, and regional conflicts [14]. Given these challenges, the World Bank released a report recommending a coordinated regional MENA trade integration framework that lays the foundation for global value chain integration and thus helps MENA’s economic prospects [14]. Income disparities abound in MENA, which is home to the wealthy Persian Gulf nations and yet is one of the only regions in the world where extreme poverty has been rising since 2011, particularly in regions with violent conflicts [6,15]. Thus, as countries in MENA combat the COVID-19 pandemic, they do so in the context of falling oil prices and disparities in wealth and political stability.

Violent conflicts significantly weakened the health infrastructure in several countries across MENA, resulting in poor health worker capacity [12,16]. However, outside of conflict regions, MENA countries have some of the lowest proportion of health workers in the world with Egypt, Morocco, Iraq, Yemen, and Djibouti subject to low levels of doctors and nurses [11]. The health system capacity in MENA is facility based, which limits the surge capacity needed to respond to a crisis such as the COVID-19 pandemic [11]. Some MENA countries have limited public health infrastructure, including Egypt, Iraq, Djibouti, and Yemen that spend less than 5% of their government funding on health [11,17].

Middle Eastern populations have a high disease burden, which suggests a reduction in routine health service utilization results in increased mortality, a trend expected to hold true for COVID-19 [11,12]. Food insecurity poses a significant public health concern in countries with high refugee per capita concentration, such as Lebanon and Jordan [18-20]. Refugee populations are particularly vulnerable to COVID-19 as they are often concentrated in densely populated camps with poor hygiene measures and fragmented access to health care [12].

On August 4, 2020, a deadly warehouse explosion, caused by 2750 tons of ammonium nitrate stored at Beirut’s port, killed hundreds, wounded and displaced thousands, left 5 hospitals in the area either nonfunctional or only partially functional, and destroyed 17 containers of WHO essential medical supplies [21-23]. The seeming carelessness with which this destructive material was stored for 6 years fueled public outrage and antigovernment protests throughout the city [24]. Since then, Lebanon has seen a 220% increase in COVID-19 cases [25]. As countries ease restrictions and a second wave of COVID-19 threatens to undo progress, it is particularly important to accurately track cases across MENA [26]. This is evident in countries such as Israel [27]. Ideally, advanced tracking and estimation of COVID-19 transmission will inform the implementation of public health policies. Though general public health surveillance is helpful and provides a proxy of the pandemic, surveillance data suffer from significant bias due to undercounts, reporting delays, testing errors, dearth of testing, asymptomatic carriers, and other types of data contamination [7,28-35]. In fact, surveillance systems are predicated on the fact that they tend to include the more severe cases and suffer from incomplete case ascertainment by missing the mild cases and undiagnosed infections and deaths.

Objective

The objective of our research is to provide additional surveillance metrics to add to the public health arsenal using dynamic panel modeling and method of moments to minimize sampling bias by measuring significant negative or positive weekly shifts in the increase, decrease, or plateaued transmission of SARS-CoV-2. We will apply novel indicators derived specifically to inform policy makers about the COVID-19 pandemic [3,36], including weekly shifts in the pandemic, speed, acceleration and deceleration, jerk, and 7-day persistence. These dynamic metrics will better inform public health policy when combined with standard surveillance measures.

Methods

This study relies on a longitudinal trend analysis of data collected from the Foundation for Innovative New Diagnostics (FIND) [37]. FIND compiles data from multiple sources across individual websites, statistical reports, and press releases; data for the most recent 8 weeks were accessed from the GitHub repository [38]. This resulted in a panel of 19 MENA countries with 30 days in each panel (n=570). An empirical difference equation was specified in which the number of new positive cases in each country on each day is a function of prior numbers of cases, level of testing, and weekly shift variables that measure whether the contagion was growing faster, at the same pace, or slower than the previous weeks. This resulted in a dynamic panel data model that was estimated using the generalized method of moments (GMM) approach by implementing the Arellano-Bond estimator in R (The R Foundation) [3,36,39].

Results

MENA Regression Results

Regression results are presented for 17 MENA countries in Table 1. Weekly surveillance data in Tables 2-7 are based on these regressions.

The regression Wald statistic was significant (χ²₅=859.5, P<.001). The Sargan test was not significant, failing to reject the validity of overidentifying restrictions (χ²₂₉₄=16, P=.99)

Tables 2 and 3 contain standard surveillance metrics for the weeks of October 5-11 and October 12-18. These metrics include new daily COVID-19 cases, cumulative COVID-19 cases, 7-day moving averages, rate of infection, new deaths, cumulative deaths, 7-day moving average of death rates, and rates of death per 1,000,000 population.

In Table 3, we see the countries with the highest cumulative caseload of the novel coronavirus, which includes Iran, Iraq, Saudi Arabia, and Israel with 530,380, 426,634, 342,202, and 303,109 cases, respectively. To eliminate the week-end effect or delayed reporting, we look at 7-day moving average to understand the 7-day average number of cases. In the second week of our study, October 12-18, Iran reported 4329, Israel reported 1802, Iraq reported 3472, and Morocco reported 3033 new cases per day (weekly averages). The next closest country in terms of average new cases only has a third of those cases. We find that many of the smaller countries in MENA have higher infection rates than those countries with the highest caseload. For example, Oman had 33.3 new infections per 100,000 population while Bahrain had 12.1, Libya had 14, and Lebanon had 14.6. In order of most to least number of cumulative deaths since January 2020, Iran, Iraq, Egypt, and Saudi Arabia had 30,375, 10,254, 6120, and 5185, respectively. These countries are outliers when comparing them to the rest of the region, which all have less than 3000 cumulative deaths

The novel surveillance metrics are presented in Tables 4 and 5 and Figure 2 [40] and Multimedia Appendices 1-5.

‎View this figure

Overall, the MENA region was more stable than other global regions, but metrics between October 5-18 indicate pending growth. The speed of new infections decelerated during the week of October 5-11 and remained stable during the week of October 11-18. In addition, in the latter week, there was a slight negative jerk. The persistence rate slightly decreased from 3.11 to 3.05 per 100,000 population over the study period, which directly measures those new cases that are statistically related to the number of new infections 7 days earlier. While the surveillance metrics for the MENA region as a whole are promising, they are averages and thus we must look to those countries with increasing rates of speed, acceleration, jerk, and persistence to understand which countries have outbreaks and less control of the pandemic. In Table 5, we note that Bahrain, Israel, and Lebanon decreased in speed or remained level between October 5 and 18. Jordan and the United Arab Emirates had significant increases in rates of infection. Tunisia had the best shift in the pandemic that resulted in a decrease in speed week over week. In addition, the rates reversed course and decelerated from October 5-11 to October 12-18, and Tunisia experienced a negative jerk in infection rates 2 weeks in a row, indicating a strong downward trend in COVID-19 infections. For Israel, the 7-day persistence effect for the week of October 12-18 (21 new cases) slightly overpredicts the speed or the actual number of new cases (n=20); coupled with a declining persistence rate relative to the prior week of October 5-11, this is indicative of a rapid descent from a recent period of explosive growth. On the other end of the spectrum, Bahrain had positive speed, acceleration, and jerk, indicating the outbreak is increasing and is leading toward a worsening pandemic.

Israel, Bahrain, Lebanon, and Oman had the highest rate of COVID-19 persistence, which is the number of new infections statistically related to new infections from 7 days ago. These 3 countries had higher persistence rates 2 weeks in a row. Finally, Egypt has the highest population in MENA (Table 7) and the potential for growth but has remained stable across the 2 weeks of this study, indicating that COVID-19 is not accelerating.

Discussion

Principal Findings

Analysis at the country level indicates there are some nations that should increase their public health efforts to gain control of the COVID-19 pandemic. Iran, Israel, Iraq, and Morocco had the highest reported 7-day average per 100,000 population, which is significantly higher than other nations in the region. Iran, Iraq, Saudi Arabia, and Israel had the highest caseload at the end of October 18. Looking toward the future, Jordan, Iraq, and the United Arab Emirates have the fastest acceleration in new COVID-19 infections while Bahrain and Israel have the largest upwards jerk in infections, which can lead to explosive growth. Iran began the pandemic with explosive growth but during the last week of this study ending on October 18, Iran’s acceleration rate has leveled off and its jerk has reversed lower; however, given the number of new cases and population size, the country could easily flare up in new outbreaks, especially considering the second wave of COVID-19 infections has just begun.

Limitation

Data are limited by granularity and collection method. Data were collected at the country level, which precludes local analysis of surveillance trends. Moreover, data collection mechanisms differ by country and may even differ by region within a given country. These different methods lead to week-end effects, missing data points, and other contamination.

Comparison With Prior Work

This study is part of a broader research program at Northwestern University Feinberg School of Medicine, The Global SARS-CoV-2 Surveillance Project: Policy, Persistence, & Transmission. This research program developed novel surveillance metrics to include rates of speed, acceleration, jerk, and 7-day persistence [3,36]. We have also derived surveillance metrics for all global regions.

Conclusion

Static and dynamic public health surveillance tools provide a more complete picture of pandemic progression across countries and regions. While static measures capture data at a given point in time, like infection rates and death rates, they are less successful at assessing population health over a period of weeks or months. By including speed, acceleration, jerk, and 7-day persistence, public health officials may design policies with an eye to the future.

MENA countries with the highest risk all shared a number of characteristics according to the surveillance data. There was a definite positive shift between October 5-11 and October 12-18. Iran, Iraq, Saudi Arabia, and Israel all demonstrated the highest numbers of cumulative infections, acceleration, jerk, and 7-day persistence rates. Looking ahead, policy makers in these countries and the region at large should be concerned about growth in the already substantial number of cases over the short term. Given the substantial 7-day persistence rates of Israel, Bahrain, and Lebanon, it is imperative that efforts be made to target super spreader events. Analysis of subsequent surveillance data using both static and dynamic tools can help confirm the efficaciousness of new policies.