This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
SARS-CoV-2, the virus that caused the global COVID-19 pandemic, has severely impacted Central Asia; in spring 2020, high numbers of cases and deaths were reported in this region. The second wave of the COVID-19 pandemic is currently breaching the borders of Central Asia. Public health surveillance is necessary to inform policy and guide leaders; however, existing surveillance explains past transmissions while obscuring shifts in the pandemic, increases in infection rates, and the persistence of the transmission of COVID-19.
The goal of this study is to provide enhanced surveillance metrics for SARS-CoV-2 transmission that account for weekly shifts in the pandemic, including speed, acceleration, jerk, and persistence, to better understand the risk of explosive growth in each country and which countries are managing the pandemic successfully.
Using a longitudinal trend analysis study design, we extracted 60 days of COVID-19–related data from public health registries. We used an empirical difference equation to measure the daily number of cases in the Central Asia region as a function of the prior number of cases, level of testing, and weekly shift variables based on a dynamic panel model that was estimated using the generalized method of moments approach by implementing the Arellano-Bond estimator in R.
COVID-19 transmission rates were tracked for the weeks of September 30 to October 6 and October 7-13, 2020, in Central Asia. The region averaged 11,730 new cases per day for the first week and 14,514 for the second week. Infection rates increased across the region from 4.74 per 100,000 persons to 5.66. Russia and Turkey had the highest 7-day moving averages in the region, with 9836 and 1469, respectively, for the week of October 6 and 12,501 and 1603, respectively, for the week of October 13. Russia has the fourth highest speed in the region and continues to have positive acceleration, driving the negative trend for the entire region as the largest country by population. Armenia is experiencing explosive growth of COVID-19; its infection rate of 13.73 for the week of October 6 quickly jumped to 25.19, the highest in the region, the following week. The region overall is experiencing increases in its 7-day moving average of new cases, infection, rate, and speed, with continued positive acceleration and no sign of a reversal in sight.
The rapidly evolving COVID-19 pandemic requires novel dynamic surveillance metrics in addition to static metrics to effectively analyze the pandemic trajectory and control spread. Policy makers need to know the magnitude of transmission rates, how quickly they are accelerating, and how previous cases are impacting current caseload due to a lag effect. These metrics applied to Central Asia suggest that the region is trending negatively, primarily due to minimal restrictions in Russia.
On December 29, 2019, 4 cases of “pneumonia of unknown etiology” were reported in Wuhan, Hubei Province, China [
Central Asia is largely composed of nation states that are former Soviet Union member countries. The Union of Soviet Socialist Republics (USSR) was dissolved in 1991 after controlling the region for 68 years, following a coup d’état during President Gorbachev’s administration. The former Soviet Union left a lasting legacy in the former national republics [
Public health systems in these new nations in Central Asia faced many challenges, including endemic infectious diseases [
Progress in food security [
Food insecurity is linked to environmental conditions caused by overuse of the Aral Sea, which has been depleted by 90% since 1960 to irrigate large areas of land [
Beyond access to food, obesity is becoming more pervasive [
Because Kazakhstan and China share a border, preventative measures in Kazakhstan were established as early as January 6, 2020, enforcing increased border sanitation and monitoring arrivals from China [
Many countries in the region have had disruptions in the labor market. The oil economies in Kazakhstan and Azerbaijan have been negatively impacted [
State governments have attempted to minimize the impact of the virus; complaints about lack of adequate personal protective equipment in Russia were suppressed [
In the electoral authoritarian regime of Azerbaijan, lockdown was established immediately following the first confirmed case of SARS-CoV-2 on February 28. The capital of Azerbaijan instated highly restrictive rules and shut down its border with the Islamic Republic of Iran, where cases were spreading quickly, the next day [
In Azerbaijan, violations of the quarantine mandate were punishable by fines, custodial restraint, and prison time [
While Azerbaijan initially slowed the spread of the virus [
Shocking the international community, Russian President Vladimir Putin announced on August 11 that their country’s health regulator was the first in the world to approve a SARS-CoV-2 vaccine for mass use [
Other than vaccine efforts, Russia has not implemented significant public health measures. During the first wave of infections in March 2020, no financial support was given to small or medium-sized businesses despite instructions for employees to stay home on paid leave [
Without an effective vaccine to prevent COVID-19, Central Asian leaders require an effective SARS-CoV-2 surveillance system that enables their governments to make safe and informed decisions [
Timeline of the COVID-19 pandemic in Central Asia.
To that end, the objective of our research was to use a longitudinal trend analysis study design in concert with dynamic panel modeling and method of moments approaches to correct for existing surveillance data limitations [
This study relies on a longitudinal trend analysis of data collected from the Foundation for Innovative New Diagnostics (FIND) [
Arellano-Bond estimation of difference equations has several statistical advantages: (1) it enables statistical examination of the predictive ability of a model and the validity of the model specification; (2) it corrects for autocorrelation and heteroscedasticity; (3) it has good properties for handling data with a small number of time periods and large number of states; and (4) it corrects for omitted variables and provides a statistical test of correction validity. With these advantages, the method is applicable to ascertaining and statistically validating changes in the evolution of the pandemic within a period of ≤1 week, such as changes in the reproduction rate. See Oehmke et al [
We analyzed the 12 countries that are included in the Central Asia region as defined by the World Bank. The results of the associated regression supporting the weekly surveillance metrics are captured in
As shown in
Arellano-Bond dynamic panel data modeling of the number of daily infections reported by country in Central Asia from September 30 to October 13, 2020.
Variable | Values | |
1-day lag coefficient | 1.075 | <.001 |
7-day lag coefficient | –0.051 | .89 |
Cumulative tests | 0.000016 | .55 |
Shift parameter, week of October 6 | –0.157 | .53 |
Shift parameter, week of October 13 | –0.417 | .12 |
Weekend effecta | –0.009 | .99 |
aWeekend effect: impact of limited testing over the weekend on case counts.
The 7-day lag and shift parameters suggest that there have been no recent changes in disease transmission rates. Additionally, there is no weekend effect or cumulative test effect.
Static and dynamic surveillance metrics for the weeks of October 6 and 13 are reflected in
Static surveillance metrics for the week of September 30 to October 6, 2020.
Country | New COVID-19 cases, n | Cumulative COVID-19 cases, n | 7-day moving average of new cases | Infection rate (per 100,000 persons) | New deaths, n | Cumulative deaths, n | 7-day moving average of the death rate | Death rate (per 100,000 persons) |
Armenia | 406 | 53,083 | 454.57 | 13.73 | 6 | 990 | 4.57 | 0.20 |
Azerbaijan | 143 | 40,931 | 116.00 | 1.43 | 2 | 600 | 1.43 | 0.02 |
Cyprus | 29 | 1876 | 19.00 | 2.42 | 1 | 23 | 0.14 | 0.08 |
Georgia | 549 | 9245 | 482.71 | 14.76 | 4 | 58 | 3.14 | 0.11 |
Kazakhstan | 66 | 108,362 | 64.86 | 0.36 | 21 | 1746 | 3.00 | 0.11 |
Kosovo | 34 | 15,889 | 45.00 | 1.89 | 2 | 635 | 1.43 | 0.11 |
Kyrgyzstan | 164 | 47,799 | 182.43 | 2.54 | 0 | 1066 | 0.29 | 0 |
North Macedonia | 223 | 19,096 | 187.14 | 10.70 | 8 | 768 | 4.43 | 0.38 |
Russia | 11481 | 1,231,277 | 9835.57 | 7.95 | 184 | 21,559 | 157.57 | 0.13 |
Tajikistan | 40 | 10,014 | 41.14 | 0.43 | 0 | 78 | 0.43 | 0 |
Turkey | 1511 | 327,557 | 1469.29 | 1.81 | 55 | 8553 | 60.43 | 0.07 |
Uzbekistan | 397 | 59,343 | 427.00 | 1.18 | 4 | 489 | 3.29 | 0.01 |
Static surveillance metrics for the week of October 7-13, 2020.
Country | New COVID-19 cases, n | Cumulative COVID-19 cases, n | 7-day moving average of new cases | Infection rate (per 100,000 persons) | New deaths, n | Cumulative deaths, n | 7-day moving average of the death rate | Death rate (per 100,000 persons) |
Armenia | 745 | 57,566 | 640.43 | 25.19 | 6 | 1032 | 6.00 | 0.20 |
Azerbaijan | 277 | 42,381 | 207.14 | 2.76 | 3 | 612 | 1.71 | 0.03 |
Cyprus | 83 | 2130 | 36.29 | 6.92 | 0 | 25 | 0.29 | 0.00 |
Georgia | 569 | 12,841 | 513.71 | 15.29 | 9 | 102 | 6.29 | 0.24 |
Kazakhstan | 83 | 108,984 | 88.86 | 0.45 | 22 | 1768 | 3.14 | 0.12 |
Kosovo | 98 | 16,345 | 65.14 | 5.46 | 1 | 649 | 2.00 | 0.06 |
Kyrgyzstan | 343 | 49,871 | 296.00 | 5.31 | 2 | 1092 | 3.71 | 0.03 |
North Macedonia | 80 | 21,193 | 299.57 | 3.84 | 3 | 800 | 4.57 | 0.14 |
Russia | 13,690 | 1,318,783 | 12,500.86 | 9.48 | 240 | 22,834 | 182.14 | 0.17 |
Tajikistan | 37 | 10,297 | 40.43 | 0.40 | 0 | 79 | 0.14 | 0.00 |
Turkey | 1632 | 338,779 | 1603.14 | 1.96 | 62 | 8957 | 57.71 | 0.07 |
Uzbekistan | 323 | 61,642 | 328.43 | 0.96 | 2 | 511 | 3.14 | 0.01 |
Novel surveillance metrics for the week of September 30 to October 6, 2020.
Country | Speeda | Accelerationb | Jerkc | 7-day persistence effect on speedd |
Armenia | 15.4 | 0.4 | 0.3 | –0.6 |
Azerbaijan | 1.2 | 0.1 | 0 | –0.1 |
Cyprus | 1.6 | 0 | –0.1 | –0.1 |
Georgia | 13.0 | 0.9 | –0.2 | –0.4 |
Kazakhstan | 0.4 | 0 | 0 | 0 |
Kosovo | 2.5 | –0.2 | –0.1 | –0.1 |
Kyrgyzstan | 2.8 | 0 | –0.2 | –0.1 |
North Macedonia | 9.0 | 0.8 | 0.5 | –0.3 |
Russia | 6.8 | 0.3 | 0.1 | –0.3 |
Tajikistan | 0.4 | 0 | 0.0 | 0 |
Turkey | 1.8 | 0 | 0.0 | –0.1 |
Uzbekistan | 1.3 | –0.1 | 0.0 | –0.1 |
aSpeed: daily positives per 100,000 persons (weekly average of new daily cases per 100,000 persons).
bAcceleration: day-to-day change in the number of positives per day (weekly average per 100,000 persons).
cJerk: week-over-week change in acceleration per 100,000 persons.
d7-day persistence effect on speed: number of new cases per day per 100,000 persons.
Novel surveillance metrics for the week of October 7-13, 2020.
Country | Speeda | Accelerationb | Jerkc | 7-day persistence effect on speedd |
Armenia | 21.7 | 1.6 | 0.7 | –0.8 |
Azerbaijan | 2.1 | 0.2 | 0.2 | –0.1 |
Cyprus | 3.0 | 0.6 | 0.4 | –0.1 |
Georgia | 13.8 | 0.1 | 0.5 | –0.7 |
Kazakhstan | 0.5 | 0 | 0 | 0 |
Kosovo | 3.6 | 0.5 | 0.3 | –0.1 |
Kyrgyzstan | 4.6 | 0.4 | 0.2 | –0.1 |
North Macedonia | 14.4 | –1.0 | –1.6 | –0.5 |
Russia | 5.2 | 0.1 | 0.0 | –0.2 |
Tajikistan | 8.7 | 0.2 | 0.0 | –0.4 |
Turkey | 0.4 | 0 | 0.0 | 0 |
Uzbekistan | 1.9 | 0 | 0.0 | –0.1 |
aSpeed: daily positives per 100,000 persons (weekly average of new daily cases per 100,000 persons).
bAcceleration: day-to-day change in the number of positives per day (weekly average per 100,000 persons).
cJerk: week-over-week change in acceleration per 100,000 persons.
d7-day persistence effect on speed: number of new cases per day per 100,000 persons.
Comparison of 1-day persistence in the four countries in Central Asia with positive significant positive accelerations for the week of October 6, 2020.
Country | 1-day persistence | |
|
Week of September 30 | Week of October 6 |
Armenia | 16.1 | 21.5 |
Georgia | 13.0 | 16.5 |
North Macedonia | 8.8 | 14.8 |
Russia | 7.0 | 9.1 |
Most populous countries in Central Asia as of 2020.
Country | Population as of 2020, n |
Russia | 145,953,632 |
Turkey | 84,621,255 |
Uzbekistan | 33,469,203 |
Kazakhstan | 18,776,707 |
Russia and Turkey had the highest 7-day moving averages in the region, at 9836 and 1469, respectively, for the week of October 6, and 12,501 and 1603, respectively, for the week of October 13 (
Russia and Turkey also had the highest 7-day moving averages of deaths in the region, with Russia at 157.57 per 100,000 persons for the week of October 6 and Turkey at 60.43. Together, they accounted for approximately 90% of the deaths reported in the region. Up to October 13, the region had reported 38,461 cumulative deaths. For the week of 10/06, North Macedonia and Armenia had the highest death rates per 100,000 persons in the region, at 0.38 and 0.20, respectively. Armenia maintained a death rate of 0.20 per 100,000 persons the following week. Georgia had the highest death rate the week of 10/13 at 0.24 per 100,000 persons, up from 0.11 the previous week.
The 1-day persistence is an indicator of a clustering effect where an event on a particular day causes an increase in the number of cases on adjoining days. As shown in
Largely consistent with infection rates, Armenia, Georgia, and North Macedonia had the highest speed or average of new daily cases per 100,000 persons. During the week of October 6, Armenia had a speed of 15.4, increasing to 21.7 the following week. Georgia had a speed of 13.0, which increased slightly the following week to 13.8. North Macedonia had a speed of 9.0, which increased to 14.4. The region overall had an increase in speed from 4.2 to 5.2.
Speed is best used in conjunction with acceleration and jerk, which can provide further insight into potential pandemic trajectory changes. Four countries in the region had significant positive accelerations for the week of October 6: Georgia at 0.9, North Macedonia at 0.8, Armenia at 0.4, and Russia at 0.3. North Macedonia, Armenia, and Russia also had positive jerks. During the following week, in addition to the highest speed, Armenia had the highest acceleration and jerk in the region.
Weekly SARS-CoV2 trends in Central Asia [
COVID-19 poses a significant threat to the Central Asian region, which is largely composed of former Soviet republics. These countries continue to suffer from food insecurity, high levels of poverty, and variation in health care quality and access as the region continues on its journey of transitioning from a centralized Soviet medical system. The population also suffers from multiple endemic infections, such as HIV/AIDS and tuberculosis. Russia and Turkey comprise the bulk of the population in Central Asia, and these countries are facing growing burdens of chronic disease and some of the highest obesity rates in Europe. Due to the combination of these factors, the region is vulnerable to negative outcomes from the COVID-19 pandemic. To date, the region has seen variations in policy intervention to control the spread of COVID-19 and mitigate outbreaks. Some countries, such as Kazakhstan and Azerbaijan, imposed strict and early lockdowns, while others, such as Russia, imposed more limited interventions.
Metrics tracking the progression of COVID-19 in Central Asia to date have largely been static, including measures such as new cases, cumulative cases, deaths, and 7-day moving averages. These metrics provide a view of the current state of the pandemic but are unable to provide any insight into the change in the speed of the pandemic over time or potential shifts in its trajectory, evolving from controlled spread to rapid growth or vice versa. These metrics also provide limited utility in comparing countries to each other and in analyzing countries with smaller populations. Novel metrics such as speed, acceleration, and jerk help contextualize static metrics and provide a view of trajectory over time, enabling potential anticipation of how the pandemic will evolve in the future.
Considering the static and dynamic metrics, it is apparent that the Central Asian region is trending negatively. The region saw an increase in 7-day moving average of new cases, infection rate, and speed for the week of October 13 compared to the week of October 6, with continued positive acceleration. This trend is largely driven by Russia and Turkey, which together encompass over 70% of the region’s population and showed the highest 7-day moving averages in the region. Russia has the fourth highest speed in the region and continues to experience positive acceleration.
Kazakhstan, the fourth most populous country in the region, had the lowest infection rate for the week of October 6. This is likely due to a continued emphasis on policy interventions to curb the spread of COVID-19. Authorities in Kazakhstan took some of the earliest precautions to prevent infections and continue to strictly enforce pandemic mitigation measures, including halting the easing of restrictions due to the global spike in COVID-19 cases in recent weeks.
Turkey, while contributing a significant portion of the total cases in the region due to its population size, has maintained relatively low infection rates of 1.81 and 1.96 per 100,000 persons for the weeks of October 6 and 13, respectively. Turkey has taken significant precautions to mitigate the spread of the virus, and authorities continue to enforce rigorous mask wearing and social distancing guidelines along with local quarantines when necessary.
Armenia is experiencing uncontrolled spread of COVID-19, with an infection rate of 13.73 per 100,000 persons for the week of October 6 quickly jumping to 25.19, the highest in the region, the following week. The pandemic speed, consistent with the infection rate trajectory, increased from 15.4 to 21.7, with an acceleration increase from 0.4 to 1.6. This change is likely due to the recent lifting of the COVID-19 state of emergency, which allowed the resumption of in-person schooling and international flights, among other activities.
After Armenia, Georgia and Russia had the highest infection rates in the region for the week of October 13. Russia continues to resist implementing interventions to curb the spread of COVID-19, with no mask mandates, capacity caps, or nightlife restrictions. In addition to the significant focus on developing an effective vaccine, there has been limited intervention to manage the spread of COVID-19 in Russia. This policy stance is impacting the trajectory of the region, which is trending negatively with no sign of a reversal in sight.
The rapid evolution, novel outbreaks, and frequently fluctuating trajectory of COVID-19 cannot be adequately assessed using static public health measures alone. Static measures, including the number of new COVID-19 cases, number of cumulative COVID-19 cases, 7-day moving average of new cases, rate of infection, number of new deaths, number of cumulative deaths, 7-day moving average of number of deaths, and death rates, provide a current view of the state of the pandemic. However, these measures do not provide any insight into how the trajectory of the pandemic may change over time.
Generally, the approach to modeling the spread of COVID-19 is to assume there is an underlying contagion model [
Novel surveillance metrics allow for a more nuanced analysis of the COVID-19 pandemic and together with static metrics, can enable policymakers to make informed decisions to control the spread of the pandemic and prevent further outbreaks. Novel dynamic metrics include speed, acceleration, jerk, and 7-day persistence, and they provide potential insight into how the pandemic will evolve in the future.
The analysis of Central Asia using static and novel surveillance metrics suggests that the region is precariously positioned and trending negatively. Russia, the largest country by population, continues to have high infection rates and one of the highest speeds of infection. With no sign of increasing restrictions, it is unlikely that this trend will reverse and that outbreaks in the region will be controlled.
Our data are limited by reporting methods across individual countries. Some countries, such as Turkmenistan, refuse to acknowledge COVID-19 cases. Variation in testing and infrastructure may impact the number of cases reported by other countries. The data are reported at a national level, which does not enable any subnational analysis.
This study is part of a broader research program at Northwestern Feinberg School of Medicine (The Global SARS-CoV-2 Surveillance Project: Policy, Persistence, & Transmission). Novel surveillance metrics, including speed, acceleration, jerk, and 7-day persistence, have been developed by this research program and are being applied to all global regions.
Weekly Cental Asia SARS-CoV-2 statistics by country.
Weekly Central Asia 7-day persistence map.
Weekly Central Asia statistics.
Weekly Central Asia jerk map.
Weekly Central Asia acceleration jerk map.
Foundation for Innovative New Diagnostics
basic reproduction number
Union of Soviet Socialist Republics
This publication was made possible through support provided by Feed the Future through the US Agency for International Development under the terms of Contract No. 7200LA1800003. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the US Agency for International Development.
None declared.